World Psychiatry June 2017

WPS-C1SP 4/26/17 7:54 AM Page 1
WPA
World Psychiatry
WORLD PSYCHIATRY
OFFICIAL JOURNAL OF THE WORLD PSYCHIATRIC ASSOCIATION (WPA)
Volume 16, Number 2 June 2017
EDITORIALS Shared decision making in mental health care 158

settings: perspective, purpose and practice
Shared decision making: everyone wants it, 117 S. TSE
so why isnt it happening?
Incorporating shared decision making in mental 160
A. COULTER
health care requires translating knowledge
Migration and psychosis: our smoking lung? 119 from implementation science
Vol. 16, No. 2, pp. 117-224

J.B. KIRKBRIDE I. SCHOLL, P.J. BARR
Mental health shared decision making 161
SPECIAL ARTICLES in the US
Etiology in psychiatry: embracing the reality 121 R.E. DRAKE
of poly-gene-environmental causation of
mental illness RESEARCH REPORTS
R. UHER, A. ZWICKER Prevalence, incidence and mortality from 163
The contemporary refugee crisis: an overview 130 cardiovascular disease in patients with pooled
of mental health challenges and specific severe mental illness: a large-scale
D. SILOVE, P. VENTEVOGEL, S. REES meta-analysis of 3,211,768 patients
and 113,383,368 controls
PERSPECTIVES C.U. CORRELL, M. SOLMI, N. VERONESE ET AL
Has the rising placebo response impacted 181
The clinical relevance of appraisals of psychotic 140 antidepressant clinical trial outcome? Data
experiences from the US Food and Drug Administration
P.A. GARETY, A. HARDY 1987-2013
Mating, sexual selection, and the evolution 141 A. KHAN, K.F. MAR, J. FAUCETT ET AL
of schizophrenia Risk of suicide, deliberate self-harm and 193
M. DEL GIUDICE psychiatric illness after the loss of a close relative:
Validity and utility of the general factor 142 a nationwide cohort study
of psychopathology M.-B. GULDIN, M.I.S. KJAERSGAARD, M. FENGER-GRN
B.B. LAHEY, R.F. KRUEGER, P.J. RATHOUZ ET AL ET AL
Neuroticism is a fundamental domain of personality 144

with enormous public health implications
REAPPRAISAL
FEBRUARY 2017
T.A. WIDIGER, J.R. OLTMANNS A critique of the ultra-high risk and transition 200
paradigm
FORUM SHARED DECISION MAKING J. VAN OS, S. GULOKSUZ
IN MENTAL HEALTH CARE
INSIGHTS
Implementing shared decision making 146
in routine mental health care Treatment of people at ultra-high risk 207
M. SLADE for psychosis
A.R. YUNG
Commentaries Persistent persecutory delusions: the spirit, 208
Shared decision making: a consideration of 154 style and content of targeted treatment
historical and political contexts D. FREEMAN, F. WAITE
G. MEADOWS Does neuroimaging have a role in predicting 209
outcomes in psychosis?
Involvement in decision making: the devil is 155
P. MCGUIRE, P. DAZZAN
in the detail
R. MCCABE The role of expectations in mental disorders 210
and their treatment
Psychiatric practice: caring for patients, 156 W. RIEF, J.A. GLOMBIEWSKI
collaborating with partners, or marketing to
consumers?
D.J. STEIN LETTERS TO THE EDITOR 212
Common sense alone is not enough 157 WPA NEWS 221
S. PRIEBE
IMPACT FACTOR: 20.205 ISSN 1723-8617

WPS-C2-C3 4/19/17 11:15 AM Page 1
The World Psychiatric Association (WPA) World Psychiatry
The WPA is an association of national psychiatric societies World Psychiatry is the official journal of the World
aimed to increase knowledge and skills necessary for work in Psychiatric Association. It is published in three issues per year
the field of mental health and the care for the mentally ill. Its and is sent free of charge to psychiatrists whose names and
member societies are presently 138, spanning 118 different addresses are provided by WPA member societies and sections.
countries and representing more than 200,000 psychiatrists. Research Reports containing unpublished data are welcome
The WPA organizes the World Congress of Psychiatry every for submission to the journal. They should be subdivided into
three years. It also organizes international and regional con- four sections (Introduction, Methods, Results, Discussion).
gresses and meetings, and thematic conferences. It has 72 sci- References should be numbered consecutively in the text and
entific sections, aimed to disseminate information and pro- listed at the end according to the following style:
mote collaborative work in specific domains of psychiatry. It 1. Cuijpers P, Sijbrandij M, Koole SL et al. Adding psychother-
has produced several educational programmes and series of apy to antidepressant medication in depression and anxi-
books. It has developed ethical guidelines for psychiatric prac- ety disorders: a meta-analysis. World Psychiatry 2014;13:
tice, including the Madrid Declaration (1996). 56-67.
Further information on the WPA can be found on the web- 2. McRae TW. The impact of computers on accounting.
site www.wpanet.org. London: Wiley, 1964.
3. Fraeijs de Veubeke B. Displacement and equilibrium mod-
els in the finite element method. In: Zienkiewicz OC,
WPA Executive Committee
Hollister GS (eds). Stress analysis. London: Wiley, 1965:145-
President D. Bhugra (UK)
97.
President-Elect H. Herrman (Australia)
All submissions should be sent to the office of the Editor.
Secretary General R.A. Kallivayalil (India)
Secretary for Finances A. Soghoyan (Armenia)
Editor M. Maj (Italy).
Secretary for Meetings M. Takeda (Japan)
Editorial Board D. Bhugra (UK), H. Herrman (Australia), R.A.
Secretary for Education E. Belfort (Venezuela)
Kallivayalil (India), A. Soghoyan (Armenia), M. Takeda (Japan), E.
Secretary for Publications M. Riba (USA)
Belfort (Venezuela), M. Riba (USA), A. Javed (UK/Pakistan).
Secretary for Sections A. Javed (UK/Pakistan)
Advisory Board H.S. Akiskal (USA), R.D. Alarcn (USA), J.A.
Costa e Silva (Brazil), J. Cox (UK), M. Jorge (Brazil), H. Katschnig
WPA Secretariat (Austria), F. Lieh-Mak (Hong Kong-China), F. Lolas (Chile), J.E.
Geneva University Psychiatric Hospital, 2 Chemin du Petit Bel- Mezzich (USA), D. Moussaoui (Morocco), P. Munk-Jorgensen
Air, 1225 Chne-Bourg, Geneva, Switzerland. Phone: (Denmark), F. Njenga (Kenya), A. Okasha (Egypt), J. Parnas
+41223055737; Fax: +41223055735; E-mail: wpasecretariat@ (Denmark), V. Patel (India), P. Ruiz (USA), N. Sartorius
wpanet.org. (Switzerland), A. Tasman (USA), S. Tyano (Israel), J. Zohar
(Israel).
Office of the Editor Department of Psychiatry, University of

Naples SUN, Largo Madonna delle Grazie, 80138 Naples,
Italy. Phone: +390815666502; Fax: +390815666523; E-mail:
majmario@tin.it.
Acknowledgement
This publication has been partially supported by an unrestricted
educational grant from Janssen-Cilag SpA,
which is hereby gratefully acknowledged.
World Psychiatry is indexed in PubMed, Current Contents/Clinical Medicine, Current Contents/Social 2017 by WPA Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons
and Behavioral Sciences, Science Citation Index, and EMBASE. or property as a matter of products liability, negligence or otherwise, or from any use or oper-
ation of any methods, products, instructions or ideas contained in the material herein.
All back issues of World Psychiatry can be downloaded free of charge from the PubMed system Because of rapid advances in the medical sciences, in particular, independent verification of
(http://www.pubmedcentral.nih.gov/tocrender.fcgi?journal=297&action=archive). diagnoses and drug dosages should be made.
EDITORIAL
Shared decision making: everyone wants it, so why isnt it happening?

M. Slade1 makes a strong case, in this issue of the journal, as anticipatory, negotiated discussions to clarify patients goals
for a wider adoption of shared decision making (SDM), while and priorities, agree realistic objectives, solve specific prob-
acknowledging the particular problems that must be overcome lems, identify relevant sources of support, document agreed
if it is to become the dominant mode in mental health care. I action plans, and implement these, including follow-up and
believe the arguments for this approach are even more com- review. Nineteen trials met our inclusion criteria. These showed
pelling than he demonstrates, but implementation remains a evidence of small, but beneficial effects on emotional health
key challenge. (depression), physical health (blood glucose, blood pressure),
It is true, as Slade argues, that evidence of the impact of and self-efficacy (self-management knowledge and skills). The
SDM on clinical outcomes in psychiatry is mixed, but the wid- effects were greater when all stages of the care planning cycle
er body of SDM research leads to a more positive assessment. were completed, when the process included more contacts over
Most of the studies he cites were primarily concerned to a longer time period, when it was fully integrated with routine
evaluate the use of specific tools to inform one-off treatment care, and when both clinicians and patients were well sup-
decisions, such as patient decision aids. These are information ported. We rated this evidence as promising, albeit not yet con-
packages designed to inform patients about their treatment clusive, but we hope we have pointed the way towards a more
options and help them determine which they would prefer. sophisticated evaluation of SDM and its effects than simply
They take a variety of forms, spanning from one-page sheets, looking at the impact of patient decision aids on one-off treat-
more detailed leaflets or computer programmes, through to ment decisions.
DVDs and interactive websites. Some are designed for use by We need to give serious consideration to the issue of how
patients at home, while others are intended to guide discus- conclusive evidence has to be before we attempt to implement
sions in medical consultations. They are not absolutely essen- it. If patients want it, the ethical case is strong, there is evi-
tial for SDM but, by packaging evidence-based information in dence of some beneficial effects and no likelihood of harm, is
an accessible form, they certainly make it easier. that sufficient? Most of those arguing for SDM base their case
A recently-updated Cochrane review across all conditions, on ethical justifications rather than clinical ones patients do
not just mental health, combined results from 115 randomized want it, it is important to respect their autonomy (right to be
controlled trials of patient decision aids, most of which focus- informed and involved in decisions that affect them), and
ed on discrete choices and decisions made at a single point in it also promotes beneficence (balancing benefits, risks and
time2. The review found that use of decision aids led to im- costs), and non-maleficence (avoiding harm)4. As Slade dem-
provements in patients knowledge, more participation, more onstrates, these arguments may be somewhat more nuanced
accurate risk perceptions, fewer people remaining undecided, in the case of people with psychosis or those who lack mental
and greater comfort with decisions. However, none of these capacity, but few would argue that people with mental health
trials had looked at whether consultations exhibited the full problems should not be given the opportunity to shape their
range of SDM characteristics (defining the problem, outlining own care whenever feasible.
the options, checking understanding, eliciting values, support- In the UK, ethical and legal guidance from various authorities,
ing deliberation, reaching mutual agreement) and few had including the UK Supreme Court5, the General Medical Council6,
looked at longer term clinical outcomes. In other words, these and the Mental Capacity Act7, is now fully aligned: SDM should
studies looked at only one component of this highly complex be the default and very few exceptions are permitted. Despite
intervention. this, as Slades paper makes clear, around half of mental health
These decision aid trials tell us something about the poten- service users said they were not involved in treatment decisions
tial benefits of SDM, but this is only part of the story. SDM to the extent they wanted to be8. Why then has it proved so diffi-
involves a conversation, or more likely a series of conversa- cult to secure widespread implementation of SDM?
tions, between patients and clinicians: it is a relationship and I agree with Slades contention that the barriers are as much
a process, not a tool or a one-off event. This is especially perti- attitudinal as practical and organizational. Commonly voiced
nent in the case of long-term conditions, such as most mental objections include concerns about lack of time, lack of skills,
health problems. Relevant outcomes may include physical and lack of resources and misapprehensions about patients ability
emotional effects, subjective health status, knowledge and to make appropriate choices, all of which act as powerful
understanding, self-management capabilities, treatment bur- disincentives to change practice. Most of these perceptions are
den and resource use, as well as experience of the decision- not supported by the evidence. For example, SDM does not
making process. necessarily require longer consultations than more traditional
We looked at these outcomes in another Cochrane review forms of decision making2, and several studies have shown
focused on people with long-term conditions3. We searched that it is possible to inform and engage patients from all ages,
for randomized controlled trials that had evaluated personal- walks of life and educational backgrounds, with benefits
ized care planning (SDM by another name), which we defined accruing to all, including those with low health literacy9.
World Psychiatry 16:2 - June 2017 117

A comprehensive strategy is required to promote wider uptake Quality-assured patient decision aids should be made avail-
of SDM. In a recent review of experience in various SDM dem- able at specific decision points via electronic medical records, so
onstration sites in North America and Europe, we described ten that they are readily accessible during clinical consultations. Ap-
components that are required to encourage widespread adop- propriate patient reported outcome measures (PROMs) should
tion10: a) research evidence showing that it can be effective in a be used in routine care as a feedback loop to check that patients
specific clinical or local context; b) medical leadership willing are actively engaged and receive treatments that reflect their
to encourage it; c) demand for SDM from patient leaders and goals and preferences.
organizations; d) incentives for clinicians to change their prac- The tensions that Slade outlines are real, but so is the need to
tice ethical, financial or professional; e) training for clinical work together to find ways to overcome them to ensure the deliv-
staff in SDM and risk communication skills, plus support and ery of appropriate, efficient and effective mental health care.
supervision for practising and maintaining these competen-
cies; f) availability of patient decision aids; g) integration of Angela Coulter
Nuffield Department of Population Health, University of Oxford, Oxford, UK
patient decision aids into electronic medical record systems; h)
institutional support for developing and updating patient deci- 1. Slade M. World Psychiatry 2017;16:146-53.
sion aids; i) certification scheme to assure the quality of patient 2. Stacey D, Legare F, Col NF et al. Cochrane Database Syst Rev 2014;1:
CD001431.
decision aids; j) validated outcome measures to monitor the 3. Coulter A, Entwistle VA, Eccles A et al. Cochrane Database Syst Rev 2015;3:
extent to which patients feel informed and involved in deci- CD010523.
sions about their care, plus feedback to enable clinicians to 4. Stiggelbout AM, Weijden TV, Wit MP et al. BMJ 2012;344:e256.
5. UK Supreme Court. Montgomery v Lanarkshire Health Board Scotland.
monitor progress. UKSC 11, 2015.
All patients, including users of mental health services, should 6. General Medical Council. Consent: patients and doctors making decisions
be encouraged and supported to prepare themselves for an together. London: General Medical Council, 2008.
7. Mental Capacity Act 2005. http://www.legislation. gov.uk/.
active role in treatment selection. SDM theory, skills and compe- 8. Care Quality Commission. 2015 community mental health survey. Statistical
tencies (risk communication, options appraisal, goal setting, care release. London: Care Quality Commission, 2015.
planning and outcomes assessment) should be taught in medical 9. Durand MA, Carpenter L, Dolan H et al. PLoS One 2014;9:e94670.
10. Coulter A, Harter M, Moumjid-Ferdjaoui N et al. Int J Pers Cent Med 2015;
schools, in post-registration training, and in continuing profes- 5:9-14.
sional development, aligned with support for self-management
and patient engagement. DOI:10.1002/wps.20407
118 World Psychiatry 16:2 - June 2017

EDITORIAL
Migration and psychosis: our smoking lung?

To read the history of humankind is to read a history of in agreeing on a specific psychotic diagnosis, but this does not
migration. From the first human exoduses out of Africa, to appear to be racially biased6. Further, few modern epidemio-
Greek and Roman empires which sought territorial expansion, logical studies rely solely on clinician-rated diagnoses to mea-
to the Ming dynastys pioneering voyages of exploration, to the sure outcomes, instead using carefully operationalized criteria
flight of ethnic, religious, political and sexual minorities escap- to reach standardized diagnoses3,7. Finally, in the UK and else-
ing persecution from various authoritarian regimes or internal where, the ethnic composition within clinical psychiatry is
conflicts, to the economic migrants from continental Europe, increasingly diverse, far from the monochromatic contrast
Asia, the Middle East, and South and Central America who that implicitly surrounds the misdiagnosis debate. In a recent
sought better lives for themselves and their families on new study, for example, which also found elevated rates of psychot-
continents, migration is arguably the defining feature of a sin- ic disorder in BME groups in rural England7, operationalized
gular human experience that binds our past, present and diagnoses were made by a panel of psychiatrists from over 13
future. The drivers and consequences of migration also leave different ethnic backgrounds.
indelible marks on the history of humankind. Perhaps in equal Further new empirical data offer important directions. For
measure, they result in leaps forward for civilization enriching example, raised rates do not seem to be entirely attributable to
cultural, social, genetic and economic diversity and human socioeconomic differences between BME groups and the major-
development and pockmarks which serve to remind us of the ity population8. Other recent research, from Sweden, has dem-
seemingly ceaseless bounds of human savagery and brutality onstrated that refugee migrants are at considerably elevated risk
(see also Silove et al1 in this issue of the journal). of non-affective psychotic disorders compared with both the
To a psychiatric epidemiologist, migration is arguably asso- Swedish-born population and, importantly, other migrants from
ciated with one of the defining public health inequalities of the the same regions of origin4. The implication is that severe expo-
last 100 years: that certain migrants, their children, and their sure to pre-migratory adversities, including war, famine and
childrens children are as much as 10 times more likely to meet persecution, or the hazards involved in the transitory process
diagnostic criteria for psychotic disorder than the majority of migration itself, may be aetiologically relevant to psychosis
(usually white Caucasian) population in a given setting2. The risk. Exposure to other severely traumatic migration-related
exact magnitude of this risk varies, depending on the given experiences, such as witnessing genocide8, also increases schizo-
migrant group and setting in which the study is conducted. In phrenia risk. Nonetheless, these data would not explain why
the UK, for example, psychosis risk ranges from slight increases elevated rates persist in successive generations following the
(of 1.5 or less) for white migrants, to 2-4 times greater risk for index immigration event. Other factors must be relevant,
people of Pakistani and Bangladeshi origin, and up to 10 times possibly including experiences of racism and discrimination,
higher rates amongst black Caribbean and African groups3. although further research is needed on this issue.
Elsewhere, elevated risk also follows historical migration flows, We also require more integration of observational data with
such as amongst the Surinamese and Moroccan populations in sociological, ethnographic, experimental psychology and neu-
the Netherlands2, or East African migrants to Sweden4. Emerg- roscience research to shed light on the possible pre-, peri- and
ing research from countries which have experienced unprece- post-migratory factors that increase psychosis risk amongst
dented contemporary immigration pressures5 also shows that BME groups. A recent study from social neuroscience, for
incidence rates are elevated amongst migrant groups. example, suggests that healthy volunteers from second gener-
It is only right that this epidemiological literature is subject to ation migrant backgrounds exhibit elevated neural responses
proper scrutiny to determine whether these patterns are causal. to stress following a sociocultural challenge9. If we can eluci-
If they are not, then the alternatives are no less palatable: that date whether these putative stress pathways also contribute to
other social or economic exposures are so entrenched within the onset of psychosis potentially encompassing complex
certain black and ethnic minority (BME) sections of society that interactions between genetic, biological and social factors
they are powerful enough to increase the chance of experienc- this will not only move us closer to understanding the excess
ing a psychotic disorder by up to 1000%; or that the tools, prac- risks among BME communities, but in society at large. Aside
titioners and institutions tasked with making reliable and valid from psychosis (and, perhaps, post-traumatic stress disorder),
diagnostic assessments are so unfit for purpose, or so grossly there is less consistent evidence that migrants are at higher risk
inept at differentiating between normal cultural mores of be- of other mental health conditions; this specificity would be one
haviour and psychotic symptoms, that for every one migrant of several important criteria helping to establish causation.
correctly diagnosed, a further nine may be misdiagnosed with Further studies are also required in settings where the in-
psychotic disorder. creased psychosis risk amongst migrants is not observed, such
Scrutiny of the evidence in relation to misdiagnosis does as in people of Indian descent in the UK3, Turkish descent in the
not strongly support this as an explanation of higher rates. Netherlands2, or Hispanic origin in the US10. Canada is another
There may be poor inter-rater reliability between psychiatrists putative counterfactual setting, given both its foundation on a

relatively recent migration history, and the effects on mental to advance our aetiological understanding on this issue in or-
health of indigenous First Nations people in this context. der to eradicate this gross social injustice.
Studies in settings where white migrants form the minority
group would also shed further light on the role of migration in James B. Kirkbride
PsyLife Group, Division of Psychiatry, University College London, London, UK
psychosis risk. South Africa provides a possible example. None-
theless, while white migrants in this context would be the minor- The author is supported by a Sir Henry Dale Fellowship, jointly funded by the
ity in terms of population size, they also continue to hold a Wellcome Trust and the Royal Society (grant no. 101272/Z/13/Z). He is grateful
to J. Hayes for his critical proof reading of an earlier draft of the paper.
disproportionate balance of socioeconomic capital, which may
negate any effect; in either case, the aetiological implications 1. Silove D, Ventevogel P, Rees S. World Psychiatry 2017;16:130-9.
would be illuminating. For various reasons, and not without 2. Cantor-Graae E, Selten JP. Am J Psychiatry 2005;162:12-24.
3. Kirkbride JB, Errazuriz A, Croudace TJ et al. PLoS One 2012;7:e31660.
considerable challenges, Brazil, China, Japan and Zimbabwe 4. Hollander A-C, Dal H, Lewis G et al. BMJ 2016;352:i1030.
present other settings for such counterfactual study. 5. Lasalvia A, Bonetto C, Tosato S et al. Br J Psychiatry 2014;205:127-34.
Using data from the UK, we have previously estimated that, 6. Hickling FW, McKenzie K, Mullen R et al. Br J Psychiatry 1999;175:283-5.
7. Kirkbride JB, Hameed Y, Ankireddypalli G et al. Am J Psychiatry 2017;174:
if we could identify the drivers of the elevated psychosis risk in 143-53.
BME groups, we could prevent up to 22% of new cases of first 8. Levine SZ, Levav I, Goldberg Y et al. Psychol Med 2015;46:855-63.
episode psychosis in the general population, and up to two 9. Akdeniz C, Tost H, Streit F et al. JAMA Psychiatry 2014;71:672-80.
10. Oh H, Abe J, Negi N et al. Psychiatry Res 2015;229:784-90.
thirds in BME groups specifically11. This major health inequal- 11. Kirkbride J, Coid JW, Morgan C et al. J Publ Ment Health 2010;9:4-14.
ity may be to psychiatry what nicotine exposure was to bron- 12. Doll R, Hill AB. BMJ 1950;2:739-48.
chogenic carcinomas over 65 years ago12: our smoking lung.
The psychiatric research community has an unparalleled duty DOI:10.1002/wps.20406

SPECIAL ARTICLE
Etiology in psychiatry: embracing the reality of

poly-gene-environmental causation of mental illness
Rudolf Uher, Alyson Zwicker
Departments of Psychiatry and Pathology, Dalhousie University, Halifax, B3H 2E2, Nova Scotia, Canada
Intriguing findings on genetic and environmental causation suggest a need to reframe the etiology of mental disorders. Molecular genetics shows
that thousands of common and rare genetic variants contribute to mental illness. Epidemiological studies have identified dozens of environmen-
tal exposures that are associated with psychopathology. The effect of environment is likely conditional on genetic factors, resulting in gene-
environment interactions. The impact of environmental factors also depends on previous exposures, resulting in environment-environment
interactions. Most known genetic and environmental factors are shared across multiple mental disorders. Schizophrenia, bipolar disorder and
major depressive disorder, in particular, are closely causally linked. Synthesis of findings from twin studies, molecular genetics and epidemio-
logical research suggests that joint consideration of multiple genetic and environmental factors has much greater explanatory power than sep-
arate studies of genetic or environmental causation. Multi-factorial gene-environment interactions are likely to be a generic mechanism
involved in the majority of cases of mental illness, which is only partially tapped by existing gene-environment studies. Future research may
cut across psychiatric disorders and address poly-causation by considering multiple genetic and environmental measures across the life course
with a specific focus on the first two decades of life. Integrative analyses of poly-causation including gene-environment and environment-
environment interactions can realize the potential for discovering causal types and mechanisms that are likely to generate new preventive and
therapeutic tools.
Key words: Psychiatric genetics, environmental risk factors, gene-environment interactions, classification of mental disorders, life course
research, schizophrenia, depression, bipolar disorder, autism
(World Psychiatry 2017;16:121129)
Major depressive disorder, schizophrenia, bipolar disorder consistently show that monozygotic twins who share 100% of
and autism are among the most disabling and costly diseases1. their nuclear DNA are more likely to be concordant on each dis-
They affect individuals from young age, and are associated order than dizygotic twins who share 50% of their genetic mate-
with physical morbidity and early death2. The causal mecha- rial5. This difference suggests that the causation of mental illness
nisms underlying mental illness may hide keys to effective pre- is to a large degree attributable to genetic factors.
vention and treatment, but remain poorly understood. There is a gradient of genetic contribution, with higher esti-
The last two decades have seen an expansion of knowledge mates of heritability for the more severe and less common dis-
punctuated by surprises that challenge previously held assump- orders (autism, schizophrenia, bipolar disorder) and a lesser
tions about mental illness. In this paper we provide a synthesis of degree of heritability for more common and less severe disor-
current knowledge and direct further research to maximize the ders (anxiety, major depressive disorder)5.
potential for meaningful discovery. While the focus is on generic The large heritability estimates promised an easy identifica-
principles underlying the causation of any mental illness, the tion of the molecular genetic variants responsible for the cau-
majority of information comes from studies of schizophrenia, sation of mental illness. Influential authorities estimated that
bipolar disorder, major depressive disorder and autism, on which severe mental illness, such as schizophrenia, was likely to be
most data have been accumulated. caused by several (2 to 9) genetic loci6, while others argued for
We first review the genetic and environmental factors impli- a single gene causing most cases of schizophrenia7.
cated in the etiology of mental illness, before adopting an integra- Three assumptions have shaped the field of genetic discovery:
a) severe mental illness is caused by a small number of genes; b)
tive perspective that jointly considers genetic and environmental
there is a specific relationship between genotype and the type of
elements of causation. We conclude by outlining a framework for
mental illness and c) the genetic variants lead to mental illness
productive causal research.
through biological pathways independent of environment. Con-
sequently, most genetic research has studied one mental disorder
at a time by comparing cases with a specific diagnosis to con-
GENETIC FACTORS IN THE CAUSATION OF MENTAL trols, without accounting for environmental influences.
ILLNESS Over the last decade, the molecular genetic technology has
offered the tools to study the genetic variants responsible for
All types of mental illness have a tendency to run in families, the transmission of liability for mental illness from parents to
and the risk of developing an illness is associated with the degree offspring. This decade of research has brought surprising find-
of biological relatedness to the affected individual3,4. This pattern ings that challenge the assumptions on which psychiatric genet-
of transmission strongly suggests genetic causation. Twin studies ics has been based. Genome-wide association studies have

Table 1 Genetic variants associated with mental illness
Autism Schizophrenia Bipolar disorder Depression
Number of individuals in largest 13,088 cases with autism 36,989 cases with 7,481 cases with 121,380 cases with
genetic sample to date spectrum disorders and schizophrenia and bipolar disorder depression and
16,664 controls 113,075 controls and 9,250 controls 338,101 controls
Number of genetic variants 4 128 18 17

associated at genome-wide
level of statistical significance
Odds ratio of the most strongly 1.17 1.21 1.15 1.05

associated genetic variant
Proportion of variance explained 14% 23% 25% 5%

by common genetic variants
across the genome
identified more than a hundred variants associated with severe even for the most heritable types of mental illness, such as
mental illness (Table 1)8-11. Each of the variants has small ef- autism or schizophrenia5. While it is not possible to complete-
fect and the number of associated variants keeps increasing ly separate the effects of environment from errors in diagnosis,
with growing sample sizes8. a realistic assessment suggests that environmental and genetic
Polygenic risk scores analyses consistently show that the factors contribute equally to the causation of mental illness.
prediction of mental illness improves by including more weak- Since the 1960s, researchers have been identifying strong
ly associated genetic variants, suggesting that many thousands relationships between adverse social environment and mental
of genetic variants are involved in shaping the risk for most illness. The bulk of the research on social causation has been
mental disorders12,13. These involve both common single based on the assumption that a single environmental factor
nucleotide polymorphisms and rare structural variants, such may explain the causation of a specific diagnosis, irrespective of
as deletions and insertions of stretches of DNA14. enduring characteristics of the exposed individual. Thus, social
Another consistent finding is that most common and rare researchers tended to examine one aspect of environment and
genetic variants are non-specifically associated with a range of one mental disorder diagnosis at a time. The highlights of this
mental disorders15,16. Overall, approximately two thirds of research included identification of strong associations between
genetic associations are common to schizophrenia, bipolar severe adverse life events and depression19.
disorder and major depressive disorder15. There are also over- A number of studies of environmental factors have included
laps with genetic variants contributing to autism, attention- longitudinal follow-ups and documented both the long-term
deficit/hyperactivity disorder, and intellectual disabilities. effects of adversity in childhood and the close temporal rela-
It has also become clear that the heritability estimates derived
from twin studies do not translate into direct effects of molecular
genetics variants15. The estimates based on case-control studies
with molecular genomic data suggest that genetic variants con-
tribute only a fraction of the effect that was suggested by herita-
bility estimates from twin studies (Figure 1). The most likely
explanation for this heritability gap is that a large fraction of
genetic effects are contingent on factors that are common to
individuals growing up in the same family but not to unrelated
individuals who participate in case-control studies17,18. A picture
is emerging of a complex etiological mechanism, where genetic
influence is thinly distributed across thousands of genetic var-
iants of small effects that are contingent on environment and
not specific to any single form of psychopathology.
Figure 1 The heritability gap. Heritability (the proportion of causation

attributable to genetic factors) has been estimated from differences of
ENVIRONMENTAL FACTORS IN THE CAUSATION OF concordance between identical and fraternal twins (twin estimates)
MENTAL ILLNESS and from hundreds of thousands of single nucleotide polymorphisms
across the human genome (molecular estimates). The large difference
between the twin and molecular estimates is referred to as the
The same twin studies which confirmed that mental illness
heritability gap. Twin estimates are based on same-sex twin pairs
is heritable have also demonstrated that environment matters. from a recent comprehensive meta-analysis5. Molecular estimates are
Concordance of genetically identical twins is far from perfect from large case-control genome-wide association studies8-11.

Table 2 Environmental factors associated with mental illness Resilience appears to be related to a number of enduring
personal characteristics that are partly heritable and partly
Bipolar
Autism Schizophrenia disorder Depression
shaped by previous environmental exposures28,29. Experiences
early in life may make a person more vulnerable or resilient to
Pregnancy risk factors exposures later in development, resulting in a sequential
Infections 1 111 11 1 environment-environment interaction. For example, exposure
to maltreatment in childhood may cause sensitization to the
Malnutrition 111 11 11
effects of specific types of stressful life events in adulthood30.
Heavy metals 111 11
The observation that unshared environment has greater influ-
Perinatal risk factors ence on intellectual ability among twins growing up in families
Preterm birth 11 11 11 11 with low socio-economic status also suggests a complex inter-
Season of birth 11 111 11 1 play between multiple environmental factors31.
Birth complications 111 111 0 A synthesis of current knowledge on environmental causation
of mental illness suggests a complex picture with a multitude of
Childhood environment
social, physical and chemical exposures occurring at different
Urbanicity 111 111 1 1
stages of life, affecting the risk for a range of mental disorders. It
Poverty 11 111 1 111 is becoming increasingly unlikely that any given environmental
Maltreatment N/A 11 11 111 factor could be a necessary and sufficient cause of any mental
Bullying N/A 11 1 111 disorder. Instead of searching for single disorder-specific envi-
Drug use in adolescence ronmental causes, researchers who want to explain or predict
mental illness may need to jointly study a multitude of environ-
Cannabis N/A 111 11 1
mental influences across the life course, that may be summed
Stimulants N/A 111 11 0
up in cumulative poly-environmental scores (E-scores)32 or
The number of 1 marks the strength of evidence (1 means some evidence of grouped in unique environment-environment constellations31.
association/single report; 11 means moderate replicated evidence of associa- While the array of environmental factors that are known to
tion/multiple reports; 111 means strong evidence of association/multiple be involved in the causation of mental illness is impressive, it
replications or good meta-analysis). Evidence of no association is noted as 0.
may still only be the tip of an iceberg. Research designs to date
Empty cells reflect absence of evidence for or against association. No factor
has been negatively associated with any of the disorders. Because of the early have only been powered to detect environmental factors that
age at onset of autism, environmental factors occurring after age 3 cannot be are harmful for the vast majority of individuals. The types of
reliably studied and are marked as not applicable (N/A). environments that may good for some and bad for others are
still waiting to be discovered.
tionship between severe life events and psychopathology onset

in adulthood20,21. With larger and more representative studies, GENE-ENVIRONMENT CAUSATION OF MENTAL
additional and diverse environmental risk factors have been ILLNESS
identified, including exposure to viral infections during gestation,
vitamin D deficiency, growing up in urban environment, ethnic No genetic variant and no environmental exposure on its own
minority status, childhood maltreatment and bullying victimiza- is a sufficient cause of mental illness. While it is possible that
tion (Table 2)22-25. some cases of mental illness are caused by a combination of
Several general principles emerged. First, the same type of many genetic variants or a combination of multiple environmen-
environmental exposure increases the risk of many different tal exposures, the most likely scenario by far is that both genetic
mental disorders. For example, urban environment was first and environmental factors jointly contribute to the causation of
identified as a risk factor for schizophrenia, but a systematic mental illness. A causal mechanism where one or more genetic
analysis showed that it is associated with increased risk of all factors and one or more environmental factors are required to
types of mental disorders25. Second, many different types of produce an outcome is gene-environment interaction (GxE).
environmental exposures contribute to the same disorder. For A ubiquitous role of GxE in the causation of mental illness
example, the risk of schizophrenia increases with maternal mal- is suggested by a contradiction between the results of epide-
nutrition, vitamin D deficiency and viral infections during preg- miological studies and twin studies that we call the shared
nancy, low socio-economic status, urban upbringing, minority environment paradox. Epidemiological research shows that a
status and childhood maltreatment, as well as exposure to substantial proportion of cases of mental illness are attribut-
stimulants, cannabis and tobacco26. Third, no constellation of able to environmental factors which are typically shared by
adverse environmental exposures will result in psychopathology whole families, such as socio-economic class, poverty, urban
among all exposed individuals. Many individuals appear to be environment, minority status, neighbourhood characteristics
resilient and do not develop any mental disorder even if they and childhood maltreatment22,25,33. Yet, twin studies allocate
are exposed to multiple adverse environmental factors27,28. only a very small role to shared environmental factors5 (Figure

and adversity leading to depression has seen similar number
of replications and non-replications and it may be specific to
childhood maltreatment leading to persistent depressive disor-
der38-40. Yet other reported GxE have proven unreplicable. For
example, the interaction between catechol-O-methyltransferase
(COMT) gene and cannabis use leading to psychotic symp-
toms has been reported, but not replicated in subsequent
studies41.
More recent studies have screened a larger number of genes
and polymorphisms to search for GxE. Such systematic search
has led to the identification, among 152 polymorphisms in
42 genes related to cannabinoid signalling, of an interaction
between a single nucleotide polymorphism in the serine/threo-
nine kinase encoding gene AKT1 and cannabis use leading to
psychosis42. This GxE has been replicated in independent sam-
Figure 2 Shared environment paradox. Twin studies have consistent- ples43,44, suggesting that this polymorphism sensitizes individu-
ly allocated little or no role in the causation of mental illness to envi-
als to the psychosis inducing effects of tetrahydrocannabinol.
ronmental factors that are shared by members of the same family.
The estimates plotted here are from a recent comprehensive meta- Finally, several genome-wide environment interaction stud-
analysis of twin studies5, based on same-sex twin pairs. Estimates for ies (GWEIS) have been completed to search for GxE without
schizophrenia and depression were actually negative, but since a neg- any pre-existing hypothesis about the genetic variants involv-
ative contribution to variance is not possible, we plotted them at 0%. ed. The first GWEIS concerned interactions of common genet-
ic variants with prenatal exposure to cytomegalovirus and with
stressful life events in the causation of schizophrenia45 and
2). One explanation of the shared environment paradox is that depression46, respectively. The existing GWEIS have limited
the impact of the family-wide environment depends on factors statistical power, because most large genotyped samples are
that are shared more between monozygotic than between di- missing adequate measures of environment. At present, it is
zygotic twins, i.e. genetic polymorphisms. If the effects of unclear whether the results of GWEIS will be more replicable
shared environment are conditional on genetic variants, the than those of candidate GxE studies. An interim synthesis sug-
statistical models used in twin studies will fully attribute the gests that multiple GxE contribute to most types of mental ill-
joint effect to the genetic component, thus inflating heritabili- ness, but no specific GxE explains a substantial proportion of
ty and reducing the estimate of shared environment34. In this cases.
way, GxE provide the most parsimonious explanation for both Several studies suggest that multiple genetic variants shape
the shared environment paradox and the heritability gap. the susceptibility to harmful and protective environmental
In the last 15 years, researchers have started to identify spe- factors. One study has shown that a score derived from over
cific genetic variants that may sensitize individuals to environ- 2,800 schizophrenia-associated variants in coding or regulato-
mental factors. Like most molecular genetic research, the ry genomic regions interacted with winter birth to increase the
search for GxE started with tests of candidate polymorphisms risk of schizophrenia47. In another study, a polygenic risk score
in candidate genes. The success of such studies depends not of tens of thousands of common variants tapping the overall
just on picking the correct combination of a genetic variant
sensitivity to environment predicted the effects of negative
and an environmental factor based on prior knowledge, but
parenting on emotional psychopathology as well as the effec-
also on sampling and design that allows an approximation of a
tiveness of intensive psychological treatment for anxiety48.
biological interaction with a statistical test.
As in studies of direct polygene-disorder associations, the
Remarkably, some of these studies appear to have been suc-
GxE increased in strength with more genetic variants being
cessful in finding causal mechanisms. Some candidate gene
included in the polygenic risk score. The emerging pattern of
GxE have been consistently replicated. For example, the inter-
findings suggests that sensitivity to environment is a highly poly-
action between low activity variants of the X-chromosome-
genic trait with contributions from thousands of common genet-
linked monoamine oxidase A (MAO-A) gene and childhood
maltreatment leading to antisocial behavior in males has been ic variants.
replicated multiple times and confirmed in meta-analyses35,36. The examination of gene-environment interplay is still in
The interaction between brain derived neurotrophic factor its infancy, and research available to date leaves many unan-
(BDNF) gene variants and severe life events leading to depres- swered questions. The specificity of polygenic GxE to the type
sion has also been replicated and confirmed in a meta-analysis37. and timing of environmental exposures, the specificity or plei-
Other GxEs have proven less robust or more specific than otropy of GxE across types of mental disorders, and the role of
originally reported. For example, the interaction between short rare structural variants in sensitivity to environment remain
variants of the serotonin transporter gene length polymorphism largely unexplored.

BOUNDARIES AND OVERLAP BETWEEN DISORDERS constructs remain more relevant to the severe types of mental
illness that are most pertinent to psychiatry. When it comes to
The review of genetic and environmental factors above has psychopathology, it is unlikely that the difference between
concluded that most factors are associated with most types of complete absence of pathological symptoms and population
mental illness. The apparent overlaps in causation has been average matters as much as the difference between average
generally ascribed either to pleiotropy, i.e. the same factors and severe psychopathology. Yet, the number of categories in
having the potential to cause multiple types of illness, or to the the current classifications are too large and the boundaries
lack of validity of the diagnostic criteria for specific disorders. between them lack validity52.
Pleiotropy at the level of a single causal factor has been well Because psychiatric research to date has been based on the
documented: for example, the same variant (A-allele of now refuted assumption of diagnostic specificity, most studies
rs1006737) of the calcium channel gene CACNA1C has been are uninformative about the validity of specific diagnostic cat-
associated with increased risk of bipolar disorder, schizophre- egories or dimensions52. The potential for discovery will likely
nia, depression, anxiety and autism49-51. While inadequate be enhanced if researchers refocus on examining broad and
validity of boundaries between diagnostic categories has also heterogeneous samples of mental illness without exclusions
been amply demonstrated52, evidence supporting validity also based on diagnostic criteria and without constraining their
exists, e.g. in the specificity of therapeutic response to lithium measurement to consensus based constructs, categorical or
in bipolar disorder but not schizophrenia. dimensional. Shedding the constraints of diagnosis-specific
While both pleiotropy and inadequate validity of categori- research does not require adopting another set of constraints
zation are likely to be at play, the multifactorial causation also and it does not necessitate transition from a categorical to a
leaves the possibility of unique combinations of causal factors. dimensional framework of inquiry. Examination of overlaps in
For example: even if most risk factors are shared between etiological factors between disorders suggest that higher level
bipolar disorder and schizophrenia, the loading and combina- broad categorical constructs may be more appropriate targets
tion of factors that give rise to each of the two disorders may of etiological research than specific diagnostic categories. For
still be unique. example, the major overlap in both genetic and environmental
Since hundreds or thousands of environmental and genetic contributors between major depressive disorder, bipolar disor-
factors are likely involved in the causation of each disorder, der, schizophrenia and other types of psychotic illness suggests
the number of possible combinations is extremely large. The that a broad category of severe mental illness that encompasses
examination of these possible combinations has only just major mood and psychotic disorders may be an appropriate unit
begun. One example involves both a mental disorder and a of investigation.
physical disorder: individuals with schizophrenia have less
than half the risk of developing rheumatoid arthritis compared
to the general population, even though schizophrenia and DEVELOPMENTAL CONTEXT AND CLINICAL COURSE
rheumatoid arthritis share environmental risk factors, includ-
ing winter birth and tobacco smoking. Recently it has been Two major discoveries in psychiatry remain underrated and
shown that a polygenic risk associated with the immune sys- are not reflected in most etiological research. The first one is
tem is associated with both increased risk of rheumatoid the continuity of pathology over the life course. From cohorts
arthritis and reduced risk of schizophrenia. In addition, a poly- with complete and intensive long-term follow up, it has
genic risk score for schizophrenia interacts with winter season become clear that most cases of mental illness start in child-
of birth to increase the likelihood of schizophrenia47. In this hood and adolescence. The early manifestations of psychopa-
case, some environmental factors are shared, but genetic dis- thology typically differ in kind from the eventual diagnoses in
position distributed across thousands of variants may deter- adulthood, yet are very strongly predictive of mental illness
mine the relative risk of two competing outcomes. diagnosed across the life course.
Even if we are able to examine combinations of genetic and Heterotypic continuity is the rule. For example, anxiety in
environmental factors, the question remains about the level of childhood is a strong predictor of both major depressive disor-
outcomes that is most likely to lead to success in etiological der and bipolar disorder in adulthood. Oppositional defiant
research. Most of the debate to date has focused on the dis- disorder in childhood predicts a broad range of psychopathol-
tinction between categorical diagnoses and dimensional mea- ogy in adults, including depression, bipolar disorder, anxiety,
sures. This may have been a false focus. At present, it is substance use disorders and antisocial personality disorder.
unclear whether one of the approaches is more advantageous Yet, there is also a degree of specificity, with a systematic cor-
than the other. The experience with dimensional constructs respondence between the profile of childhood symptoms and
introduced as part of the Research Domain Criteria framework the type of adult disorders53.
over the past five years does not inspire hopes for major The fact that most individuals with mental illness will go
advances in etiological research. While dimensional measures through a number of diagnostic categories over their life
may be more powerful for examining variation in common course adds to the problems associated with diagnosis-specific
traits across the general population, the categorical diagnostic research52. It highlights the need to examine psychopathology

broadly and in a developmental context. Since most mental ill- in psychiatry occurred before these assumptions were estab-
ness starts in childhood and retrospective report is inaccurate, lished. For example, the discovery of lithium efficacy for bipolar
etiological research needs longitudinal designs that start in disorder occurred thanks to investigation in an unselected
childhood, at birth or even earlier54. group of patients62.
The second discovery is that the course of mental illness Another limiting factor lies in omissions. The diagnosis-
varies between individuals and is only loosely related to the focused approach of the 20th century and the ensuing cate-
diagnostic category. Traditionally, some disorders have been gories-vs.-dimensions discussion might have led to the neglect
conceptualized as episodic and other disorders as persistent, of the developmental context and the temporal dimension of
but longitudinal follow-ups suggest that this conceptual dis- psychopathology.
tinction has limited validity. Mood disorders have been codi- The final limitations we will discuss are related: statistical
fied as episodic, yet on follow-up they are marked by power and quality of measurement. Since many genetic and
chronicity, with most individuals spending most of their time environmental factors contribute to most cases of mental ill-
with depressive symptoms55,56. Personality disorders have ness, large representative samples with accurate measure-
been conceptualized as persistent, but on follow-up their symp- ments of genetic variation, environmental exposures and
toms show similar rates of remission and relapses as mood disor- psychopathology over the life course are needed for etiological
ders do57. research. We have many studies with good measurement of
Yet, within the same disorder, episodic and persistent cases environment, but they do not overlap much with studies with
may have distinct etiologies. For example, episodic cases of high standard of genetic measurement. We have some longitu-
major depressive disorder are more strongly heritable58 and dinal studies with high quality measurements and we have
persistent cases are more strongly linked to childhood mal- some studies with a large number of individuals. Unfortunate-
treatment59. The interaction between serotonin transporter ly, there has been little overlap between the two. The largest
gene length polymorphism and childhood maltreatment also studies in psychiatry are either pulled together from many var-
appears to be specific to persistent depressive disorder39,40. iably assessed samples or they suffer from large dropout rate
On the other hand, there is evidence that cycloid psychosis, a and less accurate measurement. We may not get the answers
type of mental illness marked by a characteristic highly epi- about causation of mental illness unless experts in develop-
sodic course in spite of varied symptom content, may have mental psychopathology, environment and genetics join forces
distinct genetic underpinning60,61. These examples suggest to work together on large longitudinal studies. Early examples
that time course of illness may be at least as important in etio- of such collaborations are emerging and hopefully will be
logical studies as the symptom profile. completed.
The findings of longitudinal research outlined above high-
light a massive caveat in prior etiological studies that grouped
individuals into cases and controls based on symptom content
FRAMEWORK FOR DISCOVERY
in adulthood without reference to developmental context or
time course of symptoms. Future etiological research will be
To substantially advance our understanding of mental ill-
improved by the systematic incorporation of a temporal
ness, the next generation of studies will need to embrace the
dimension that has been conspicuously missing from both
complexity of poly-gene-environmental causation. The tech-
categorical and dimensional classifications used by most etio-
nology and methodological knowledge available today enables
logical studies to date.
studies of multiple environmental and genetic factors without
assumptions of independence. It is essential that the research
studies are designed in a way that maximizes the potential for
LIMITS OF CURRENT APPROACHES meaningful discovery by avoiding the pitfalls of assumptions,
omissions, inadequate measurement and statistical power
The last decade has seen a large amount of criticism of psy- (Figure 3).
chiatric research. It may be important to own up to both suc- Large longitudinal studies of samples that are not selected
cesses and failures and take a stock of what might have for a particular diagnosis are needed to enable new discovery.
hindered the field from knowing more. Based on the review of These studies should start in pregnancy, childhood or adoles-
etiological research in psychiatry outlined above, we conclude cence to capture the development of psychopathology and
that four factors are limiting further progress. allow separation of cases from consequences. Repeated assess-
One of the major limiting factors is assumed knowledge. ments of multiple aspects of environment during the individu-
Over the past five decades, psychiatry researchers have built als development should cover known environmental risk
their studies around the following assumptions: causes are factors as well as key factors of environment that may be good
diagnosis-specific, disorders are caused by a small number of for some and bad for others. Regular assessments of psychopa-
factors, and genetic factors have primacy over environmental thology across the life course are needed to establish true age
influences. It is remarkable that some of the greatest discoveries at onset, track the course and record sequential comorbidity.

Figure 3 Framework for discovery. Life-course developmental perspective and an open search space for unique combinations of genetic and
environmental factors (including gene-gene, gene-environment, and environment-environment interactions) are core elements that will
enhance the potential for discovery in etiological research. Genetic and environmental data reduction polygenic sensitivity scores, polygenic
risk scores, genetic pathway scores, poly-environmental scores (E-scores) may be a necessary intermediate step towards the discovery of
broad poly-gene-environmental causal mechanisms, but the reduction process should be reversible to enable fine mapping of specific molecu-
lar and behavioral mechanisms.
Measurement of environment and psychopathology should should be open to identify combinations of early and late envi-
use multiple independent sources of information to maximize ronmental factors as well as of environmental and genetic
objectivity and avoid common source bias (e.g., predictably factors.
high but uninformative correlations between questionnaires Tools for such analyses are becoming widely available. For
completed by the same individual at the same time). Instead example, statistical learning offers a set of tools that are de-
of case-control studies, genetic measurement should concen- signed to maximize the use of rich datasets in the prediction
trate on samples with high-quality longitudinal data on envi- and explanation of outcomes and at the same time provide
ronment and psychopathology. understanding of how individual factors contribute to the pre-
With broadly based assumption-free designs, the onus will diction63. Methods are also being developed that make it
be on data analysis to make use of the resulting data in a way possible to distinguish between causal heterogeneity and poly-
that can identify complete etiological mechanisms leading to factorial causation64. While the available methods potentially
mental illness. The key challenge for data analysts will be to offer many ways of analyzing rich datasets, the model com-
embrace the complexity of causation while retaining the plexity will have to be kept proportional to available sample
capacity to trace specific causal mechanisms. The data analy- sizes. Given the vast number of potential factors to be consid-
sis may need to move from theory-driven hypothesis testing ered, data analysis process will require a degree of data reduc-
focus to a theory-free explanatory framework that aims to tion in the initial stages. This may take the form of polygenic
explain the causation of a large proportion of cases. It will be risk scores of disorder liability or environmental sensitivity48,
important to identify unique combinations of genetic and genetic pathway scores47 and poly-environmental risk scores32.
environmental variables that lead to mental illness, irrespec- The degree of data reduction should not be excessive and
tive of whether the biological mechanism corresponds to the the process may need to preserve developmental specificity:
constrained concept of statistical interaction. The framework e.g., with separate procedures for childhood, adolescent and

adulthood exposures. Data reduction also needs to be trans- 15. Cross-Disorder Group of the Psychiatric Genomics Consortium. Identifi-
cation of risk loci with shared effects on five major psychiatric disorders: a
parent, so that it is possible to follow a positive result back to genome-wide analysis. Lancet 2013;381:1371-9.
the molecules and specific factors that drive the causal mecha- 16. Lee SH, Ripke S, Neale BM et al. Genetic relationship between five psychi-
nisms. Specific constellations of molecular genetic and envi- atric disorders estimated from genome-wide SNPs. Nat Genet 2013;45:
984-94.
ronmental factors will be needed to inform prevention and 17. Uher R. Gene-environment interactions in common mental disorders: an
treatment65. update and strategy for a genome-wide search. Soc Psychiatry Psychiatr
Eventually, the role of each genetic and environmental vari- Epidemiol 2014;49:3-14.
18. Uher R. Gene-environment interactions in severe mental illness. Front
able has to be understood in a way that enables independent Psychiatry 2014;5:48.
replication and examination of the underlying biological mech- 19. Brown GW, Harris TO. Social origins of depression. A study of psychiatric
anism. Embracing complexity in a transparent and assumption- disorder in women. London: Routledge, 1978.
20. Brown GW, Craig TK, Harris TO. Parental maltreatment and proximal risk
free framework will enable researchers to map complete causal factors using the Childhood Experience of Care & Abuse (CECA) instru-
mechanisms that explain why large groups of individuals devel- ment: a life-course study of adult chronic depression - 5. J Affect Disord
op mental illness. While this may be a bigger task than what pre- 2008;110:222-33.
21. Arseneault L, Cannon M, Fisher HL et al. Childhood trauma and childrens
vious generations of psychiatrists had envisioned, knowledge of emerging psychotic symptoms: a genetically sensitive longitudinal cohort
complete causal mechanisms is necessary to meaningfully trans- study. Am J Psychiatry 2011;168:65-72.
form classification, prevention and treatment. 22. Davis J, Eyre H, Jacka FN et al. A review of vulnerability and risks for
schizophrenia: beyond the two hit hypothesis. Neurosci Biobehav Rev
2016;65:185-94.
23. Arseneault L. The long-term impact of bullying victimization on mental
ACKNOWLEDGEMENTS health. World Psychiatry 2017;16:27-8.
24. Rai D, Lewis G, Lundberg M et al. Parental socioeconomic status and risk
The work reported in this manuscript has been completed thanks to funding of offspring autism spectrum disorders in a Swedish population-based
from the Canada Research Chairs Program, the Canadian Institutes for Health
study. J Am Acad Child Adolesc Psychiatry 2012;51:467-76.
Research (grant nos. 124976, 142738 and 148394), Nova Scotia Health Research
Foundation and the Dalhousie Medical Research Foundation. 25. Vassos E, Agerbo E, Mors O et al. Urban-rural differences in incidence
rates of psychiatric disorders in Denmark. Br J Psychiatry 2015;208:435-40.
26. Davis J, Eyre H, Jacka FN et al. A review of vulnerability and risks for
schizophrenia: Beyond the two hit hypothesis. Neurosci Biobehav Rev
2016;65:185-94.
27. Collishaw S, Pickles A, Messer J et al. Resilience to adult psychopathology
REFERENCES
following childhood maltreatment: evidence from a community sample.
1. Whiteford HA, Degenhardt L, Rehm J et al. Global burden of disease attrib- Child Abuse Negl 2007;31:211-29.
utable to mental and substance use disorders: findings from the Global 28. Cicchetti D. Resilience under conditions of extreme stress: a multilevel
Burden of Disease Study 2010. Lancet 2013;382:1575-86. perspective. World Psychiatry 2010;9:145-54.
2. Walker ER, McGee RE, Druss BG. Mortality in mental disorders and global 29. Rutten BP, Hammels C, Geschwind N et al. Resilience in mental health:
disease burden implications: a systematic review and meta-analysis. JAMA linking psychological and neurobiological perspectives. Acta Psychiatr
Psychiatry 2015;72:334-41. Scand 2013;128:3-20.
3. Gottesman II, Laursen TM, Bertelsen A et al. Severe mental disorders in off- 30. Starr LR, Hammen C, Conway CC et al. Sensitizing effect of early adversity
spring with 2 psychiatrically ill parents. Arch Gen Psychiatry 2010;67:252-7. on depressive reactions to later proximal stress: moderation by polymor-
4. Rasic D, Hajek T, Alda M et al. Risk of mental illness in offspring of parents phisms in serotonin transporter and corticotropin releasing hormone
with schizophrenia, bipolar disorder, and major depressive disorder: a receptor genes in a 20-year longitudinal study. Dev Psychopathol 2014;26:
meta-analysis of family high-risk studies. Schizophr Bull 2014;40:28-38. 1241-54.
5. Polderman TJ, Benyamin B, de Leeuw CA et al. Meta-analysis of the herita- 31. Hanscombe KB, Trzaskowski M, Haworth CM et al. Socioeconomic status
bility of human traits based on fifty years of twin studies. Nat Genet 2015; (SES) and childrens intelligence (IQ): in a UK-representative sample SES
47:702-9. moderates the environmental, not genetic, effect on IQ. PLoS One 2012;7:
6. Gershon ES. Bipolar illness and schizophrenia as oligogenic diseases: e30320.
implications for the future. Biol Psychiatry 2000;47:240-4. 32. Padmanabhan JL, Shah JL, Tandon N et al. The polyenviromic risk score:
7. Crow TJ. A continuum of psychosis, one human gene, and not much else aggregating environmental risk factors predicts conversion to psychosis in
the case for homogeneity. Schizophr Res 1995;17:135-45. familial high-risk subjects. Schizophr Res (in press).
8. Schizophrenia Working Group of the Psychiatric Genomics Consortium. 33. Johnson SB, Riis JL, Noble KG. State of the art review: poverty and the
Biological insights from 108 schizophrenia-associated genetic loci. Nature developing brain. Pediatrics 2016;137:1-10.
2014;511:421-7. 34. Taylor PJ. The unreliability of high human heritability estimates and small
9. Hyde CL, Nagle MW, Tian C et al. Identification of 15 genetic loci associat- shared effects of growing up in the same family. Biol Theory 2008;2:387-97.
ed with risk of major depression in individuals of European descent. Nat 35. Byrd AL, Manuck SB. MAOA, childhood maltreatment, and antisocial
Genet 2016;48:1031-6. behavior: meta-analysis of a gene-environment interaction. Biol Psychia-
10. Psychiatric Genomics Consortium. Large-scale genome-wide association try 2014;75:9-17.
analysis of bipolar disorder identifies a new susceptibility locus near 36. Taylor A, Kim-Cohen J. Meta-analysis of gene-environment interactions in
ODZ4. Nat Genet 2011;43:977-83. developmental psychopathology. Dev Psychopathol 2007;19:1029-37.
11. Robinson EB, St Pourcain B, Anttila V et al. Genetic risk for autism spec- 37. Hosang GM, Shiles C, Tansey KE et al. Interaction between stress and the
trum disorders and neuropsychiatric variation in the general population. BDNF Val66Met polymorphism in depression: a systematic review and
Nat Genet 2016;48:552-5. meta-analysis. BMC Med 2014;12:7.
12. Dudbridge F. Power and predictive accuracy of polygenic risk scores. PLoS 38. Karg K, Burmeister M, Shedden K et al. The serotonin transporter promot-
Genet 2013;9:e1003348. er variant (5-HTTLPR), stress, and depression meta-analysis revisited: evi-
13. Wray NR, Lee SH, Mehta D et al. Research review: Polygenic methods and dence of genetic moderation. Arch Gen Psychiatry 2011;68:444-54.
their application to psychiatric traits. J Child Psychol Psychiatry 2014;55: 39. Brown GW, Ban M, Craig TK et al. Serotonin transporter length polymor-
1068-87. phism, childhood maltreatment, and chronic depression: a specific gene-
14. CNV and Schizophrenia Working Groups of the Psychiatric Genomics environment interaction. Depress Anxiety 2013;30:5-13.
Consortium. Contribution of copy number variants to schizophrenia from 40. Uher R, Caspi A, Houts R et al. Serotonin transporter gene moderates
a genome-wide study of 41,321 subjects. Nat Genet 2017;49:27-35. childhood maltreatments effects on persistent but not single-episode

depression: replications and implications for resolving inconsistent re- 53. Stringaris A, Goodman R. Longitudinal outcome of youth oppositionality:
sults. J Affect Disord 2011;135:56-65. irritable, headstrong, and hurtful behaviors have distinctive predictions.
41. Zammit S, Spurlock G, Williams H et al. Genotype effects of CHRNA7, J Am Acad Child Adolesc Psychiatry 2009;48:404-12.
CNR1 and COMT in schizophrenia: interactions with tobacco and canna- 54. Jaddoe VW, van Duijn CM, Franco OH et al. The Generation R Study:
bis use. Br J Psychiatry 2007;191:402-7. design and cohort update 2012. Eur J Epidemiol 2012;27:739-56.
42. van Winkel R. Family-based analysis of genetic variation underlying 55. Judd LL, Akiskal HS, Maser JD et al. A prospective 12-year study of subsyn-
psychosis-inducing effects of cannabis: sibling analysis and proband fol- dromal and syndromal depressive symptoms in unipolar major depressive
low-up. Arch Gen Psychiatry 2011;68:148-57. disorders. Arch Gen Psychiatry 1998;55:694-700.
43. Di Forti M, Iyegbe C, Sallis H et al. Confirmation that the AKT1 56. Judd LL, Akiskal HS, Schettler PJ et al. The long-term natural history of the
(rs2494732) genotype influences the risk of psychosis in cannabis users. weekly symptomatic status of bipolar I disorder. Arch Gen Psychiatry
Biol Psychiatry 2012;72:811-6. 2002;59:530-7.
44. Morgan CJ, Freeman TP, Powell J et al. AKT1 genotype moderates the acute 57. Zanarini MC, Frankenburg FR, Reich DB et al. Attainment and stability of
psychotomimetic effects of naturalistically smoked cannabis in young sustained symptomatic remission and recovery among patients with bor-
cannabis smokers. Transl Psychiatry 2016;6:e738. derline personality disorder and axis II comparison subjects: a 16-year
45. Borglum AD, Demontis D, Grove J et al. Genome-wide study of association prospective follow-up study. Am J Psychiatry 2012;169:476-83.
and interaction with maternal cytomegalovirus infection suggests new 58. McGuffin P, Katz R, Watkins S et al. A hospital-based twin register of the
schizophrenia loci. Mol Psychiatry 2014;19:325-33. heritability of DSM-IV unipolar depression. Arch Gen Psychiatry 1996;53:
46. Dunn EC, Wiste A, Radmanesh F et al. Genome-wide association study 129-36.
(GWAS) and genome-wide by environment interaction study (GWEIS) of 59. Nanni V, Uher R, Danese A. Childhood maltreatment predicts unfavorable
depressive symptoms in African American and Hispanic/Latina women. course of illness and treatment outcome in depression: a meta-analysis.
Depress Anxiety 2016;33:265-80. Am J Psychiatry 2012;169:141-51.
47. Lee SH, Byrne EM, Hultman CM et al. New data and an old puzzle: the 60. Maj M. Cycloid psychotic disorder: validation of the concept by means of
negative association between schizophrenia and rheumatoid arthritis. Int a follow-up and a family study. Psychopathology 1990;23:196-204.
J Epidemiol 2015;44:1706-21. 61. Pfuhlmann B, Jabs B, Althaus G et al. Cycloid psychoses are not part of a
48. Keers R, Coleman JR, Lester KJ et al. A genome-wide test of the differential bipolar affective spectrum: results of a controlled family study. J Affect
susceptibility hypothesis reveals a genetic predictor of differential re- Disord 2004;83:11-9.
sponse to psychological treatments for child anxiety disorders. Psychother 62. Cade JF. Lithium salts in the treatment of psychotic excitement. Med J
Psychosom 2016;85:146-58. Aust 1949;2:349-52.
49. Green EK, Grozeva D, Jones I et al. The bipolar disorder risk allele at CAC- 63. Iniesta R, Stahl D, McGuffin P. Machine learning, statistical learning and the
NA1C also confers risk of recurrent major depression and of schizophre- future of biological research in psychiatry. Psychol Med 2016;46:2455-65.
nia. Mol Psychiatry 2010;15:1016-22. 64. Han B, Pouget JG, Slowikowski K et al. A method to decipher pleiotropy by
50. Li J, Zhao L, You Y et al. Schizophrenia related variants in CACNA1C also detecting underlying heterogeneity driven by hidden subgroups applied to
confer risk of autism. PLoS One 2015;10:e0133247. autoimmune and neuropsychiatric diseases. Nat Genet 2016;48:803-10.
51. Pasparakis E, Koiliari E, Zouraraki C et al. The effects of the CACNA1C 65. Uher R. The implications of gene-environment interactions in depression:
rs1006737 A/G on affective startle modulation in healthy males. Eur Psy- will cause inform cure? Mol Psychiatry 2008;13:1070-8.
chiatry 2015;30:492-8.
52. Uher R, Rutter M. Basing psychiatric classification on scientific founda-
tion: problems and prospects. Int Rev Psychiatry 2012;24:591-605. DOI:10.1002/wps.20436

SPECIAL ARTICLE
The contemporary refugee crisis: an overview of mental health

challenges
Derrick Silove1, Peter Ventevogel2, Susan Rees1
1
School of Psychiatry, University of New South Wales, and Psychiatry Research and Teaching Unit, Academic Mental Health Centre, Southwestern Sydney Local Health
District, Sydney, Australia; 2Public Health Section, United Nations High Commissioner for Refugees, Geneva, Switzerland
There has been an unprecedented upsurge in the number of refugees worldwide, the majority being located in low-income countries with lim-
ited resources in mental health care. This paper considers contemporary issues in the refugee mental health field, including developments in
research, conceptual models, social and psychological interventions, and policy. Prevalence data yielded by cross-sectional epidemiological
studies do not allow a clear distinction to be made between situational forms of distress and frank mental disorder, a shortcoming that may
be addressed by longitudinal studies. An evolving ecological model of research focuses on the dynamic inter-relationship of past traumatic
experiences, ongoing daily stressors and the background disruptions of core psychosocial systems, the scope extending beyond the individual to
the conjugal couple and the family. Although brief, structured psychotherapies administered by lay counsellors have been shown to be effective
in the short term for a range of traumatic stress responses, questions remain whether these interventions can be sustained in low-resource set-
tings and whether they meet the needs of complex cases. In the ideal circumstance, a comprehensive array of programs should be provided,
including social and psychotherapeutic interventions, generic mental health services, rehabilitation, and special programs for vulnerable
groups. Sustainability of services, ensuring best practice, evidence-based approaches, and promoting equity of access must remain the goals of
future developments, a daunting challenge given that most refugees reside in settings where skills and resources in mental health care are in
shortest supply.
Key words: Refugees, displacement, asylum seekers, ecological models, trauma, stress, mental health, post-traumatic stress disorder, depression,
social interventions, brief psychotherapy
The upsurge in the number of refugees over recent years is THE SCALE OF THE PROBLEM
unprecedented in the modern world. If current trends con-
tinue, one in a hundred persons will be a refugee in the near The United Nations estimate that over 65 million persons
future1. At present, responsibility for mental health support to worldwide are currently displaced by war, armed conflict or
refugees is shared by a network of agencies, including the persecution. In total, 16.5 million fall under the mandate of
United Nations High Commissioner for Refugees (UNHCR) the UNHCR. Although the flow has slowed somewhat, 3.2 mil-
and the World Health Organization (WHO), government and lion persons were displaced in 2016 alone, the leading source
non-for profit organizations, mainstream mental health and countries being Syria and South Sudan1. More than 80% of ref-
specialist refugee services and voluntary organizations. Yet, ugees are displaced internally or have fled across national bor-
the ineluctable reality is that most refugees with mental health der to neighbouring countries, the majority being located in
problems will never receive appropriate services. low- and lower middle-income countries.
The chief reason for this is the scarcity and inequitable Half of the worlds refugees remain in protracted situations,
distribution of services, but other factors contribute to the unstable and insecure locations, most commonly in dense
situation, including difficulties in coordinating national and urban areas, but also in refugee camps. For example, 314,000
international efforts, barriers to accessing care even when persons remain displaced from Darfur in Eastern Chad, and
services are available, and persisting stigma associated with more than a million Somalis live as displaced persons in Kenya,
being both a refugee and mentally ill2. Notwithstanding, ad- Ethiopia, Djibouti and Yemen. Dadaab, a vast refugee camp in
vances have been made in research, theory, policy and models Kenya, houses families that have been sequestrated in this re-
of treatment. Importantly, there is evidence of growing conver- mote and insecure location for more than three generations.
gence in these areas, a consensus that is likely to gradually In 2016, Europe confronted the largest single inflow of refu-
build to the more effective use of scarce resources to achieve gees since the World War II, with over a million Syrians and
better mental health outcomes for this population. others from the Middle East entering the region1. Oscillations
The present paper focuses on issues of general concern in public opinion and government policies resulted at times in
amongst adult refugees. The reader is referred to the special- chaotic responses in which authorities attempted to halt or
ized literature on vulnerable sub-populations (child soldiers, divert the influx, indicating the lack of preparedness of even
unaccompanied minors, children and youth, single or widowed advanced nations to deal with this humanitarian crisis.
women) and specific geographical situations around the To place the European situation in perspective, a total of 13
world3-7. million Syrians have been displaced by the war, the majority to

neighbouring countries. Lebanon, a small country of 4.5 mil- resistance to immigration, the distinction between refugees
lion persons, now accommodates as many Syrian refugees as and voluntary migrants becoming blurred in the process21-24.
the whole of Europe1,8. The wars in the Middle East also tend For all these reasons, although the Refugee Convention is still
to overshadow lesser known refugee crises around the world, in force, there are unprecedented pressures to dilute if not to
for example in West Papua, Myanmar and Western Sahara9-14. dismantle the key provisions for protecting the rights of refu-
gees, irrespective of their backgrounds or countries of origin25.
OSCILLATIONS IN PUBLIC PERCEPTIONS AND

NATIONAL POLICIES EPIDEMIOLOGY OF MENTAL HEALTH PROBLEMS
AMONGST REFUGEES
Throughout history, recipient societies have responded in
ambivalent ways to refugees, at times greeting them as heroes, Prior to the 1970s, the field lacked robust scientific data de-
and at others as interlopers who threaten the peace, integrity, tailing the nature, prevalence and determinants of mental
cultural identity and economic stability of the host country15. health problems amongst refugees. Pioneering studies under-
The policies applied to refugees by host countries are crucial taken in the US, Canada, Norway and Southeast Asia identified
to the mental health of that population. The United Nations what appeared to be substantial symptom levels of anxiety and
Refugee Convention (1951) and later Protocol (1967) ushered in depression amongst Indochinese refugees, but the absence of
a progressive era in the international response to this problem. closely matched comparison groups limited interpretation of
The essential principles established by these instruments in- the findings.
clude: a) that persons with a well-founded fear of being The inclusion of post-traumatic stress disorder (PTSD) in
persecuted for reasons of race, religion, nationality, membership the DSM-III set the stage for the modern era of research in the
of a particular social group or political opinion have an inalien- refugee field, the first studies being conducted amongst South-
able right to seek asylum in signatory countries; b) that refugees east Asian refugees26-28. For example, a study conducted in a
are protected from refoulement or forced return to places of refugee camp for Cambodian survivors of the Khmer Rouge
danger in their homeland; and c) that host countries have a autogenocide found that half of respondents met threshold
responsibility to provide favourable conditions for refugees, criteria for depression and 15% for PTSD27.
including, inter alia, the right to work, to freedom of association In the following two decades, there was a burgeoning of
and movement, and to appropriate services. epidemiological studies in the refugee mental health field,
The Convention proved effective in the early decades follow- prompting two systematic reviews of the cumulative findings
ing the World War II, when refugee flows were small, newcomers in 200530-32. The first, which was limited to studies of refugees
were mainly of European origin, and recipient societies reso- in Western countries, yielded an average prevalence of 9% for
nated positively with their reasons for fleeing, usually based on PTSD and 5% for depression, noting that lower rates were
their opposition to the ideology of totalitarian regimes in the found amongst the larger, more rigorously conducted studies.
countries of origin. The popular campaign against torture in the These findings provided a corrective to the tendency to regard
1970s further strengthened public compassion for survivors all refugees as traumatized and in need of counselling. The
who in most instances were refugees. second review, based on studies that included comparison
The large exodus of Southeast Asian refugees in the 1970s and groups, showed that refugees had a modestly elevated risk
1980s created a new challenge for the Convention16, but after a (effect size of 0.41) of a range of adverse mental health out-
period of inertia and dissension, leading Western nations finally comes. Factors associated with poor mental health amongst
accepted most of the displaced persons for resettlement. Never- refugees included socio-demographic characteristics (being
theless, the crisis underscored a pattern that has been repeated older, a woman, from rural background, well educated, and
in Europe in contemporary times, that is, that the willingness of coming from a higher socio-economic status), and stressors in
recipient countries to accept refugees is inversely related to the the post-displacement environment (living in institutions,
rate of influx and ethnic difference of the incoming group17. restrictions in economic opportunities, being internally dis-
The distinction made in the 1980s onwards between asylum placed or involuntarily repatriated, and coming from a country
seekers (persons arriving without prior authorization) and that remained in conflict).
Convention refugees (those granted residency visas prior to The largest review of its kind, published in 2009, identified
arrival17) further put to test the viability of existing international 181 surveys undertaken amongst 81,866 refugees and other
procedures. Australia implemented stringent policies of deter- conflict-affected populations from 40 countries29. The prevalen-
rence to asylum seekers, and other countries of Europe and ces of PTSD and depression were similar, approximating 30%,
North America instituted similar policies and practices18-20 although there was substantial heterogeneity in rates across
The spirit of the Convention was further eroded by the phe- studies. Exposure to torture and the total number of trauma
nomenon of terrorism. Several factors, including the ethnic events experienced emerged as the strongest predictors of PTSD
and religious stereotyping of terrorists, increased communal and depression, respectively. Larger, more rigorously designed

studies yielded lower prevalence rates, reducing the estimate for frank mental disorder that risks becoming chronic and dis-
PTSD to 15%, a finding broadly supported by a more recent abling, in part independent of the context54. Longitudinal
review33. Even so, the PTSD prevalence greatly exceeds the esti- studies assist to some extent in addressing this problem, in
mate of 1.1% recorded across non-refugee populations in coun- that they are capable of distinguishing between symptom tra-
tries participating in the WHO World Mental Health Surveys34. jectories that indicate recovery as opposed to chronicity, path-
The body of research focusing on asylum seekers served to ways that may be predicted to some extent by the profile of
highlight the impact of the post-migration environment on baseline risk and protective factors55. Short-term follow-up
the mental health of displaced populations35-43. A growing studies (1-3 years) may not distinguish these trajectories with
number of studies in recipient countries found that imposed any accuracy, particularly if the follow-up extends through a
conditions of adversity, including prolonged detention, inse- period of ongoing instability, for example, in the immediate
cure residency status, challenging refugee determination pro- post-displacement phase56-59.
cedures, restricted access to services, and lack of opportunities Only a small number of studies have followed up refugees
to work or study, combined in a way that compounded the for 10 years or longer, in all instances being limited to the mea-
effects of past traumas in exacerbating symptoms of PTSD and surement of general symptoms of anxiety and depression
depression29,36,39,44-48. Yet, in spite of widespread concerns, using screening instruments57-60. Broadly interpreted, these
these practices continue. As a corollary, mental health profes- studies suggest a common pattern of outcome: most refugees
sionals keep on confronting ethical challenges when working continue to show low or no symptoms; a significant minority
within detention centre hierarchies, and practical questions show a pattern of gradual recovery; and a small group remain
persist regarding the effectiveness of offering counselling to chronic. This picture was supported by a large cross-sectional
persons forced to live under such restrictive conditions49. study using a retrospective quasi-longitudinal analysis37. A
similar set of trajectories has been found in a six-year follow-
up study amongst a post-conflict population in Timor-Leste61.
This tripartite pattern of low or no symptoms, gradual recovery
TRANSLATING EPIDEMIOLOGICAL DATA INTO
and chronicity, although tentative, has important implications
POLICY AND PRACTICE
from a public health perspective in judging which populations
will benefit from programs of social reconstruction and which
Translating epidemiological data into estimates of service
might require more intensive psychotherapeutic interventions,
needs requires careful consideration. As indicated, prevalence
as discussed hereunder.
rates of common mental disorders such as depression and
Estimating service needs also depends on a range of other
PTSD have shown wide variation across the body of refugee
factors, including help-seeking behaviour. Stigma, mistrust
studies reported. Methodological factors are partly responsi-
and lack of knowledge of services may limit the extent to
ble, including transcultural measurement error, biases related
which refugees access mental health services, even if available.
to non-probabilistic sampling, and the use of screening mea-
Taking all factors into account, modelling based on the Global
sures which tend to overestimate the prevalence of disor-
Burden of Disease Study has illustrated how large the gap is
der50,51. In addition, populations from some regions of the
between the existing number of mental health professionals
world (East Asia, Sub-Saharan Africa and the Pacific) tend to
and the service needs of low-income countries and regions
record lower symptom levels compared to high-income coun-
that have large populations exposed to mass conflict and dis-
tries52. Failure to include indigenously derived measures that
placement62. There is no realistic prospect, therefore, of formal
capture local expressions or idioms of distress also can lead to
mental health services, whether generic or specialized, meet-
the under-enumeration of mental health problems37,53.
ing the mental health needs of refugees, noting that the major-
Notwithstanding these sources of heterogeneity, substantive
ity reside in low-income countries. Creative solutions are thus
issues of a universal nature, such as the extent of exposure to tor-
necessary, including networking of all agencies to ensure the
ture, the severity and number of trauma events experienced, the
sharing of responsibility of care for refugees with mental disor-
socio-demographic characteristics of the population, the level of
der, and task-shifting, i.e., the transfer of skills to primary care
ecosocial threat that the community continues to face, and the
and lay workers in order to undertake specific mental health
nature and extent of post-migration stressors, all make a major
interventions of various types under supervision.
contribution to the prevalence of disorders across populations.
Given the variation in these substantive factors across contexts,
it should not be surprising that prevalence rates of common
symptoms of mental distress differ from one population to BROADENING KNOWLEDGE OF MENTAL HEALTH
another. OUTCOMES
The greatest obstacle in translating epidemiological data
into service needs arises from the difficulty in differentiating, Recent research in the refugee field has widened the scope
in cross-sectional surveys, between reactions which may be of interest to disorders and reactions that extend beyond the
commensurate with the level of stress being encountered and conventional focus on PTSD and depression, and to a lesser

extent anxiety and somatic symptoms. There is a resurgence environments in which most refugees live72. The chief ongoing
of interest in the construct of prolonged or complicated grief, division in the mainstream was between advocates of individ-
given the importance of this reaction to refugees, the majority ualized, trauma-focused psychotherapeutic approaches and
of whom have experienced multiple losses and separations in those arguing in favour of psychosocial models that focus on
the context of gross human rights violations63. In addition, the the community as a whole and that aim to promote self-
long-debated category of complex PTSD, comprising elements directed recovery and build resilience.
of disrupted self-organization (negative self-concept, affective Contemporary models address these issues by providing a
dysregulation, interpersonal difficulties) will be included for comprehensive account of the refugee experience. Most adopt
the first time in the forthcoming ICD-1164, early evidence sug- a multisystem, ecosocial framework, drawing on established
gesting that the diagnosis can be identified amongst refugees. models in the social sciences73. Within these broad frame-
There is also a growing body of studies documenting cases works, mental disorder is regarded as the endpoint of an im-
in which PTSD is associated with psychotic-like symptoms or balance in the multiplicity of countervailing environmental
frank psychosis amongst refugees and post-conflict popula- factors that impact on refugees rather than an expression of
tions65-67. Recognition of the prevalence and salience of these innate or intrapsychic problems at an individual level. In that
symptom constellations adds further complexity to the field, sense, the distinction between normative and pathological
particularly in relation to the need to tailor interventions to responses is somewhat blurred and fluid, the vicissitudes of
individual patterns of comorbidity and disability. the ecological context determining the direction and extent to
which individuals shift on a continuum of stress.
An example of prevailing models includes Hobfolls Conser-
vation of Resources theory74, which gives centrality to the effects
TOWARDS AN ECOLOGICAL EPIDEMIOLOGY
of objective losses, and the shared meanings of these depriva-
tions within each culture and context in determining mental
The massive disruptions to family and social networks in the
health outcomes and resilience. From that perspective, resil-
context of extreme human rights violations undermines the
ience is regarded both as the capacity of the individual to with-
fundamental sense of coherence of refugees, many becoming
stand experiences of trauma and stress and as the capacity to
isolated and losing trust in authority structures. Chronic anger
remain vigorously engaged with lifes tasks, principally, the pur-
is one potential outcome that has important social implications.
suit of restoring resources that have been lost in times of adver-
For example, amongst West Papuan refugees, a constellation of
sity. The guiding assumption is that all humans have a natural
mistrust, resentment and anger is embodied in an idiom of
drive to obtain, retain, foster and protect resources, defined
distress, Sakit Hati, literally meaning sick heart68.
widely to include a range of domains including the personal
A focus on states of chronic and uncontrollable anger in
(health, well-being, positive sense of self), familial, and social
survivors of extreme trauma creates an important bridge that
(preservation of peace, capacity to work, access to facilities and
links individual reactions to the stability of the family and the
services). Maintaining adequate resources is essential to fulfill-
wider social network. A cycle of violence model posits that, in
ing the task of self-regulation and a sense of control. The refugee
some instances, aggressive outbursts amongst survivors may
situation typifies conditions in which there is a sudden and
be implicated in family conflict, generating a multiplier effect
often massive loss of resources, the pattern of deprivation po-
of mental health problems in intimate partners and potentially
tentially compounding over time. Interventions should focus on
children, a cycle of violence that may have profound transge-
providing the supportive environments that allow refugees (and
nerational effects69. Recent applications of multilevel statisti-
other trauma survivors) to restore their resource base (personal,
cal techniques allow examination of these transactional effects
familial, social, material), a prerequisite for addressing mental
both within conjugal couples and families, thereby broadening
health problems. The model offers the potential to make an
the scope of epidemiology to increase its ecological and con-
objective assessment of the resource losses experienced by indi-
textual significance70,71.
viduals and the community, the totality of the losses indicating
the likely degree of mental distress that will be identified in the
populations. Social interventions aimed at creating a supportive
CONCEPTUAL FRAMEWORKS environment which facilitates the capacity of refugees to restore
their lost resources will advance the overall aim of promoting
From a theoretical perspective, the formative period of the resilience and mental health.
refugee mental health field (broadly the 1970s to 2000) was In their ecological model, Miller et al75,76 give emphasis to
marked by spirited and at times divisive debates in relation to the impact of daily stresses on the mental health of refugees
theory and models of intervention. Those adopting a critical, and asylum seekers. The authors draw on data indicating that
transcultural perspective questioned, and in the most extreme daily stressors partly or wholly mediate the effects of past war-
case rejected, the tendency by Western mental health profes- related trauma in shaping mental health outcomes such as
sionals to transfer Western diagnostic categories such as PTSD PTSD symptoms77. Examples of these stresses include living in
and associated trauma-focused therapies to the culturally distinct unsafe environments, challenges in meeting basic survival needs

(inadequate access to food, water, shelter, health care); inability mass displacement. During these periods, psychiatric hospitals
to pursue income-generating activities; and isolation from family and clinics often close, leaving patients without protection or
and traditional social supports. Vulnerable groups such as medication.
women exposed to gender-based violence, former child soldiers, The reality for psychiatrists and other mental health profes-
unaccompanied and orphaned minors, and persons with physi- sionals working in clinics in Africa and other refugee situations
cal and mental disabilities all face exceptional levels of ongoing is that a large proportion of the patients they consult manifest
stress. Based on this conceptualization, the emphasis of interven- one or more of these forms of severe mental disorder. There is
tions is on creating supportive social environments that reduce now compelling evidence that schizophrenia and other psy-
daily stressors rather than on providing individual psychotherapy chotic disorders are more prevalent amongst refugees resettled
focusing on past trauma experiences. in high-income countries compared to other immigrants and
The Adaption and Development After Persecution and Trau- host populations81. Therefore, the field of refugee mental health
ma (ADAPT) model78,79 identifies five core psychosocial pillars should include consideration of this subpopulation in mounting
disrupted by conflict and displacement, that is, systems of safety comprehensive programs of mental health care, an issue that is
and security, interpersonal bonds and networks, justice, roles now more widely recognized and acknowledged in policy and
and identities, and existential meaning and coherence. These planning exercises82.
pillars form the bedrock on which stable societies are grounded
and on which civilians depend for their mental equilibrium. The
refugee experience, which involves a sequence of adversity that
INTERVENTIONS
traverses epochs of conflict, dislocation, flight, transition and
resettlement, erodes the integrity of all five psychosocial sys-
Brief psychotherapies
tems, thereby weakening social structures and institutions and
exerting deleterious effects on the mental health of individuals.
Counselling and psychotherapy remain the mainstay of
Although the relationship is indirect, the erosion of each pillar
treatment for common mental disorders such as PTSD,
can have broad representations in the symptom patterns identi-
depression and anxiety or combinations of these symptom
fied in refugees. For example, the combination of traumatic loss
profiles in refugees. Most commonly, workers apply a flexible
and extreme injustice may result, via several intermediate path-
combination of supportive counselling and cognitive behav-
ways, in comorbid symptoms of complicated grief and explosive
ioural therapies. In spite of variability in the quality of existing
anger. The ADAPT framework has been used as a conceptual
studies, the overall evidence suggests that various forms of
foundation for formulating and implementing a comprehensive
psychotherapy are relatively effective in ameliorating symp-
refugee mental health program amongst Iraqi refugees in Syr-
toms of PTSD, depression and anxiety83.
ia79. In support of the model, a recent study showed that a mea-
Over the past two decades, a series of brief, structured, man-
sure of the ADAPT construct moderated the effects of past
ualized psychotherapeutic packages have been devised for use
trauma and ongoing adversity in shaping PTSD symptoms80.
amongst refugee and post-conflict populations. Most models
The ADAPT model alerts clinicians and planners to the impor-
draw on evidence from Western contexts supporting trauma-
tance of understanding the overall social ecology of the refugee
focused cognitive behavioural therapies84. The strengths of
experience and contextualizes the array of interventions which
these newer programs include that: a) they can be adapted to
may assist in repairing each pillar, thereby creating the context
local cultures; b) they allow rapid training of front-line person-
for promoting mental health recovery.
nel; and c) they facilitate task-shifting, that is, the transfer of
skills from professionals such as psychologists to lay or commu-
nity workers, a vital provision to allow uptake and dissemination
THE GLOBAL MENTAL HEALTH PERSPECTIVE in settings where there is a severe lack of mental health special-
ists. The time-limited nature and low cost of these interventions
The refugee mental health field overlaps considerably with increase the potential for dissemination (or scalability) and for
the larger movement of Global Mental Health, both focusing integrating the procedures within routine public health or com-
on the mental health needs of deprived populations from low- munity centre settings.
income countries (noting that one of several distinctions is the Most approaches use standard cognitive behavioural com-
substantial number of refugees relocated to high-income coun- ponents including stress management, prolonged exposure,
tries, where they confront special conditions). cognitive restructuring, behavioural strategies, and mindfulness
There has been a tendency in the refugee field to limit interest or related de-arousal techniques. Increasingly, activation thera-
in severe mental illnesses such as schizophrenia and related psy- pies are used for depression. The most widely tested method,
choses, bipolar disorder, melancholic forms of depression, drug narrative exposure therapy, draws on the principles of testi-
and alcohol problems, and organic brain disorders. Persons with mony therapy in tracing the persons chronological life course,
psychosis in particular are at risk of neglect, exploitation and embedding imaginal exposure to trauma memories in the natu-
abuse in acute humanitarian settings and other situations of ral course of this sequence85. A common elements treatment

approach is designed to accommodate common patterns of in routine psychiatric practice, although adjusting dosage ac-
comorbidity, allowing the therapist flexibility in selecting mod- cording to ethno-pharmacological considerations.
ules (for example, for traumatic stress, depression, anxiety) to In general, for common patterns of major depressive disor-
match the particular symptom constellation of each patient. der, PTSD and anxiety disorders, the most commonly used med-
Trials in several settings attest to the efficacy of this method86. ications are the first generation (tricyclic) drugs and, where
More recently, the WHO has established a brief intervention, available, the newer antidepressants (selective serotonin reup-
Problem Management Plus (PM1), drawing once again on the take inhibitors, serotonin and noradrenaline reuptake inhibitors,
core principles and strategies of cognitive behavioural therapy. and their variants), the latter recommended for PTSD by the
The first studies examining this method have yielded positive WHOs Mental Health Gap Action Programme (mhGAP) guide-
findings87. lines89. In many low-income countries, first generation antipsy-
An important next step is to establish that these brief pack- chotic medications (haloperidol, chlorpromazine) are the only
aged interventions can be embedded in routine primary care ones available for psychoses, although atypical antipsychotics,
services in low-income countries in a manner that is sup- including clozapine, are becoming more widely available.
ported by local structures and hence sustainable. Apart from Difficulties are frequently encountered in humanitarian
securing resources and the commitment of the hierarchy to and acute refugee settings in ensuring continuity in the supply
these mental health initiatives, there is a major challenge in of medications. A further challenge is the provision of ongoing
providing ongoing supervision and mentoring of workers, an supervision and in-service training of nurses and other front-
essential provision to avert attrition of skills and motivation line community health workers who commonly oversee the
and to avoid burnout. The increasingly wide reach of the Inter- use of psychotropic medications in low-income countries.
net and telecommunication systems improves opportunities There is a risk, therefore, that practices will be constrained to
to provide supervision from afar to remote locations where standard dosing and that side effects may receive inadequate
many refugee populations are located. attention.
A further concern is whether brief or even extended inter-
ventions based on contemporary approaches to psychother-
Psychosocial interventions
apy are effective for the significant minority of refugees with
complex traumatic stress presentations. A controlled trial from
As indicated, research findings are consistent with contem-
Denmark88 offering a comprehensive array of interventions
porary ecological models in demonstrating the powerful im-
(medical and psychiatric assessment and consultation, psy-
pact that ongoing social conditions exert on the mental health
chopharmacology, social worker assistance, and individualized
and psychosocial well-being of refugees. In addition to the
psychotherapy) found no change in baseline high levels of
effects of past trauma, refugees commonly confront important
PTSD symptoms over a one-year course of follow-up, and only
challenges and stressors in their new environments, including
modest reductions in symptoms of depression. The most likely
ongoing insecurity, restricted access to essential services (health,
reason is that the majority of participants came from the poor
mental health, education), lack of opportunities for employ-
prognostic subpopulation provisionally identified in epidemi-
ment, and more generally, host society attitudes of racism and
ological studies. Participants had extensive exposure to torture
xenophobia. Death, disappearances and separations result in
and other forms of abuse; high rates of head injury, chronic
persisting grief and loss. The ongoing consequence of these
pain and physical disability; a chronic pattern of persisting
losses is that refugees commonly lack the support of nuclear
symptoms; and a history of failed response to past treatments.
and extended families and other traditional networks, a pro-
Most were socially isolated, marginalized and unemployed.
found challenge for communities with strong collectivist val-
Patients with these complex characteristics may not have
ues. Even in intact families, relationships can be undermined
the motivation, resilience or cognitive capacity to engage in
by the cumulative effects of past trauma and ongoing stres-
exposure therapies or to implement the techniques of cogni-
sors, resulting in conflict and, at worst, intimate partner
tive behavioural therapy which require active practice to be
violence90.
effective. Questions remain, therefore, as to the best strategies
Social programs for refugees have the potential to revive a
to assist these complex cases. It may be that more graduated
sense of connectedness, re-establish social networks, and pro-
rehabilitation approaches are needed to encourage what may
mote self-help activities. Strategies that foster community ini-
be a slow recovery trajectory in this subpopulation.
tiatives encourage a sense of control and engagement in the
task of self-directed recovery, counteracting the inertia, depen-
Pharmacotherapies dency, and inter-group divisions that characterize many tran-
sitional refugee settings. There are compelling theoretical,
There is a dearth of research focusing on specific psycho- economic, and strategic reasons, therefore, to give priority to
pharmacological issues amongst refugee populations. Practi- social interventions in the array of strategies aimed at relieving
tioners apply the same range of psychotropic medications used distress and promoting well-being amongst refugees.

At the most general level, psychosocial programs focus on POLICY, LEADERSHIP AND COORDINATION
the population as a whole, examples being community-wide
truth and reconciliation programs, income generation activi- The pioneering phase of the refugee mental health field was
ties, and the development of participatory processes to foster driven by a high level of passion and commitment, in a context
democratic decision-making and self-governance. Practical where program leaders and clinicians were working from a
programs include setting aside child friendly spaces, develop- low knowledge base. The past two decades have witnessed a
ing teams of refugee outreach volunteers to assist families maturing of the field, an era when lead agencies (the United
confronting a range of economic or social problems, and Nations, international non-governmental organizations, univer-
establishing community centres where individuals can obtain sities, amongst others) have established close working relation-
assistance in relation to housing, other basic needs, education, ships that have allowed the gradual building of an international
and referral to other services91-93. consensus on issues that previously were divisive.
Special populations or vulnerable groups such as former The fruits of these endeavours include the formulation and
child soldiers and survivors of gender-based violence may wide adoption of influential policies and guidelines that assist
require specifically designed programs. In some instances, the planning and implementation of programs, for example, the
however, social programs may have paradoxical effects. For Inter-Agency Standing Committee (IASC) Guidelines for Mental
example, participation in truth and reconciliation processes Health and Psychosocial Support in Emergency Settings and the
can improve community cohesion, but result in worsening of SPHERE handbook100,101. A further major achievement has been
mental health. These findings reinforce the need for rigorous the clinical guidelines produced by the WHOs mhGAP, especially
research to test both the benefits and disadvantages of various the module focusing on emergencies102,103. In addition, United
psychosocial programs. Nations agencies have produced and disseminated a range of
Sociotherapy is one of the few well researched group psy- assessment and monitoring tools to encourage standardization
chosocial interventions94, the primary focus being the fostering of assessments across programs around the globe104. There
of connections between people. The method was developed in also have been important consensus building activities in rela-
the post-genocidal context of Rwanda and has since been ap- tion to setting priorities for research105.
plied in other settings including amongst refugees95. Groups
share and discuss daily problems ranging from interpersonal dis-
putes, feelings of marginalization, and strategies to deal with TOWARDS THE FUTURE
gender-based violence and poverty at the community level.
Trained facilitators create a safe therapeutic environment which As indicated, there are growing points of convergence a-
nurtures trust, mutual care and community-wide respect. The cross activities (research, development of conceptual frame-
restorative experience of participating in the group itself may works and policies) in the refugee field, although tensions
assist in repairing disrupted social relationships, although in all remain in some areas. For example, there is clearly a dysjunc-
groups of this kind there should be agreed limits to disclosure, ture between the breadth and complexity of extant ecological
for example, discussing and revealing specific instances of inti- models of mental health and the more limited assumptions
mate partner violence is contraindicated in the group setting. In underpinning the implementation of brief, symptom-focused
general, however, the process may foster supportive peer rela- packages of intervention that continue to be trialled in a range
tionships that endure beyond the life of the group program. Pre- of refugee settings.
liminary research suggests that sociotherapy has the dual effect An important direction for research is to distinguish the
of increasing civic participation (and hence social capital) and needs of the various subpopulations of interest: those with dis-
improving participants mental health96,97. tress reactions that are responsive to environmental factors,
Related models have been trialled, including use of multi- for whom broader social programs as well as more targeted
family interventions in which several families share experien- non-clinical group interventions may be of assistance; those
ces of traumatic stress and chronic adversity. The aim is to whose traumatic stress reactions are severe, disabling and
reduce isolation, create a sense of shared experiences and soli- unlikely to resolve spontaneously and who may benefit from
darity, and foster supportive connections. Preliminary findings brief structured psychotherapies; more complex trauma-
indicate that such methods are effective in improving self- related cases who may benefit from longer-term rehabilitation;
confidence, decreasing social isolation and increasing access the severely mentally ill who need an array of mainstream
to mental health services98,99. interventions; persons with drug and alcohol problems requir-
In relation to future developments, a stepped care model in ing specific attention; and special groups such as women
which refugees first attend social programs which address exposed to domestic violence who may require a gender-
general levels of distress, while at the same time those with sensitive approach to care.
more severe mental health problems are identified, offers an In relation to advocacy, awareness-raising and embedding
integrated approach to maximizing resources and a non- mental health programs within the existing institutional struc-
stigmatizing referral pathway to specialist services. ture, the refugee field can learn a great deal from the general

field of Global Mental Health106,107. Without establishing a firm 18. Morales K. Australias Guantanamo Bay: how Australian migration laws
violate the United Nations Convention against torture. Am Univ Int Law
foothold for refugee mental health in existing primary care and Rev 2016;31:327.
other public health services, issues of sustainability will persist. 19. Fleay C, Cokley J, Dodd A et al. Missing the boat: Australia and asylum
Showing that treatments work under controlled research condi- seeker deterrence messaging. Int Migr 2016;54:60-73.
20. Canetti D, Snider KL, Pedersen A et al. Threatened or threatening? How
tions is only the first step in ensuring that effective interven- ideology shapes asylum seekers immigration policy attitudes in Israel
tions actually reach the majority of populations in need. and Australia. J Refug Stud 2016:29:583-606.
A major challenge that the field confronts at a global level is 21. Silove DM, Rees S, Steel Z. Descent into the dark ages: torture and its per-
ceived legitimacy in contemporary times. In: Dudley M, Silove D, Gale F
that most refugee populations reside in locations where the (eds). Mental health and human rights: vision, praxis, and courage.
resource base in mental health is extremely low. Theoretical Oxford: Oxford University Press, 2012:255-63.
debates aside, the reality is that, in these contexts, no single 22. Silove DM, Rees SJ. Interrogating the role of mental health professionals
in assessing torture. BMJ 2010;340:c124.
agency or program can provide for all the inter-related psycho- 23. Esses VM, Hamilton LK, Gaucher D. The global refugee crisis: empirical
social and mental health needs of refugees. The success of the evidence and policy implications for improving public attitudes and
overall program will be gauged not by the accomplishments of facilitating refugee resettlement. Soc Issues Policy Rev 2017;11:78-123.
24. Kotisova J. Cynicism ex machina: the emotionality of reporting the refu-
one component but by the extent to which all contributors coor- gee crisis and Paris terrorist attacks in Czech television. Eur J Commun
dinate to establish the most comprehensive, inclusive, and inte- (in press).
grated response, which includes networking of mental health 25. Garcia-Zamor J-C. The global wave of refugees and migrants: complex
challenges for European policy makers. Publ Org Rev (in press).
agencies with social, community, and general health services. 26. Mollica RF, McInnes K, Poole C et al. Dose-effect relationships of trauma
Within the mix, the voice of the refugee communities is to symptoms of depression and post-traumatic stress disorder among
vital. Mental health cannot be conferred, it must be regained Cambodian survivors of mass violence. Br J Psychiatry 1998;173:482-8.
27. Mollica RF, Donelan K, Tor S et al. The effect of trauma and confinement
by the communities that have temporarily lost their equilib- on functional health and mental health status of Cambodians living in
rium as a consequence of overwhelming circumstances. Thailand-Cambodia border camps. JAMA 1993;270:581-6.
28. Beiser M. The health of immigrants and refugees in Canada. Can J Publ
Health 2005;96(Suppl. 2):S30-44.
29. Steel Z, Chey T, Silove D et al. Association of torture and other potentially
REFERENCES traumatic events with mental health outcomes among populations
1. United Nations Refugee Agency. Global trends report: world at war. exposed to mass conflict and displacement. JAMA 2009;302:537-49.
Geneva: United Nations High Commissioner for Refugees, 2016. 30. de Jong JT, Komproe IH, van Ommeren M et al. Lifetime events and post-
2. Silove D. The best immediate therapy for acute stress is social. Bull World traumatic stress disorder in 4 postconflict settings. JAMA 2001;286:555-62.
Health Org 2005;83:75-6. 31. Porter M, Haslam N. Predisplacement and postdisplacement factors
3. Sharma M, Fine S, Brennan R et al. Coping and mental health outcomes associated with mental health of refugees and internally displaced per-
among Sierra Leonean war-affected youth: results from a longitudinal sons: a meta-analysis. JAMA 2005;294:602-12.
study. Dev Psychopathol 2017;29:11-23. 32. Fazel M, Wheeler J, Danesh J. Prevalence of serious mental disorder in
4. Tol WA, Stavrou V, Greene MC et al. Mental health and psychosocial 7000 refugees resettled in western countries: a systematic review. Lancet
support interventions for survivors of sexual and gender-based violence 2005;365:1309-14.
during armed conflict: a systematic review. World Psychiatry 2013;12: 33. Priebe S, Giacco D, El-Nagib R. Public health aspects of mental health
179-80. among migrants and refugees: a review of the evidence on mental health
5. Betancourt TS, Williams T. Building an evidence base on mental health care for refugees, asylum seekers and irregular migrants in the WHO
interventions for children affected by armed conflict. Intervention 2008; European region. Copenhagen: WHO Regional Office for Europe, 2016.
6:39-56. 34. Karam EG, Friedman MJ, Hill ED et al. Cumulative traumas and risk
6. Slone M, Mann S. Effects of war, terrorism and armed conflict on young thresholds: 12-month PTSD in the World Mental Health (WMH) surveys.
children: a systematic review. Child Psychiatry Hum Dev 2016;47:950-65. Depress Anxiety 2014;31:130-42.
7. Vostanis P. New approaches to interventions for refugee children. World 35. Steel Z, Silove D, Bird K et al. Pathways from war trauma to posttraumatic
Psychiatry 2016;15:75-7. stress symptoms among Tamil asylum seekers, refugees, and immigrants.
8. United High Commissioner for Refugees. Forced displacement 2015. J Trauma Stress 1999;12:421-35.
Geneva: United High Commissioner for Refugees, 2016. 36. Silove D, Steel Z, Watters C. Policies of deterrence and the mental health
9. Rees S, Silove DM, Tay K et al. Human rights trauma and the mental of asylum seekers. JAMA 2000;284:604-11.
health of West Papuan refugees resettled in Australia. Med J Aust 2013; 37. Steel Z, Silove D, Phan T et al. Long-term effect of psychological trauma
199:280-3. on the mental health of Vietnamese refugees resettled in Australia: a
10. Tay AK, Rees S, Chan J et al. Examining the broader psychosocial effects population-based study. Lancet 2002;360:1056-62.
of mass conflict on PTSD symptoms and functional impairment amongst 38. Momartin S, Silove D, Manicavasagar V et al. Dimensions of trauma asso-
West Papuan refugees resettled in Papua New Guinea (PNG). Soc Sci Med ciated with posttraumatic stress disorder (PTSD) caseness, severity and
2015;132:70-8. functional impairment: a study of Bosnian refugees resettled in Australia.
11. Chantavanich S, Kamonpetch A. Refugee and return: protracted conflict Soc Sci Med 2003;57:775-81.
and displacement in Myanmar. London: Springer, 2017. 39. Steel Z, Silove D, Brooks R et al. Impact of immigration detention and
12. Riley A, Varner A, Ventevogel P et al. Daily stressors, trauma exposure temporary protection on the mental health of refugees. Br J Psychiatry
and mental health among stateless Rohingya refugees in Bangladesh. 2006;188:58-64.
Transcult Psychiatry (in press). 40. Schweitzer R, Melville F, Steel Z et al. Trauma, post-migration living diffi-
13. Martin C. Designing homes to welcome refugees. Lancet 2016;388:1150. culties, and social support as predictors of psychological adjustment in
14. Belloso ML, Hidalgo EG. The role of European institutions in the defense resettled Sudanese refugees. Aust N Z J Psychiatry 2006;40:179-88.
of human rights in the Western Sahara. Estud Deusto 2016;64:329-60. 41. Laban CJ, Gernaat HB, Komproe IH et al. Impact of a long asylum proce-
15. Pandya K. The 1951 Refugee Convention is Janus-faced: it asserts as well dure on the prevalence of psychiatric disorders in Iraqi asylum seekers in
as undermines state sovereignty. https://ssrn.com/abstract=2763095. The Netherlands. J Nerv Ment Dis 2004;192:843-51.
16. Ghosh PS. Migrants, refugees and the stateless in South Asia. New Delhi: 42. Robjant K, Hassan R, Katona C. Mental health implications of detaining
SAGE Publications India, 2016. asylum seekers: systematic review. Br J Psychiatry 2004;1994:306-12.
17. Joly D. Heaven or hell?: asylum policies and refugees in Europe. London: 43. Cleveland J, Rousseau C. Psychiatric symptoms associated with brief deten-
Springer, 2016. tion of adult asylum seekers in Canada. Can J Psychiatry 2013;58:409-16.

44. Rees S. Refuge or retrauma? The impact of asylum seeker status on the 69. Rees S, Thorpe R, Tol W et al. Testing a cycle of family violence model in
wellbeing of East Timorese women asylum seekers residing in the Austra- conflict-affected, low-income countries: a qualitative study from Timor-
lian community. Australas Psychiatry 2003;11:S96-101. Leste. Soc Sci Med 2015;130:284-91.
45. Li SS, Liddell BJ, Nickerson A. The relationship between post-migration 70. Nickerson A, Bryant RA, Brooks R et al. The familial influence of loss and
stress and psychological disorders in refugees and asylum seekers. Curr trauma on refugee mental health: a multilevel path analysis. J Trauma
Psychiatry Rep 2016;18:82. Stress 2011;24:25-33.
46. Momartin S, Steel Z, Coello M et al. A comparison of the mental health of 71. Silove D, Tay A, Steel Z et al. Symptoms of post-traumatic stress disorder,
refugees with temporary versus permanent protection visas. Med J Aust severe psychological distress, explosive anger and grief amongst partners
2006;185:357. of survivors of high levels of trauma in post-conflict Timor-Leste. Psychol
47. Fazel M, Silove D. Detention of refugees: Australia has given up mandatory Med 2017;47:149-59.
detention because it damages detainees mental health. BMJ 2006;332:251. 72. Summerfield D. A critique of seven assumptions behind psychological
48. Bosworth M. Mental health in immigration detention: a literature review. trauma programmes in war-affected areas. Soc Sci Med 1999;48:1449-62.
London: Her Majestys Stationery Office, 2016. 73. Bronfenbrenner U. Ecological systems theory. London: Jessica Kingsley,
49. Brooker S, Albert S, Young P et al. Challenges to providing mental health 1992.
care in immigration detention. Geneva: Global Detention Project, 2016. 74. Hobfoll SE. Conservation of resources. A new attempt at conceptualizing
50. Rodin D, van Ommeren M. Commentary: explaining enormous varia- stress. Am Psychol 1989;44:513-24.
tions in rates of disorder in trauma-focused psychiatric epidemiology 75. Miller KE, Rasmussen A. War exposure, daily stressors, and mental health
after major emergencies. Int J Epidemiol 2009;38:1045-8. in conflict and post-conflict settings: bridging the divide between
51. Ventevogel P. Borderlands of mental health: explorations in medical anthro- trauma-focused and psychosocial frameworks. Soc Sci Med 2010;70:7-16.
pology, psychiatric epidemiology and health systems research in Afghani- 76. Miller K, Rasmussen A. The mental health of civilians displaced by armed
stan and Burundi. Amsterdam: Universiteit van Amsterdam, 2016. conflict: an ecological model of refugee distress. Epidemiol Psychiatr Sci
52. Steel Z, Marnane C, Iranpour C et al. The global prevalence of common 2017;26:129-38.
mental disorders: a systematic review and meta-analysis 1980-2013. Int J 77. Miller KE, Omidian P, Rasmussen A et al. Daily stressors, war experiences,
Epidemiol 2014;43:476-93. and mental health in Afghanistan. Transcult Psychiatry 2008;45:611-38.
53. Steel Z, Silove D, Chey T et al. Mental disorders, disability and health ser- 78. Silove D. The psychosocial effects of torture, mass human rights viola-
vice use amongst Vietnamese refugees and the host Australian populations, and refugee trauma: toward an integrated conceptual framework.
tion. Acta Psychiatr Scand 2005;111:300-9. J Nerv Ment Dis 1999;187:200-7.
54. Ventevogel P, De Vries G, Scholte WF et al. Properties of the Hopkins 79. Quosh C. Mental health, forced displacement and recovery: integrated
Symptom Checklist-25 (HSCL-25) and the Self-Reporting Questionnaire mental health and psychosocial support for urban refugees in Syria.
(SRQ-20) as screening instruments used in primary care in Afghanistan. Intervention 2013;11:295-320.
Soc Psychiatry Psychiatr Epidemiol 2007;42:328-35. 80. Tay AK, Rees S, Chen J et al. The structure of post-traumatic stress disor-
55. Tol WA, Rees SJ, Silove DM. Broadening the scope of epidemiology in der and complex post-traumatic stress disorder amongst West Papuan
conflict-affected settings: opportunities for mental health prevention and refugees. BMC Psychiatry 2013;15:111.
promotion. Epidemiol Psychiatr Sci 2013;22:197-203. 81. Hollander A-C, Dal H, Lewis G et al. Refugee migration and risk of schizo-
56. Mollica RF, Sarajlic N, Chernoff M et al. Longitudinal study of psychiatric phrenia and other non-affective psychoses: cohort study of 1.3 million
symptoms, disability, mortality, and emigration among Bosnian refugees. people in Sweden. BMJ 2016;352:i1030.
JAMA 2001;286:546-54. 82. Jones L, Asare JB, El Masri M et al. Severe mental disorders in complex
57. Hauff E, Vaglum P. Organised violence and the stress of exile. Predictors emergencies. Lancet 2009;374:654-61.
of mental health in a community cohort of Vietnamese refugees three 83. Nickerson A, Bryant RA, Silove D et al. A critical review of psychological
years after resettlement. Br J Psychiatry 1995;166:360-7. treatments of posttraumatic stress disorder in refugees. Clin Psychol Rev
58. Lie B. A 3-year follow-up study of psychosocial functioning and general 2011;31:399-417.
symptoms in settled refugees. Acta Psychiatr Scand 2002;106:415-25. 84. Hinton DE, Pich V, Hofmann SG et al. Acceptance and mindfulness tech-
59. Steel Z, Momartin S, Silove D et al. Two year psychosocial and mental niques as applied to refugee and ethnic minority populations with PTSD:
health outcomes for refugees subjected to restrictive or supportive immi- examples from culturally adapted CBT. Cogn Behav Pract 2013;20:33-46.
gration policies. Soc Sci Med 2011;72:1149-56. 85. Neuner F, Schauer M, Klaschik C et al. A comparison of narrative expo-
60. Beiser M, Hou F. Language acquisition, unemployment and depressive sure therapy, supportive counseling, and psychoeducation for treating
disorder among Southeast Asian refugees: a 10-year study. Soc Sci Med posttraumatic stress disorder in an African refugee settlement. J Consult
2001;53:1321-34. Clin Psychol 2004;72:579.
61. Silove D, Liddell B, Rees S et al. Effects of recurrent violence on post- 86. Murray LK, Dorsey S, Haroz E et al. A common elements treatment
traumatic stress disorder and severe distress in conflict-affected Timor- approach for adult mental health problems in low-and middle-income
Leste: a 6-year longitudinal study. Lancet Glob Health 2014;2:e293-300. countries. Cogn Behav Pract 2014;21:111-23.
62. Charlson FJ, Diminic S, Lund C et al. Mental and substance use disorders 87. Dawson KS, Bryant RA, Harper M et al. Problem Management Plus
in Sub-Saharan Africa: predictions of epidemiological changes and men- (PM1): a WHO transdiagnostic psychological intervention for common
tal health workforce requirements for the next 40 years. PLoS One 2014;9: mental health problems. World Psychiatry 2015;14:354-7.
e110208. 88. Buhmann CB, Nordentoft M, Ekstroem M et al. The effect of flexible
63. Momartin S, Silove D, Manicavasagar V et al. Complicated grief in Bos- cognitive-behavioural therapy and medical treatment, including antide-
nian refugees: associations with posttraumatic stress disorder and depressants on post-traumatic stress disorder and depression in traumatised
pression. Compr Psychiatry 2004;45:475-82. refugees: pragmatic randomised controlled clinical trial. Br J Psychiatry
64. First MB, Reed GM, Hyman SE et al. The development of the ICD-11 clin- 2016;208:252-9.
ical descriptions and diagnostic guidelines for mental and behavioural 89. World Health Organization. Mental health gap action programme (mhGAP):
disorders. World Psychiatry 2015;14:82-90. close the gap, dare to care. Geneva: World Health Organization, 2002.
65. Tay AK, Rees S, Chen J et al. The coherence and correlates of intermittent 90. Rees S, Tol W, Mohammad M et al. A high-risk group of pregnant women
explosive disorder amongst West Papuan refugees displaced to Papua with elevated levels of conflict-related trauma, intimate partner violence,
New Guinea. J Affect Disord 2015;177:86-94. symptoms of depression and other forms of mental distress in post-
66. Ayazi T, Swartz L, Eide AH et al. Psychotic-like experiences in a conflict- conflict Timor-Leste. Transl Psychiatry 2016;6:e725.
affected population: a cross-sectional study in South Sudan. Soc Psychia- 91. United Nations High Commissioner for Refugees. Community-based
try Psychiatr Epidemiol 2016;51:971-9. protection and mental health & psychosocial support. Geneva: United
67. Nygaard M, Sonne C, Carlsson J. Secondary psychotic features in refu- Nations High Commissioner for Refugees (in press).
gees diagnosed with post-traumatic stress disorder: a retrospective co- 92. Mirghani Z. Healing through sharing: an outreach project with Iraqi refu-
hort study. BMC Psychiatry 2017;17:5. gee volunteers in Syria. Intervention 2013;11:321-9.
68. Rees S, Silove D. Sakit Hati: a state of chronic mental distress related to 93. Ager A, Metzler J, Vojta M et al. Child friendly spaces: a systematic review
resentment and anger amongst West Papuan refugees exposed to perse- of the current evidence base on outcomes and impact. Intervention 2013;
cution. Soc Sci Med 2011;73:103-10. 11:133-47.

94. Richters A, Dekker C, Scholte WC. Community based sociotherapy in 102. World Health Organization, United Nations High Commissioner for
Byumba, Rwanda. Intervention 2008;6:100-16. Refugees. mhGAP Humanitarian Intervention Guide (mhGAP-HIG):
95. Duhumurizanye Iwacu Rwanda. Community Based Sociotherapy Pilot clinical management of mental, neurological and substance use condi-
project Kiziba-Nyabiheke refugee camps. Kigali: Duhumurizanye Iwacu tions in humanitarian emergencies. Geneva: World Health Organization,
Rwanda, 2015. 2015.
96. Scholte WF, Verduin F, Kamperman AM et al. The effect on mental health 103. World Health Organization, United Nations High Commissioner for Refu-
of a large scale psychosocial intervention for survivors of mass violence: gees. Assessment and management of conditions specifically related to
a quasi-experimental study in Rwanda. PLoS One 2011;6:e21819. stress. mhGAP Intervention Guide Module. Geneva: World Health Organi-
97. Verduin F, Smid GE, Wind TR et al. In search of links between social capi- zation, 2013.
tal, mental health and sociotherapy: a longitudinal study in Rwanda. Soc 104. United Nations High Commissioner for Refugees. Operational guidance.
Sci Med 2014;121:1-9. Mental health & psychosocial support programming for refugee opera-
98. Van Ee E, Mooren T, Kleber R. Broken mirrors: shattered relationships tions. Geneva: United Nations High Commissioner for Refugees, 2013.
within refugee families. In: Pat-Horenzcyk R, Brom D, Vogel JM (eds). 105. Tol W, Barbui C, Galappatti A et al. Mental health and psychosocial sup-
Helping children cope with trauma, individual, family and community port in humanitarian settings: linking practice and research. Lancet 2011;
perspectives. New York: Routledge, 2014:146-62. 378:1581-91.
99. Weine S, Kulauzovic Y, Klebic A et al. Evaluating a multiple-family group 106. Betancourt TS, Chambers DA. Optimizing an era of global mental health
access intervention for refugees with PTSD. J Marital Fam Ther 2008;34: implementation science. JAMA Psychiatry 2016;73:99-100.
149-64. 107. Murray L, Jordans M. Rethinking the service delivery system of psycho-
100. Batniji R, van Ommeren M, Saraceno B. Mental and social health in dis- logical interventions in low and middle income countries. BMC Psychia-
asters: relating qualitative social science research and the Sphere stan- try 2016;16:234.
dard. Soc Sci Med 2006;62:1853-64.
101. Inter-Agency Standing Committee. IASC guidelines on mental health and
psychosocial support in emergency settings. Geneva: Inter-Agency Stand- DOI:10.1002/wps.20438
ing Committee, 2007.

PERSPECTIVES
The clinical relevance of appraisals of psychotic experiences

It is not psychotic experiences in themselves but the way in The focus of cognitive-behavioural therapy for psychosis
which we appraise, or make sense of, them that determines (CBTp) is therefore on understanding and exploring these
their clinical relevance, and provides the key focus of psycho- appraisals of psychotic experiences and the thinking contrib-
logical therapy. Psychotic experiences do not inevitably cause uting to them, with the goal of supporting people to become
distress, impair functioning or result in psychiatric diagnosis. less distressed and more able to live a personally meaningful
Extensive empirical findings indicate that these experiences life. The evidence base for CBTp is now consistent in demon-
can occur in the absence of a need for care1. strating benefits for psychotic symptoms7. Developing trust
What, therefore, determines clinical pathological outcomes? and safety in the therapeutic relationship is the foundation of
Cognitive models of psychosis2 outline how the appraisals which CBTp, as for other therapies, and requires skilful competence,
people make shape both the content of psychotic experiences given the nature of peoples beliefs and the marked interper-
and the meaning that is attributed to them, bridging the gap sonal difficulties they have often experienced.
between phenomenological and neurobiological accounts of An empathic and collaborative approach is essential, con-
their occurrence3. Characteristic appraisals, for example, of veying a spirit of open enquiry, including the suspension of
psychotic experiences as betokening threat, and rendering the disbelief regarding the veracity of appraisals8. Directly chal-
self as vulnerable or worthless, are associated with need for lenging these appraisals and presenting contradictory evidence
care. These appraisals in turn are influenced by the psycho- is counter-therapeutic, as it risks invalidating peoples subjec-
logical (i.e., cognitive, affective and behavioural) processes tive experience, and may paradoxically increase their convic-
which have developed in the context of a persons genes, biol- tion and distress.
ogy and socio-environmental experiences4.
However, empathic engagement alone is insufficient to bring
A case example illustrates our proposition. James grew up in
about clinically significant improvements in people with psy-
poverty, experienced bullying and was raped during his teenage
chosis. A key mechanism of change in CBTp, consistent with
years. These early experiences led to distressing beliefs that he
psychodynamic approaches, is the development of reflective
was weak and others would harm him, and he tended to be alert functioning or the ability to make sense of ones own mind and
to potential threats. As adolescence developed into adulthood, that of others, in order to understand behaviour9. Specifically,
jobless, James became increasingly isolated and rarely went belief flexibility or slow thinking is fundamental to adaptive
outside. James felt very on-edge, and his sleep was disturbed. psychological functioning, and involves reflective curiosity and
One day, he heard whispers that sounded critical, which he was generation of alternative ideas5. There is now evidence that ther-
sure were people talking about him. He became more anxious apy which targets improvements in belief flexibility specifically
and struggled to take care of himself. He started using cannabis.
diminishes paranoia10.
The voices suddenly got more intense, telling him you are
So, whilst a developmental perspective is valuable in aiding self-
nothing and are going to get it. James just knew this was a sign
understanding, the key therapeutic focus is on identifying and
he would never escape others persecution, and he became even
modifying day-to-day cycles which maintain occurrence of dis-
more guarded and avoidant. James felt completely helpless and
tressing appraisals of psychotic experiences. As well as fast thinking
had no hope for his future.
processes, these include sensitivity to stress, threat anticipation,
Jamess difficulties highlight how adverse life experiences con-
negative affect, ruminative worrying and safety behaviours6.
tribute to negative appraisals about the self and others, which
The synthesis of an individualized narrative provides an ac-
can in the presence of a range of affective, cognitive, behaviour-
count of the range of probable factors that contribute to dis-
al, social and biological factors trigger and shape psychotic
tressing appraisals, with the goal of increasing peoples aware-
experiences and the meaning that is attributed to them. Jamess
ness of the mechanisms by which they attribute meaning to their
voices reflect the themes of how he views himself and others; and
experiences. CBTp can be seen as a process of sowing seeds to
his appraisals (I am cursed) and their consequences (I am
support the germination of alternative, less distressing ex-
helpless) also mirror his negative beliefs.
planations, which over time become more adaptive appraisals of
But note it is not just the content of appraisals that is of
psychotic experiences11. This then supports behavioural experi-
clinical relevance, but also the processes by which people
mentation in daily life, to explore the impact of modifying these
reach such conclusions and how they react to them. A certain
and trying out different ways of managing stressful, but valued
type of thinking style, fast thinking5, is particularly associated
with threatening appraisals in psychosis, and is characterized activities, with experiential learning gradually reinforcing safer
by a tendency to jump to conclusions, to have high convic- appraisals of experience.
tion in ones instincts, and to fail to consider alternative CBTp mirrors the naturalistic process through which we
explanations6. Worry and ruminative thinking further main- derive meaning from our life experiences to support adaptive
tain distressing interpretations, together with threat-focused functioning. However, sustaining this without support, given
attention, memory biases and understandable, but unhelpful, heightened vulnerability to stress, is a significant challenge. An
avoidant safety behaviours which act to prevent disconfir- important target for future research is the facilitation of endur-
mation of fears6. ing generalization of therapy gains to everyday life. To address

this, our research team is trialling a digital therapy called Maudsley NHS Foundation Trust and the Institute of Psychiatry, Psychology and
Neuroscience, Kings College London.
SlowMo that targets problematic fast thinking to modify dis-
tressing appraisals of psychotic experiences and thereby reduce 1. Linscott RJ, van Os J. Psychol Med 2013;43:1133-49.
paranoia10. A SlowMo mobile app (see www.slowmotherapy.co. 2. Garety PA, Kuipers E, Fowler D et al. Psychol Med 2001;31:189-95.
3. Howes OD, Nour M. World Psychiatry 2016;15:3-4.
uk) assists people to slow down for a moment in their daily life
4. Peters E, Ward T, Jackson M et al. World Psychiatry 2016;15:41-52.
to notice new information and develop safer thoughts, thereby 5. Kahneman D. Thinking, fast and slow. New York: Farrar, Strauss, Giroux,
aiming to optimize the clinical relevance of adaptive appraisals 2011.
6. Garety P, Freeman D. Br J Psychiatry 2013;203:327-33.
of psychotic experiences to real life.
7. van der Gaag M, Valmaggia L, Smit F. Schizophr Res 2014;156:30-7.
8. Garety P. World Psychiatry 2015;14:180-1.
Philippa A. Garety1,2, Amy Hardy1 9. Fonagy P, Gergely G, Jurist E et al. Affect regulation, mentalization and the
1
Department of Psychology, Kings College London, Institute of Psychiatry, Psychology development of the self. New York: Other Press, 2002.
and Neuroscience, London, UK; 2National Institute for Health Research Biomedical 10. Garety P, Waller H, Emsley R et al. Schizophr Bull 2015;41:400-10.
Research Centre for Mental Health, South London and Maudsley NHS Foundation 11. Moritz S, Andreou C, Schneider BC et al. Clin Psychol Rev 2014;34:358-66.
Trust, London, UK
P.A. Garety acknowledges support from the National Institute for Health Re-
search Biomedical Research Centre for Mental Health at the South London and DOI:10.1002/wps.20408
Mating, sexual selection, and the evolution of schizophrenia

For over fifty years, evolutionary theorists have sought to But how does this model account for the role of rare muta-
understand the biological roots of our species vulnerability to tions in schizophrenia? Most sexually selected traits are fitness
schizophrenia a debilitating disorder that has a relatively high indicators in that they correlate with the organisms underlying
incidence despite being associated with markedly reduced fer- condition, including good nutrition, absence of parasites, low
tility (the so-called schizophrenia paradox). While some mod- levels of harmful mutations, and so on. Other traits may evolve
els treat the entire spectrum of schizophrenia as a manifestation as amplifiers by further increasing the condition sensitivity of fit-
of biological dysfunction, others postulate that psychosis proneness indicators. In a nutshell, the SSM hypothesizes that verbal
ness (schizotypy) or even psychotic symptoms may confer adap- and artistic creativity are fitness indicators, whereas schizotypy
tive benefits through enhanced survival or reproduction (or they functions at least in part as an amplifier trait4. In other words,
used to do so during our evolutionary history)1. high schizotypy increases the risk of schizophrenia in people
Adaptive models of this kind face some formidable challenges. who carry many harmful mutations and/or are exposed to high
In addition to the low fertility of patients which is not balanced levels of stress and infections; however, the same traits boost
out by that of their close relatives and the evidence of reduced mating success in people with low mutation load and few devel-
IQ and neural integrity in schizophrenia, they need to account opmental stressors. Of course, contraception and other evolu-
for the role played by deleterious de novo mutations (including tionary novel aspects of modern societies may attenuate or break
rare copy number variations), which explain a larger share of the link between mating success and actual reproduction.
schizophrenia risk than common genetic variants1,2. The SSM potentially explains the logic of several risk factors
Schizophrenia is a heterogeneous category, and any compre- for schizophrenia, from harmful mutations and low IQ (which
hensive explanation is likely to require a combination of models. is also affected by mutation load, especially at the low end of the
At the same time, theory and evidence increasingly point to mat- distribution) to early infections and stressful life events. In addi-
ing as a contributing factor in the evolution of psychosis prone- tion, specific stressors such as migration into a minority popula-
ness. The sexual selection model (SSM) was first advanced by tion may partly operate by exacerbating competition for mates in
Nettle3 and refined by Shaner et al4. According to this model, adolescence and early adulthood. Most importantly, the SSM
schizophrenia is a maladaptive condition, but schizotypal traits offers a potential solution to the paradox of low fertility in patients
in particular positive schizotypal traits such as magical thinking, and their close relatives. According to the model, the low fertility
ideas of reference, and unusual perceptual experiences are of patients is not caused by schizotypy alone, but rather by the
associated with enhanced verbal and artistic creativity and, as a interaction between schizotypy and fitness-reducing factors such
result, lead to increased success in courtship and mating. as mutations and adversity. Close relatives of schizophrenics are
The hypermentalistic cognitive style of schizotypal individuals likely to share some of the same factors, both genetic and environ-
involves a heightened focus on others thoughts and emotions, mental. As a result, they can also be expected to show reduced
which may also contribute to courtship success5,6. Consistent fertility, though less dramatically so than patients9. If the model is
with this hypothesis, several studies have shown that positive correct, the crucial comparison would be that between the close
schizotypy is associated with artistic creativity, a larger number relatives of schizophrenics and people with similarly high levels of
of sexual partners, and a preference for uncommitted sexual rela- schizotypy but without a diagnosed relative.
tioships7. Also, a moderate degree of reduction in white matter At the genetic level, it is important to appreciate that the
integrity has been linked to creative thinking and imagination8. SSM postulates the existence of at least two distinct sources of

schizophrenia risk: a) rare and de novo mutations, which are these disorders share a functional link with fast life history-
overwhelmingly harmful and subject to negative selection; related traits such as heightened mating effort and impulsivity,
and b) schizotypy-increasing alleles, which should be relatively and form a comorbidity network with common risk factors and
common and evolve under balancing selection (a regime of developmental correlates6,7,10. They can be contrasted with slow
alternating positive and negative selection on the same allele)9. spectrum conditions, such as obsessive-compulsive personality
Common variants that influence IQ may represent a third inde- disorder, at least a subtype of autism spectrum disorder (mainly
pendent source of risk. The SSM also predicts a unique pattern in the high-functioning range), and eating disorders character-
of genotype-by-genotype interaction namely, the same delete- ized by elevated conscientiousness and self-control.
rious mutations should have a stronger effect on the risk for The taxonomy sketched above is still provisional and open
schizophrenia when they occur on a background of schizotypy- to substantial revisions. Even so, simulations show that the life
increasing common variants. To my knowledge, this hypothesis history model is already capable of reproducing the large-
has never been tested in genetic research. A recent study found scale empirical structure of mental disorders, including the
that strong negative selection on rare mutations contributes to internalizing-externalizing distinction and the emergence of a
maintain variation in other, physically close genes on the same general p factor of psychopathology10. A life history approach
chromosomes2. However, the authors did not test whether bal- recasts the SSM within a broader theoretical framework and
ancing selection may also contribute to maintain a certain integrates its insights with those of other evolutionary models,
amount of common genetic variation, independent of deleteri-
such as the diametrical model of autism and psychosis ad-
ous mutations.
vanced by Crespi and Badcock5. Together, these developments
Mating is only one component of an organisms fitness, and
are starting an exciting new chapter in the evolutionary study
needs to be balanced against other critical tasks. Examples are
of schizophrenia, with novel predictions to test and unexplored
skills acquisition, feeding, and protection of the offspring. The
implications for epidemiology, prevention, and treatment.
decisions made in allocating time and energy to these invest-
ments determine an individuals life history strategy. Life history Marco Del Giudice
strategies have wide-ranging implications for personality, behav- Department of Psychology, University of New Mexico, Albuquerque, NM, USA
ior, and physiology. In humans, fast strategies are associated
with heightened mating effort, precocious sexuality, low invest- 1. van Dongen J, Boomsma DI. Am J Med Genet 2013;162:122-36.
ment in stable couple relationships (which are conducive to par- 2. Pardi~nas AF, Holmans P, Pocklington AJ et al. bioRxiv 2016;10.1101/068593.
3. Nettle D. Strong imagination: madness, creativity and human nature. New York:
enting effort), impulsivity and risk-taking, and broad personality Oxford University Press, 2001.
traits such as low agreeableness and conscientiousness. Slow 4. Shaner A, Miller GF, Mintz J. Schizophr Res 2004;70:101-9.
strategies are associated with lower mating and higher parenting 5. Crespi B, Badcock C. Behav Brain Sci 2008;31:241-61.
effort, delayed sexuality, fewer partners, self-control and risk 6. Del Giudice M, Angeleri R, Brizio A et al. Front Psychol 2010;1:41.
7. Del Giudice M, Klimczuk ACE, Traficonte DM et al. Evol Hum Behav 2014;
aversion, and high agreeableness and conscientiousness. 35:415-24.
Life history concepts can be used to develop a broad-band 8. Jung RE, Grazioplene R, Caprihan A et al. PLoS One 2010;5:e9818.
evolutionary taxonomy of mental disorders10. In this framework, 9. Del Giudice M. PLoS One 2010;5:e16040.
10. Del Giudice M. Clin Psychol Sci 2016;4:299-311.
schizophrenia spectrum disorders can be classified as fast spec-
trum conditions, together with borderline personality disorder,
antisocial and conduct disorders, and eating disorders marked
by behavioral dysregulation. Above and beyond their differences, DOI:10.1002/wps.20409
Validity and utility of the general factor of psychopathology

Psychopathology can be viewed as a variety of symptoms that the correlations between internalizing and externalizing
that are organized into first-order dimensions by their correla- factors can be explained by a general factor of psychopathol-
tions. Critically, these first-order dimensions are themselves ogy on which every first-order dimension loads7. This finding
robustly correlated1. These correlations are problematic for has been replicated many times across the lifespan4. Most
categorical taxonomies2, but provide essential information studies examined only prevalent forms of psychopathology,
about the nature of psychopathology3-5. Correlations among but several showed that bipolar disorder, schizophrenia and
first-order dimensions vary in magnitude, with stronger corre- autism are strongly related to the general factor of psychopa-
lations among some dimensions yielding second-order factors, thology, suggesting that this factor is very general indeed4.
particularly internalizing and externalizing factors6. Before deciding that the general factor of psychopathology is
These second-order factors do not completely capture the useful, we must know if it is only an artifact of systematic mea-
correlations among dimensions of psychopathology, however. surement error. The general factor almost certainly partly reflects
Rather, second-order internalizing and externalizing factors nuisance correlations due to the same informant reporting on all
are themselves substantially correlated. We provided evidence psychopathology dimensions, but it must also capture something

substantive to have utility. We have addressed this issue ratio- nisms underlying each first-order dimension of psychopathology.
nally4, but ultimately it reduces to an empirical question of crite- Each first-order dimension is heterogeneous in its etiologies and
rion validity. If the general factor is more than a measurement mechanisms for the same reasons that different dimensions
artifact, it will be significantly correlated with variables that are are correlated. That is, the etiologic influences on each first-
external to its definition but central to its validity. Critically, the order dimension of psychopathology are heterogeneous large-
general factor is robustly correlated with measures of cognitive ly because they arise from (at least) three separate and largely
ability and the dispositional dimension of negative emotionality. orthogonal sources. Some persons exhibiting high levels of
Furthermore, controlling for internalizing and externalizing psy- symptoms in any dimension of psychopathology may carry
chopathology, demographic factors and intelligence, the general only risk genotypes that pleiotropically increase risk for all
factor robustly predicts both concurrent and future adaptive func- dimensions of psychopathology through the general factor.
tioning, even when symptoms and functioning are measured by Other persons with the same symptoms may carry only geno-
different informants4. types that increase risk for all externalizing (or all internaliz-
Can the general factor facilitate studies of the nature of ing) dimensions, and others may carry only genotypes that are
psychopathology and ultimately improve prevention and treat- specific to that dimension of symptoms. Many others will
ment? We have hypothesized that first-order dimensions of psy- carry varying combinations of genotypes from each of these
chopathology are correlated because they have shared causes. sources. The result is an intractable degree of heterogeneity in
Large twin and sibling studies of children, adolescents and the genetic influences if first-order dimensions are studied
adults indicate that the general factor is moderately heritable8 individually. It should be far more efficient to identify such
and that phenotypic correlations among the first-order dimen- diverse etiologic influences and their related mechanisms at
sions are largely attributable to shared genetic influences9, with their source by modeling higher-order phenotypes than by
less than half of the genetic variance on most first-order dimen- attempting to fractionate each first-order dimension into its
sions being dimension-specific5. diverse etiologies and mechanisms.
These findings support the view that genetic risk factors This causal taxonomy suggests the need for major changes in
for psychopathology often function pleiotropically10, but they how the etiologies and mechanisms of apparently diverse forms
suggest a previously unsuspected breadth of pleiotropy, with of psychopathology are conceptualized and studied. Case-control
a significant proportion of genetic factors non-specifically in- samples are the current standard for such research. They are
creasing risk for all dimensions of psychopathology. This optimized for identifying dimension-specific causes, but bias
implies that genetic research will be facilitated by letting correlations among first-order dimensions of psychopathology,
genetic correlations rather than ICD and DSM committees making the modeling of higher-order phenotypes complicated or
define optimal phenotypes. In concrete terms, if a genetic impossible. In contrast, large representative samples that include
variant that is robustly related to the general factor were sufficient variation in all psychopathology dimensions to model
instead tested for association with, say, depression, all cases in higher-order factors of psychopathology can inform every level
which the variant was present but the individual exhibited
of the hierarchy.
high levels of any other dimension of psychopathology would
erroneously counted as misses instead of hits. Benjamin B. Lahey1, Robert F. Krueger2, Paul J. Rathouz3,
The general factor of psychopathology also implies that Irwin D. Waldman4, David H. Zald5
1
first-order dimensions of psychopathology do not each have University of Chicago, Chicago, USA; 2University of Minnesota, Minneapolis, MN,
USA; 3University of Wisconsin, Madison, WI, USA; 4Emory University, Atlanta, GA,
their own entirely unique pathophysiologies. Dimensions of USA; 5Vanderbilt University, Nashville, TN, USA
psychopathology are too highly correlated and there is too
much sharing of genetic and environmental influences at the The authors are supported by grant R01-MH098098 from the US National
Institute of Mental Health.
level of higher-order factors not to hypothesize that variations
in some neurobiological systems non-specifically underlie mul-
tiple dimensions of psychopathology. 1. Krueger RF, Markon KE. Ann Rev Clin Psychol 2006;2:111-33.
2. Meehl PE. Clin Psychol Sci Pract 2001;8:507-19.
We recently proposed a formal causal taxonomy of psycho- 3. Angold A, Costello EJ. J Child Psychol Psychiatry 2009;50:9-15.
pathology in which the robust correlational structure of first- 4. Lahey BB, Krueger RF, Rathouz PJ et al. Psychol Bull 2017;143:142-86.
order dimensions is attributed to a hierarchy of increasingly 5. Lahey BB, Van Hulle CA, Singh AL et al. Arch Gen Psychiatry 2011;68:
181-9.
specific etiologic influences4. In this model, some non-specific
6. Achenbach TM, Conners CK, Quay HC et al. J Abnorm Child Psychol 1989;
etiologic factors increase risk for all first-order dimensions of 17:299-323.
psychopathology to varying degrees through the general fac- 7. Lahey BB, Applegate B, Hakes JK et al. J Abnorm Psychol 2012;121:
tor. Other non-specific etiologic factors increase risk only for 971-7.
8. Waldman ID, Poore H, Van Hulle C et al. J Abnorm Psychopathol 2016;125:
all first-order dimensions within the internalizing or the exter- 1053-66.
nalizing domains, and each first-order dimension has its own 9. Pettersson E, Larsson H, Lichtenstein P. Mol Psychiatry 2016;21:717-21.
unique causal influences. 10. Kendler KS. Am J Psychiatry 2005;162:1243-52.
This causal taxonomy addresses more than just the sharing
of causal influences. It also supports novel hypotheses regard-
ing the equally important heterogeneity of causes and mecha- DOI:10.1002/wps.20410

Neuroticism is a fundamental domain of personality with enormous
public health implications
Neuroticism is the trait disposition to experience negative physical and mental health care concerns, the authors of the UP
affects, including anger, anxiety, self-consciousness, irritability, again note that the public-health implications of directly treating
emotional instability, and depression1. Persons with elevated and even preventing the development of neuroticism would be
levels of neuroticism respond poorly to environmental stress, substantial7.
interpret ordinary situations as threatening, and can experience Neuroticism has long been recognized since the beginning
minor frustrations as hopelessly overwhelming. Neuroticism of basic science personality research and may even be the first
is one of the more well established and empirically validated domain of personality that was identified within psychology1.
personality trait domains, with a substantial body of research to Given its central importance for so many different forms of men-
support its heritability, childhood antecedents, temporal stabil- tal and physical dysfunction, it is not surprising that neuroticism
ity across the life span, and universal presence1,2. is evident within the predominant models of personality, person-
Neuroticism has enormous public health implications3. It ality disorder, and psychopathology.
provides a dispositional vulnerability for a wide array of different Neuroticism is one of the fundamental domains of general
forms of psychopathology, including anxiety, mood, substance, personality included within the five-factor model or Big Five2. It
somatic symptom, and eating disorders1,4. Many instances of is also within the dimensional trait model included in Section
maladaptive substance use are efforts to quell or quash the III of the DSM-5 for emerging measures and models8. This
dismay, anxiousness, dysphoria, and emotional instability of neu- trait model consists of five broad domains, including negative
roticism. Clinically significant episodes of anxiety and depressed affectivity (along with detachment, psychoticism, antagonism,
mood states will often represent an interaction of the trait or and disinhibition). As expressed in the DSM-5, these five broad
temperament of neuroticism with a life stressor1. domains are maladaptive variants of the five domains of the
Neuroticism is comparably associated with a wide array of extensively validated and replicated personality model known
physical maladies, such as cardiac problems, disrupted immune as the Big Five or Five Factor Model of personality8.
functioning, asthma, atopic eczema, irritable bowel syndrome, Neuroticism is likewise aligned with the negative affective
and even increased risk for mortality2. The relationship of neu- domain included within the dimensional trait model of per-
roticism to physical problems is both direct and indirect, in that sonality disorder proposed for the ICD-119. Finally, it is also
neuroticism provides a vulnerability for the development of these evident within the transdiagnostic Research Domain Criteria
conditions, as well as a disposition to exaggerate their impor-
(RDoC) of the National Institute of Mental Health, as RDoC
tance and a failure to respond effectively to their treatment.
negative valence encapsulates such constructs as fear, distress,
Neuroticism is also associated with a diminished quality of life,
frustration, and perceived loss10. It would be inaccurate to sug-
including feelings of ill-will, excessive worry, occupational failure,
gest that RDoC negative valence is equivalent to neuroticism,
and marital dissatisfaction5. High levels of neuroticism will con-
but it is self-evident that they are closely aligned.
tribute to poor work performance due to emotional preoccupa-
Currently, there is considerable interest in the general fac-
tion, exhaustion, and distraction. Similar to the duel-edged effect
tors of psychopathology, personality disorder, and personality.
of neuroticism on physical conditions, high levels of neuroticism
To the extent that degree of impairment and dysfunction
will result in actual impairment to marital relationships but al-
(which largely defines the general factors) is associated with
so subjective feelings of marital dissatisfaction even when there is
level of distress and dismay, which is quite likely to be the case,
no objective basis for such feelings, which can though in turn
we would propose that neuroticism will explain a substantial
lead to actual spousal frustration and withdrawal.
proportion of the variance in those general factors.
Given the contribution of neuroticism to so many negative life
In sum, neuroticism is a fundamental domain of personality
outcomes, it has been recommended that the general population
that has enormous public health implications, impacting a wide
be screened for clinically significant levels of neuroticism during
array of psychopathological and physical health care concerns. It
routine medical visits1,6. Screening in the absence of available
contributes to the occurrence of many significantly harmful life
treatment would be problematic. However, neuroticism is respon-
outcomes, as well as impairing the ability of persons to ade-
sive to pharmacologic intervention1. Pharmacotherapy can and
quately address them. It has long been recognized as one of the
does effectively lower levels of the personality trait of neuroti-
more important and significant domains of personality and is
cism. Barlow et al7 have also developed an empirically-validated
being increasingly recognized as a fundamental domain of per-
cognitive-behavioral treatment of neuroticism, called the Unified
sonality disorder and psychopathology more generally.
Protocol (UP). They have suggested that current psychological
treatments have become overly specialized, focusing on disorder- Thomas A. Widiger, Joshua R. Oltmanns
specific symptoms. The UP was designed to be transdiagnostic. Department of Psychology, University of Kentucky, Lexington, KY, USA
Recognizing the impact of neuroticism across a diverse array of

1. Widiger TA. In: Leary MR, Hoyle RH (eds). Handbook of individual 7. Barlow DH, Sauer-Zavala S, Carl JR et al. Clin Psychol Sci 2014;2:344-65.
differences in social behavior. New York: Guilford, 2009:129-46. 8. American Psychiatric Association. Diagnostic and statistical manual of
2. Tackett JL, Lahey BB. In: Widiger TA (ed). The Oxford handbook of the five mental disorders, 5th ed. Arlington: American Psychiatric Association, 2013.
factor model. New York: Oxford University Press (in press). 9. Tyrer P, Reed GM, Crawford MJ. Lancet 2015;385:717-26.
3. Lahey BB. Am Psychol 2009;64:241-56. 10. Sanislow CA, Pine DS, Quinn KJ et al. J Abnorm Psychol 2010;119:631-9.
4. Bagby RM, Uliaszek AA, Gralnick TM et al. In: Widiger TA (ed). The Oxford
handbook of the five factor model. New York: Oxford University Press
(in press). DOI:10.1002/wps.20411
5. Ozer DJ, Benet-Martinez V. Annu Rev Psychol 2006;57:401-21.
6. Widiger TA, Trull TJ. Am Psychol 2007;62:71-83.

FORUM SHARED DECISION MAKING IN MENTAL HEALTH CARE
Implementing shared decision making in routine mental health care

Mike Slade
Institute of Mental Health, School of Health Sciences, University of Nottingham, Nottingham, UK
Shared decision making (SDM) in mental health care involves clinicians and patients working together to make decisions. The key elements of
SDM have been identified, decision support tools have been developed, and SDM has been recommended in mental health at policy level. Yet
implementation remains limited. Two justifications are typically advanced in support of SDM. The clinical justification is that SDM leads to
improved outcome, yet the available empirical evidence base is inconclusive. The ethical justification is that SDM is a right, but clinicians
need to balance the biomedical ethical principles of autonomy and justice with beneficence and non-maleficence. It is argued that SDM is
polyvalent, a sociological concept which describes an idea commanding superficial but not deep agreement between disparate stakeholders.
Implementing SDM in routine mental health services is as much a cultural as a technical problem. Three challenges are identified: creating
widespread access to high-quality decision support tools; integrating SDM with other recovery-supporting interventions; and responding to
cultural changes as patients develop the normal expectations of citizenship. Two approaches which may inform responses in the mental
health system to these cultural changes social marketing and the hospitality industry are identified.
Key words: Shared decision making, mental health care, ethics, implementation, routine outcome monitoring, social marketing
Decision making is a complex and dy- complete a diagnostic test, undergo a health care settings, identifying five ran-
namic social interaction1. The balance of medical procedure, receive a particular domized controlled trials of interven-
involvement between clinician and patient pharmacological or psychological treat- tions to improve clinicians adoption of
can be conceptualized as lying on a con- ment, or attempt a lifestyle change. In SDM7. Training of clinicians and use of
tinuum from clinician-led/passive/pater- mental health, decisions relating to in- decision aids (structured approaches to
nalistic, through shared, to patient-led/ patient care are broadly similar. When facilitate SDM) were tentatively recom-
informed/active2. Clinician-led decision asked to name recent clinical decisions, mended, though none of the studies
making occurs when the clinician makes inpatients with a diagnosis of schizophre- related to mental health populations.
the decision for the patient, possibly after nia (N560) and their psychiatrists (N530) Patients want SDM8. A systematic re-
consulting with him/her. Patient-led de- consistently mentioned categories such as view of 199 analyses from 115 studies of
cision making occurs when the patient medication, leave from ward/hospital, decision-making style preference con-
makes the decision, possibly having re- non-pharmacological therapies and cluded that patients prefer shared to
ceived information from the clinician. changes in treatment setting4. By con- clinician-led decision making, with the
The intermediate position of shared deci- trast, decision making in community men- preference proportion higher in studies
sion making (SDM) involves collaboration. tal health settings is more wide-ranging; a carried out in patients with cancer or
A widely used definition of SDM is that principal component analysis of topics dis- undergoing invasive procedures, compared
it is a process in which clinicians and cussed in routine consultations between to those conducted in non-disease specific
patients work together to select tests, treat- community patients (N5418) and their cli- study populations or patients with other
ments, management or support packages, nicians found a three-factor solution com- chronic conditions9.
based on clinical evidence and the patients prising treatment, social (family, friends, Overall, there is international consen-
informed preferences; it involves the provi- leisure) and financial (work, benefits)5. sus across medicine about the importance
sion of evidence-based information about The essential elements of SDM have of SDM10, and it is widely supported11. It
options, outcomes and uncertainties, to- been identified. A systematic review syn- is argued that SDM leads to better out-
gether with decision support counselling thesized 161 conceptual models of SDM comes, including help-seeking behaviour12,
and a system for recording and imple- to identify eight characteristics of clini- increased compliance with decisions13,
menting patients informed preferences3. cian behaviour: define/explain the health reduction in errors14, reduced stigma and
This definition focuses, as does the present care problem, present options, discuss increased involvement15. In 2010, a gath-
paper, on interactions between clinicians benefits/risks/costs, clarify patient val- ering of 58 experts from 18 countries pro-
and patients, but SDM also has relevance ues/preferences, discuss patient ability/ duced the Salzburg Statement on Shared
to decision making between clinicians self-efficacy, present what is known and Decision Making16. This included a call
and family members, and perhaps also to make recommendations, clarify the pa- for clinicians to recognize SDM as an eth-
clinical discussion between different pro- tients understanding, and make or ex- ical imperative, stimulate two-way flow
fessional groups. plicitly defer a decision6. This framework of accurate and tailored information, and
What is a decision? In physical health underpinned a systematic review of im- give patients and their families resources
care, decisions might include whether to plementation of SDM across different and help to reach decisions. The statement

also exhorted action by researchers, editors, Both studies took place in Germany, one a patient preference for SDM in 63% of
journalists, patients (to speak up, to expect involving 107 patients with a schizophre- studies, but a time trend was evident, with
to be an equal partner, to seek and use high- nia diagnosis22 and the other 405 patients 50% of studies before 2000 and 71% after
quality information) and policy makers. with depression23. The Cochrane review 2000 showing this preference.
concluded that there was no evidence for Reviews of SDM in persons with schiz-
harm, but the weak evidence base meant ophrenia30 and depression31 showed that
that no firm conclusions could be drawn. patients and clinicians found SDM ac-
SHARED DECISION MAKING IS
Since that review, one randomized con- ceptable and did in fact engage in SDM,
RECOMMENDED IN MENTAL
trolled trial involving 80 community pa- which resulted in improvements in pa-
HEALTH
tients24, also showing advantages for de- tients knowledge about their illness and
cision aids, has been published. a higher level of perceived involvement in
SDM is promoted in mental health
Other reviews have reached similar decision making.
systems17. It is advocated as an impor-
conclusions. A systematic review25 identi- The ethical justification is often posi-
tant approach in the mental health poli-
fied eleven randomized controlled trials, tioned as a solution to the suggested prob-
cy of many countries internationally10.
including two in mental health, one lem of an assumption that the clinician is
For example, in England it is recom-
focussing on schizophrenia26 and the oth- the only competent decision maker, who
mended that a shared decision making
er on depression27. Five trials, including will make decisions for rather than with
approach should be facilitated across
the two mental health trials, showed posi- the patient. Ethical justifications empha-
all adult mental health services18.
tive outcomes associated with SDM, but size that clinicians and patients bring
Why is SDM in mental health so widely
the reviewers concluded that the overall different but equally important forms of
recommended? The standard argument
evidence is encouraging but inconclusive. expertise to the decision-making proc-
made to support SDM is that clinicians
It should be noted that this conclu- ess3. Arguments made from this perspec-
have expertise in diagnosis, etiology, pro-
sion is not unique to mental health. The tive often focus on values and power
gnosis, treatment options and outcome
most recent systematic review of trials relationships, for example by linking SDM
probabilities, whereas patients have ex-
(N522) testing the impact of SDM on with values-based practice32. SDM is un-
pertise in illness experience, social cir-
outcome in physical health concluded: derstood primarily as a process involving
cumstances, attitudes to risk, values and
The trials performed to date to address the expert-by-training (the clinician) and
preferences3. Bringing these two types of
expertise together can, when informed by the effect of SDM on patient-relevant, the expert-by-experience (the patient) both
research evidence, produce better deci- disease-related endpoints are insufficient contributing their expertise, committing to
sions. However, this standard argument in both quantity and quality. Although decision-making responsibility, and being
conflates two overlapping but separate just under half of the trials reviewed here respectful of the others perspective. This
justifications: the clinical and the ethical. indicated a positive effect, no final con- transactional focus contrasts with the clin-
clusion can be drawn28. But available ical justification emphasis on producing
evidence does suggest that SDM in men- better outcomes.
The clinical justification tal health is particularly challenging. For
example, SDM leads to a greater increase Shared decision making is a
The clinical justification put forward
in treatment adherence in general medi- polyvalent concept
for SDM is that patients who are active
cine than in mental health29.
participants in managing their care have
Overall, the totality of evidence is in- SDM is thus supported both by those
better outcomes. Increased involvement
conclusive about the impact of SDM on who prioritize clinical expertise and ex-
will lead to better engagement, higher-
patient outcomes in mental health. pertise-by-experience. In this sense, the
quality decision making, and increased
treatment adherence all of which will term is what sociologists call a polyvalent
improve outcome. There is some evidence The ethical justification concept33 one which commands super-
supporting this justification. For example, ficial agreement and apparent consensus
a trial in the Netherlands involving 220 The ethical justification put forward between disparate stakeholders, but which
psychiatric inpatients showed that SDM for SDM is that it is a human right. conceals incompatible assumptions and
led to reduced substance use and im- Sometimes expressed as No decision expectations. Put concretely, does the cli-
proved quality of life19. A follow-up study about me without me3, the right to self- nician still support SDM if it leads to em-
found that SDM was also associated with determination implies full involvement powered patients who are less adherent to
increases in patient autonomy20. in decisions affecting the person. This treatment recommendations? Does the
However, critical appraisal of all avail- seems to be a view increasingly taken by patient still support SDM if apparently in-
able evidence is less positive. A Cochrane patients: the above-mentioned 2012 sys- volving conversations that seem somehow
review of SDM in mental health21 identified tematic review of 115 studies investigating always to end up with the clinicians view
only two randomized controlled trials. decision-making preferences9 identified prevailing34?

There are particular challenges in men- Only 50% fully agreed with the statement understandings of, and low commitment
tal health care35. Is SDM still the best Were you involved as much as you wanted to, SDM54, lack of a rights discourse in
approach to decision making with non- to be in decisions about which medicines the culture55, and challenges of avoiding
capacitous adults, such as those with ad- you receive? (N59,775), and among pa- inequities when access to support tools is
vanced dementia or acute psychosis36? Is tients who received non-pharmacological through insurance-funded health systems56.
it appropriate in a forensic context, where treatments, only 55% fully agreed with
the decisions that the person makes may Were you involved as much as you wanted
RESEARCH IN ROUTINE CLINICAL
fall slightly or greatly outside social norms? to be in deciding what treatments or ther-
SETTINGS
These tensions between different jus- apies to use?.
tifications for shared decision making Is there a difference between SDM in
Given these implementation challenges,
also occur in other initiatives in mental mental versus physical health? A study
research in routine mental health services
health. The same features of apparent uni- in the Canary Islands compared experi-
is needed. The European Union-funded
versal agreement occur in relation to the ence of decision making between pa-
Clinical decision making and outcome in
service agenda and rights agenda which tients attending psychiatric outpatient
routine care for people with severe mental
both provide support for anti-stigma ini- clinics and primary care (N51,477)49. It
illness (CEDAR) study took place in six
tiatives37. Other polyvalent constructs in- found no difference in overall score, but
European countries (Denmark, Germany,
clude self-management, advance directives differences at the item level. Participants
Hungary, Italy, Switzerland and UK) from
and social inclusion. using psychiatric outpatient services
2009 until 201457. The study had two aims.
For example, recovery has emerged as said that they were helped to understand
The first aim was to establish a meth-
a guiding vision for mental health sys- the information, but were more likely to
odology to assess clinical decision mak-
tems38. Like the ethical justification for say that they were not asked about which
ing in people with severe mental illness.
SDM, a recovery orientation involves a re- treatment option they preferred, that there
This aim was met by the development
focussing on subjectively-defined process was no negotiation, and that the selection and cross-cultural validation of three
rather than clinician-defined outcome. of treatment was not a consensus deci- new measures. All of them comprised
The relevance of recovery to dementia39, sion. There may be challenges specific to parallel clinician and patient versions,
forensic40 and mental health inpatient SDM in mental health. and were developed in English followed
services41, however, has been questioned. A qualitative investigation of the views by rigorous translation and cultural adap-
A focus on recovery creates challenges for of experienced psychiatrists (N526) iden- tation using good practice guidelines58
clinicians and patients. Clinicians have tified barriers to its use in relation to pre- into Danish, German, Hungarian and
the uncomfortable experience of compet- scribing50. The most frequently identified Italian. The Clinical Decision Making in
ing priorities42 leading to role tensions43, barrier was beliefs about the insight of Routine Care (CDRC) measure assesses
yet advocates raise concerns that recov-
the patient, which in some cases was seen the content and implementation of deci-
ery is being commandeered44 to indi-
as an absolute barrier. Other challenges sions59. The Clinical Decision Making
vidualize social problems, to de-politicize
were societal expectations about mental Style (CDMS) measure assesses prefer-
individual experience and to remain fo-
disorder (so statutory powers are held by ence for different styles of decision mak-
cussed on deficit amelioration45. The rec-
the psychiatrist), beliefs about the primacy ing60. The Clinical Decision-making In-
ommendation that sociological research
and the tranquillizing effects of antipsy- volvement and Satisfaction (CDIS) mea-
is needed to understand the socio-cultural
chotic medication, and financial pressures sure assesses involvement and satisfac-
meaning and implications of recovery46
limiting options. tion in a specific decision. All measures
is probably equally applicable to SDM.
These barriers may lead to SDM con- are available at www.cedar-net.eu/instru-
versations in mental health being more ments.html.
factual than values-based. An explora- The second aim was to investigate deci-
HOW IS SHARED DECISION tion using factor analysis of decision sion making in routine adult community-
MAKING IMPLEMENTED IN making in psychiatric visits in the US based mental health services, using a six-
MENTAL HEALTH? (N5191) found that discussions about country prospective observational design.
the science (pros and cons, clinical issues A total of 588 patients met inclusion crite-
SDM is not yet widely implemented and uncertainties, consumers goals and ria, primarily aged 18-60, with a diagnosis
across mental health systems. For exam- understanding) were more common than of a mental disorder (established using re-
ple, in the National Health Service (NHS) about preferences (the consumers role in search criteria61) severe62 and enduring for
Community Mental Health Survey 2015 decision making, consideration of alter- two years. After giving consent, patients
in England47, only 42% a reduction with natives, exploration of preferences)51. identified a clinician, and these clinician-
respect to 201448 fully agreed with the Other implementation challenges have patient dyads were then asked to complete
statement Have you agreed with some- been identified in physical health10 and bimonthly assessments for one year.
one from NHS mental health services mental health52 settings, such as hierarchi- The main study investigated the rela-
what care you will receive? (N512,695). cal doctor-patient relationships53, differing tionship between decision making style

and outcome63. A preference for shared, decision making. This is consistent with do we minimize harm, balancing the re-
rather than patient-led or clinician-led, findings from other health sectors. For ality that being allowed to disengage from
decision making was reported by both example, a primary care study (N51,913) services leads to the best outcome for
patients (v25135.08, p<0.001) and clini- in Germany found that high experienced some people70 and to avoidable tragedies
cians (v25368.17, p<0.001). SDM was involvement predicted higher patient satis- for others?
also the dominant experience, with a faction65. The second implication is that an ori-
10% increase in the proportion of both The CEDAR study has two implica- entation towards SDM is an empirically
groups reporting SDM over the one-year tions for routine practice. First, if the in- defensible goal in mental health systems
study period. Hierarchical linear model- tention is to reduce patient-rated unmet which have traditionally used clinician-
ling found that the decision-making style needs and to maximize satisfaction, then led decision making. An SDM orientation
of clinicians significantly affected patient- the empirical findings indicate that long- will improve both patient experiences
rated unmet needs over time, with unmet term efforts should be oriented towards and outcomes, indicating an alignment
needs decreasing more in patients whose developing patient-led rather than shared between the clinical and ethical justifica-
clinicians preferred patient-led to clini- decision making. This is challenging to tions for SDM as a more beneficial style
cian-led (20.406 unmet needs per two the current culture of health services. than clinician-led decision making. If it is
months, p50.007) or shared (20.303 un- Patient-led decision making is not always accepted that SDM is a necessary compo-
met needs per two months, p50.015) valued by the system; a patient prefer- nent of a modern mental health system,
decision making. In other words, outcomes ence for involvement has been found to then three challenges can be identified:
were best when clinicians supported be negatively associated with experienced the technical problems of access to ap-
patient-led decision making. involvement65. Socio-political debate would propriate tools and integration with other
A second study investigated the rela- be needed about the purpose of the mental innovations, and addressing the implica-
tionship between decision-making in- health system to what extent is the core tions of changing culture.
volvement and satisfaction64. Patients business of the system keeping people
(N5445) were partitioned based on in- (patients and others) safe, which may nec-
volvement preferences (assessed using essarily involve some clinician-led deci- DECISION SUPPORT TOOLS
CDMS) and experiences (assessed using sion making, versus supporting them to
CDIS). The preference hypothesis was live as well as possible? Can and should Changing practice often involves the
that satisfaction with a specific decision we socialize clinicians into a professional use of formal decision support tools,
will be higher if it is made using the role which gives primacy to patient-led and resources exist to support SDM. For
patients preferred decision-making style decision making? Clinical practice would example, online decisions support sys-
(patient-led, shared or clinician-led). This need to be oriented towards supporting tems are available which are both gener-
was not confirmed. Overall, 90 patients this type of patient empowerment, with a ic (e.g., optiongrid.org) and condition-
(20%) had less involvement than pre- recovery-oriented culture in mental health specific (e.g., sdm.rightcare.nhs.uk/pda
ferred (disempowered), 190 (43%) were systems which promotes the normal enti- for depression).
matched and 162 (37%) were em- tlements of citizenship66. We know that These tools may target behaviour change
powered. Empowered patients, who ex- the desire to participate in decision mak- in either clinicians or patients. Clinician-
perienced more involvement in decision ing is higher in some groups of patients, focussed approaches typically involve
making than they desired, rated highest e.g., inpatients with experiences of invol- training and support for practice change.
satisfaction (OR52.47, p50.005, 95% CI: untary treatment, with negative attitudes These approaches have been evaluated in
1.32-4.63). The agreement hypothesis was toward medication, with a higher level of depression, and (when augmented with
that satisfaction will be higher when deci- education, with lower treatment satisfac- patient information leaflets giving infor-
sions are made with a clinician with the tion, with better perceived decision-making mation and encouragement towards in-
same preferred decision-making style. skills, in patients of female gender and in volvement) they lead to improved patient
This was also not confirmed, since ordinal younger patients30. Should efforts to sup- participation and satisfaction without add-
logistic regression modelling showed that port patient-led decision making be tar- ing to consultation time23.
decisions made with clinicians whose geted at these patient subgroups, or at all A good example of a patient-focussed
decision-making style preference was for patients? approach is the Common Ground sys-
more active involvement than the patient Also, patients may bring expectations tem, which is an online peer-delivered
preference were rated with highest satis- about being looked after whilst unwell. system to support patient involvement
faction (OR53.17, p50.003, 95% CI: 1.48- When is this expectation helpful, and and empowerment in psychopharma-
6.82). So, higher satisfaction was experi- when is it ultimately harmful? Recovery cology consultations71.
enced following more active involvement is far more common than often under- Widespread access to generic and
in decision making than the patient stated stood in mental health systems67,68, and condition-specific decision support tools
as desired, and with a clinical orientation access to peer workers can powerfully is needed. Tools need to be of a high qual-
towards empowering, rather than shared, transform these role expectations69. How ity: a systematic review of decision aids

across medicine found a tendency to due to mental illness. An emergent prob- factors including the patient-clinician rela-
under-specify the procedure, to empha- lem with this patient-led approach was tionship, fear of being judged, perceived
size benefits more than harms, and to that the clinician might not be involved in, inadequacy, and a history of substance
focus more on false positives than on false or even aware of, the directive in advance, abuse87. The use of clinician-led decision
negatives in screening tools72. Develop- leading to low implementation78. A variant making is most pronounced in treatment-
ment of reporting guidelines for decision involving SDM has emerged, called joint related decisions5.
aid studies would be one approach to im- crisis plans. These are developed through One reason for low implementation is
proving quality73. facilitated meetings between the patient represented by ethical tensions. A widely-
Decision support tools also need to and involved clinicians79. A randomized used biomedical ethical framework iden-
be small in number: the same systematic controlled trial involving 569 patients in tifies four principles: respect for au-
review identified 68 tools relating to 64 community mental health teams in tonomy, justice, beneficence and non-
treatment and 30 relating to screening. England found that implementation by cli- maleficence88. Skilled clinicians attempt
This variation makes benchmarking and nicians was the main challenge, with no to integrate these principles, for example
comparison between services and sys- significant treatment effect for the primary supporting patient participation not just
tems more difficult28. Finally, there needs outcome of compulsory admissions, or for reasons of autonomy but also justi-
to be a focus on tailoring and testing any secondary outcome with the exception fied by beneficence (as well as other
tools in different clinical groups and geo- of improved therapeutic relationships80. influences, such as avoiding legal liabili-
graphical locations. The extent to which Qualitative investigation identified four ty)89. However, engagement remains chal-
patients expect to be actively involved in barriers to clinician engagement: ambiva- lenging90. The potential conflict between
treatment decisions varies according to lence about care planning; perceptions these principles has been characterized
the prevailing culture74. In paternalistic that they were already doing SDM; con- in relation to antipsychotic prescribing
cultures, both clinicians and patients are cerns regarding the clinical appropri- for a patient who lacks insight; the psy-
likely to assume that decisions are the ateness of service users choices; and chiatrist may think: If I leave it up to the
responsibility of the clinician only, where- limited availability of service users choic- patient, he would certainly choose not to
as in more egalitarian cultures a partner- es81. initiate treatment. Symptoms would per-
ship or SDM approach may be jointly Another example of integration is sist or even worsen, and thus I would
preferred75. Translation processes there- with the emergent field of routine out- harm the patient. If I apply pressure and
fore need to address these cultural factors come monitoring82, which involves the he accepts antipsychotics, he may
in ensuring both linguistic and conceptual longitudinal collection of patient-level respond to treatment and likely gain
equivalence58. outcome information to inform individ- insight. Then he will later be thankful
ualized care. There is strong evidence of that I proceeded in the way I did91. This
short-term benefit and moderate evi- reflects the tension between deontolo-
dence of longer-term benefit from rou- gical (duty-based) ethical frameworks
INTEGRATION WITH OTHER tine outcome monitoring83. A study is emphasized in the training of many pro-
RECOVERY-SUPPORTING now underway which integrates SDM
INNOVATIONS fessional groups and teleological (rights-
and that monitoring84. Routinely collect- based) frameworks emphasized by citi-
ed outcome data are fed into the SDM
Implementation of SDM will involve zens.
process, with the intervention supported
the integration of the relevant technolo- A second reason for low implementa-
by a quality improvement collaborative
gies with wider innovations, and the tion is cultural. An asylum-based system
programme involving a national and lo-
application of improvement science to creates a micro-culture (a total insti-
cal implementation strategy.
support evaluation and sustainable im- tution92) which can be out of step with
plementation. A number of measures of wider cultural values. Institutional struc-
SDM now exist: a structured review iden- tures can powerfully socialize a patient
tified 19 measures, and a move towards ETHICAL AND CULTURAL into a moral duty to be treatment-adher-
measuring processes from both patient CHALLENGES OF ent (a good patient) and respectful of
and clinician perspectives76. These pro- IMPLEMENTATION the clinicians sapiential expertise and
vide standardized approaches to evaluate professional authority. When the domi-
complex interventions which integrate Although most clinicians believe that nant discourse is clinician-led, a primary
SDM with other established innovations. they are using the SDM approach, there flow of information from clinician to
Advanced directives and joint crisis is evidence to the contrary85. Perceptions patient means that the patients values
plans are examples of established innova- about level of involvement differ, with and treatment preferences are given less
tions77. Advance directives involve the patients identifying more clinician-led importance93. Overall, it is difficult to
patients pre-specifying their preferences and clinicians identifying more shared avoid clinician-led decision making being
for what should occur if they lose capacity approaches86. Patients report inhibiting the default choice in institution-based

mental health services, because SDM in- participants (patients), and in which part- CONCLUSION
volves a shift in power arrangements94. nership working (shown for example by
SDM) is the foundation rather than a fea- In this paper, the case has been made
ture to be added on. that SDM is part of a broader movement
TRANSFORMATION IN THE Participatory approaches to service de- of change in the mental health system110.
MENTAL HEALTH SYSTEMS velopment already exist in mental health There are implementation challenges, but
services. Peer support theories such as in- these are ethical and cultural as well as
The world is changing. Mental health tentional mutuality emphasize relation- technical.
systems internationally are transitioning ships in which both people have value It is worth addressing these complex
towards community-based services95-101, and reciprocity is possible105. Recovery issues relating to power, control, exper-
which involve interactions with patients Colleges are based on principles of col- tise and valued knowledge, because SDM
who are more influenced by citizenship laboration, co-production, inclusiveness has the potential to contribute to sup-
expectations relating to consumerism, self- and a community focus106. Similarly, a porting people to live as well as possible
determination and empowerment102. Pa- majority of participants in user-run pro- in communities of their own choosing.
tients increasingly expect as a right to be grammes value role equity, the mutuality
active participants in decisions about their and reciprocity of relationships and the
lives, with a greater emphasis on the bio- non-hierarchical organization107.
REFERENCES
medical ethical principles of autonomy Market segmentation is a well-estab-
1. Karnieli-Miller O, Eisikovits Z. Physician as
and justice. lished business technique used to identi-
partner or salesman? Shared decision-making
The implications of this shift for men- fy and manage diverse customer needs in real-time encounters. Soc Sci Med 2009;69:
tal health systems are profound, and ex- and to target marketing resources108. Po- 1-8.
2. Charles C, Gafni A, Whelan T. Decision-mak-
tend well beyond discussion of approaches sitioning similar groups of people into
ing in the physician-patient encounter: revis-
to decision making. Organizational trans- market segments, and then focusing mar- iting the shared treatment decision-making
keting efforts at these different segments model. Soc Sci Med 1999;49:651-61.
formation may be needed if the mental
3. Coulter A, Collins A. Making shared decision-
health systems are to survive this transi- as appropriate can manage heterogeneity
making a reality. No decision about me, with-
tion to engaging with patients holding cit- in preferences. By developing marketing out me. London: Kings Fund, 2011.
strategies and behaviour change strate- 4. Hamann J, Mendel R, Fink B et al. Patients
izenship expectations. A readiness to draw
and psychiatrists perceptions of clinical deci-
in insights and use language and constructs gies for distinct groups of patients who
sions during schizophrenia treatment. J Nerv
from other sectors will be needed to inform have specific needs or values, it becomes Ment Dis 2008;196:329-32.
possible to influence culture and create 5. Freidl M, Konrad J, Pesola F et al. Effects of
this transformation. This can be illustrated
clinical decision topic on patients involve-
by two examples, both of which are poten- demand for SDM in clinicians working
ment in and satisfaction with decisions and
tially relevant but currently almost unused with, and patients coming from, different their subsequent implementation. Psychiatr
clinical populations. Serv 2016;67:658-63.
in planning and developing mental health
6. Makoul G, Clayman ML. An integrative model
systems. The second example is given by the
of shared decision making in medical encoun-
The first example is given by the aca- expertise held by the hospitality industry ters. Patient Educ Couns 2006;60:301-12.
in working with disparate customers: 7. Legare F, Ratte S, Stacey D et al. Interventions
demic discipline of social marketing103,
for improving the adoption of shared decision
which could be used as an approach to Key values, such as the importance of making by healthcare professionals. Cochrane
fostering culture change in mental health welcome, the customer always being Database Syst Rev 2010;5:CD006732.
right and the job being to provide help to 8. Schattner A, Bronstein A, Jellin N. Informa-
systems. Social marketing involves the
tion and shared decision-making are top
application of marketing principles and meet the customers needs, underpin the patients priorities. BMC Health Serv Res
practices to advance social good, in this best interactions in this service industry. 2006;6:21.
Hospitality workers are skilled in recog- 9. Chewning B, Bylund CL, Shah B et al. Patient
case participation in decision making. It
preferences for shared decisions: a systematic
takes a citizen-centred approach in which nizing how customers like to be engaged review. Patient Educ Couns 2012;86:9-18.
insights developed with citizens and stake- with from face-to-face to elbow-to- 10. Health Foundation. Helping people share deci-
sion making. London: Health Foundation, 2012.
holders inform the process104. An orienta- elbow. Workers are not doing their job if
11. Anonymous. Taking shared decision making
tion towards mutuality, exchange and reci- customer care is poor109. If patients more seriously. Lancet 2011;377:784.
procity differentiates social marketing from achieve similar levels of emancipation and 12. Wakefield P, Read S, Firth W et al. Clients per-
agency as other citizens, then patient ceptions of outcome following contact with a
other social intervention approaches, par-
community mental health team. J Ment
ticularly in traditional expert-driven, top- choices and preferences become central. If Health 1998;7:375-84.
down public health approaches. So, social clinicians dont work in partnership with 13. Hamann J, Leucht S, Kissling W. Shared deci-
patients to ensure they have a positive sion making in psychiatry. Acta Psychiatr
marketing provides an approach to devel-
Scand 2003;107:403-9.
oping citizen-centred mental health sys- experience, then patients will and should 14. Crumlish N, Kell B. How psychiatrists think.
tems oriented around the preferences of choose to go elsewhere for support. Adv Psychiatr Treat 2009;15:72-9.

15. Hamann J, Langer B, Winkler V et al. Shared is associated with improvement in depression. psychotic prescribing, a qualitative study.
decision making for in-patients with schizo- Med Care 2006;44:398-405. BMC Psychiatry 2014;14:127.
phrenia. Acta Psychiatr Scand 2006;114:265-73. 32. Stacey G, Felton A, Morgan A et al. A critical 51. Fukui S, Matthias M, Salyers M. Core
16. Elwyn G. Salzburg statement on shared deci- narrative analysis of shared decision-making domains of shared decision-making during
sion making. BMJ 2011;342:d1745. in acute inpatient mental health care. J Inter- psychiatric visits: scientific and preference-
17. Del Piccolo L, Goss C. People-centred care: prof Care 2016;30:35-41. based discussions. Adm Policy Ment Health
new research needs and methods in doctor- 33. Pilgrim D. Recovery and current mental 2015;42:40-6.
patient communication. Challenges in men- health policy. Chronic Illn 2008;4:295-304. 52. Center for Mental Health Services. Shared
tal health. Epidemiol Psychiatr Sci 2012;21: 34. Quirk A, Chaplin R, Lelliott P et al. How pres- decision-making in mental health care: prac-
145-9. sure is applied in shared decisions about tice, research, and future directions. Rock-
18. National Institute for Health and Clinical antipsychotic medication: a conversation ville: Substance Abuse and Mental Health
Excellence. Service user experience in adult analytic study of psychiatric outpatient con- Services Administration, 2010.
mental health: improving the experience of sultations. Sociol Health Illn 2012;34:95-113. 53. Cornuz J, Kuenzi B, Krones T. Shared decision
care for people using adult NHS mental 35. Morant N, Kaminskiy E, Ramon S. Shared making development in Switzerland: room
health services. CG136. London: National decision making for psychiatric medication for improvement! Z Evid Fortbild Qual Gesund-
Institute for Health and Clinical Excellence, management: beyond the micro-social. Health hwes 2011;105:296-9.
2011. Expect 2016;19:1002-14. 54. Moumjid N, Br emond A, Mignotte H et al.
19. Joosten EAG, de Jong CAJ, de Weert-van Oene 36. Hamann J, Mendel R, Cohen R et al. Psychia- Shared decision-making in the physician-
GH et al. Shared decision-making reduces drug trists use of shared decision making in the patient encounter in France: a general over-
use and psychiatric severity in substance- treatment of schizophrenia: patient charac- view. Z Evid Fortbild Qual Gesundhwes 2007;
dependent patients. Psychother Psychosom teristics and decision topics. Psychiatr Serv 101:223-8.
2009;78:245-53. 2009;60:1107-12. 55. Perestelo-Perez L, Rivero-Santana A, Perez-
20. Joosten E, De Jong C, de Weert-van Oene G 37. Corrigan P. Lessons learned from unintended Ramos J et al. Shared decision making in
et al. Shared decision making: increases auton- consequences about erasing the stigma of Spain: current state and future perspectives.
omy in substance-dependent patients. Subst mental illness. World Psychiatry 2016;15:67- Z Evid Fortbild Qual Gesundhwes 2011;105:
Use Misuse 2011;46:1037-8. 73. 289-95.
21. Duncan E, Best C, Hagen S. Shared decision 38. Slade M, Amering M, Farkas M et al. Uses and 56. Holmes-Rovner M, Gruman J, Rovner D.
making interventions for people with mental abuses of recovery: implementing recovery- Shared decision-making in the US research
health conditions. Cochrane Database Syst oriented practices in mental health systems. & development outpaces delivery. Z Evid For-
Rev 2010;1:CD007297. World Psychiatry 2014;13:12-20. tbild Qual Gesundhwes 2007;101:254-8.
22. Hamann J, Cohen R, Leucht S et al. Shared 39. Daley S, Newton D, Slade M et al. Develop- 57. Puschner B, Steffen S, Slade M et al. Clinical
decision making and long-term outcome in ment of a framework for recovery in older decision making and outcome in routine care
schizophrenia treatment. J Clin Psychiatry people with mental disorder. Int J Geriatr Psy- for people with severe mental illness (CEDAR):
2007;68:992-7. chiatry 2013;28:522-9. study protocol. BMC Psychiatry 2010;10:90.
23. Loh A, Simon D, Wills CE et al. The effects of 40. Drennan G, Alred D (eds). Secure recovery. 58. Wild D, Grove A, Martin M et al. Principles of
a shared decision-making intervention in pri- Approaches to recovery in forensic mental good practice for the translation and cultural
mary care of depression: a cluster-randomized health settings. Oxon: Willan, 2012. adaptation process for patient-reported out-
controlled trial. Patient Educ Couns 2007;67: 41. Kidd S, McKenzie K, Virdee G. Mental health comes (PRO) measures: report of the ISPOR
324-32. reform at a systems level: widening the lens Task Force for Translation and Cultural Adap-
24. Woltmann EM, Wilkniss SM, Teachout A et al. on recovery-oriented care. Can J Psychiatry tation. Value Health 2005;8:94-104.
Trial of an electronic decision support system 2014;59:243-9. 59. Konrad J, Loos S, Neumann P et al. Content
to facilitate shared decision making in com- 42. Le Boutillier C, Chevalier A, Lawrence V et al. and implementation of clinical decisions in
munity mental health. Psychiatr Serv 2011;62: Staff understanding of recovery-orientated the routine care of people with severe mental
54-60. mental health practice: a systematic review and illness. J Ment Health 2015;24:15-9.
25. Joosten EA, DeFuentes-Merillas L, de Weert narrative synthesis. Implement Sci 2015;10:87. 60. Puschner B, Neumann P, Jordan H et al. De-
GH et al. Systematic review of the effects of 43. Le Boutillier C, Slade M, Lawrence V et al. velopment and psychometric properties of a
shared decision-making on patient satisfac- Competing priorities: staff perspectives on five-language multiperspective instrument to
tion, treatment adherence and health status. supporting recovery. Adm Policy Ment Health assess clinical decision making style in the
Psychother Psychosom 2008;77:219-26. 2015;42:429-38. treatment of people with severe mental ill-
26. Malm U, Ivarsson B, Allebeck P et al. Integrat- 44. Mental Health Recovery Study Working ness (CDMS). BMC Psychiatry 2013;13:48.
ed care in schizophrenia: a 2-year random- Group. Mental health recovery: users and 61. First MB, Spitzer RL, Gibbon M et al. Struc-
ized controlled study of two community- refusers. Toronto: Wellesley Institute, 2009. tured Clinical Interviews for DSM-IV Axis
based treatment programs. Acta Psychiatr 45. Harper D, Speed E. Uncovering recovery: the I Disorders Clinical Version (SCID-CV).
Scand 2003;107:415-23. resistible rise of recovery and resilience. Stud Washington: American Psychiatric Association,
27. Ludman E, Katon W, Bush T et al. Behavioural Soc Justice 2012;6:9-25. 1997.
factors associated with symptom outcomes 46. Watson DP. The evolving understanding of 62. Slade M, Cahill S, Kelsey W et al. Threshold 4:
in a primary care-based depression preven- recovery: what the sociology of mental health an evaluation of the Threshold Assessment
tion intervention trial. Psychol Med 2003;33: has to offer. Humanity Soc 2012;36:290-308. Grid as an aid to mental health referrals. Pri-
1061-70. 47. Care Quality Commission. 2015 community mary Care Ment Health 2003;1:45-54.
28. Hauser K, Koerfer A, Kuhr K et al. Outcome- mental health survey. Statistical release. Lon- 63. Puschner B, Becker T, Mayer B et al. Clinical
relevant effects of shared decision making. A don: Care Quality Commission, 2015. decision making and outcome in the routine
systematic review. Dtsch Arztebl Int 2015;112: 48. Care Quality Commission. CQCs response to care of people with severe mental illness
665-71. the 2015 community mental health survey. across Europe (CEDAR). Epidemiol Psychiatr
29. Thompson L, McCabe R. The effect of clini- London: Care Quality Commission, 2016. Sci 2016;25:69-79.
cian-patient alliance and communication on 49. De las Cuevas C, Pen ~ ate W, Perestelo-Pe
rez L 64. Clarke E, Puschner B, Jordan H et al. Empow-
treatment adherence in mental health care: a et al. Shared decision making in psychiatric erment and satisfaction in a multinational
systematic review. BMC Psychiatry 2012;12:87. practice and the primary care setting is study of routine clinical practice. Acta Psy-
30. Hamann J, Cohen R, Leucht S et al. Do unique, as measured using a 9-item Shared chiatr Scand 2015;131:369-78.
patients with schizophrenia wish to be Decision Making Questionnaire (SDM-Q-9). 65. Ho lzel L, Kriston L, Ha
rter M. Patient prefer-
involved in decisions about their medical Neuropsychiatr Dis Treat 2013;9:1045-52. ence for involvement, experienced involve-
treatment? Am J Psychiatry 2005;162:2382-4. 50. Shepherd A, Shorthouse O, Gask L. Consul- ment, decisional conflict, and satisfaction
31. Clever S, Ford D, Rubenstein L et al. Primary tant psychiatrists experiences of and atti- with physician: a structural equation model
care patients involvement in decision-making tudes towards shared decision making in anti- test. BMC Health Serv Res 2013;13:231.

66. Le Boutillier C, Leamy M, Bird VJ et al. What psychosis: a randomised controlled trial. Lan- health care in Latin American and Caribbean
does recovery mean in practice? A qualitative cet 2013;381:1334-41. countries. World Psychiatry 2012;11:191-5.
analysis of international recovery-oriented 81. Farrelly S, Lester H, Rose D et al. Barriers to 96. Ito H, Setoya Y, Suzuki Y. Lessons learned in
practice guidance. Psychiatr Serv 2011;62: shared decision making in mental health developing community mental health care in
1470-6. care: qualitative study of the Joint Crisis East and South East Asia. World Psychiatry
67. Slade M, Longden E. Empirical evidence Plan for psychosis. Health Expect 2015;19: 2012;11:186-90.
about mental health and recovery. BMC Psy- 448-58. 97. McGeorge P. Lessons learned in developing
chiatry 2015;15:285. 82. Thornicroft G, Slade M. New trends in assess- community mental health care in Australasia
68. Zipursky R, Agid O. Recovery, not progressive ing the outcomes of mental health interven- and the South Pacific. World Psychiatry 2012;
deterioration, should be the expectation in tions. World Psychiatry 2014;13:118-24. 11:129-32.
schizophrenia. World Psychiatry 2015;14:94-6. 83. Knaup C, Koesters M, Schoefer D et al. Effect 98. Drake RE, Latimer E. Lessons learned in
69. Davidson L, Bellamy C, Guy K et al. Peer sup- of feedback of treatment outcome in special- developing community mental health care in
port among persons with severe mental ill- ist mental healthcare: meta-analysis. Br J Psy- North America. World Psychiatry 2012;11:47-
nesses: a review of evidence and experience. chiatry 2009;195:15-22. 51.
World Psychiatry 2012;11:123-8. 84. Metz M, Franx G, Veerbeek M et al. Shared 99. Okasha A, Karam E, Okasha T. Mental health
70. Peters E, Ward T, Jackson M et al. Clinical, decision making in mental health care using services in the Arab world. World Psychiatry
socio-demographic and psychological char- routine outcome monitoring as a source of 2012;11:52-4.
acteristics in individuals with persistent psy- information: a cluster randomised controlled 100. Semrau M, Barley EA, Law A et al. Lessons
chotic experiences with and without a need trial. BMC Psychiatry 2015;15:313. learned in developing community mental
for care. World Psychiatry 2016;15:41-52. 85. Hamann J, Maris N, Iosifidou P et al. Effects health care in Europe. World Psychiatry 2011;
71. Deegan PE. A web application to support of a question prompt sheet on active patient 10:217-25.
recovery and shared decision making in psy- behaviour: a randomized controlled trial with 101. Hanlon C, Wondimagegn D, Alem A. Lessons
chiatric medication clinics. Psychiatr Rehabil depressed outpatients. Int J Soc Psychiatry learned in developing community mental
J 2010;34:23-8. 2014;60:227-35. health care in Africa. World Psychiatry 2010;9:
72. Feldman-Stewart D, Brennenstuhl S, McIssac 86. Seale C, Chaplin R, Lelliott P et al. Sharing 185-9.
K et al. A systematic review of information in decisions in consultations involving anti- 102. Leamy M, Bird V, Le Boutillier C et al. A concep-
decision aids. Health Expect 2006;10:46-61. psychotic medication: a qualitative study of tual framework for personal recovery in mental
73. Moher D, Schulz KF, Simera I et al. Guidance psychiatrists experiences. Soc Sci Med 2006; health: systematic review and narrative synthe-
for developers of health research reporting 62:2861-73. sis. Br J Psychiatry 2011;199:445-52.
guidelines. PLoS Med 2010;7:e1000217. 87. Eliacin J, Salyers M, Kukla M et al. Factors 103. French J, Gordon R. Strategic social market-
74. Bar Deucher A, Hengartner MP, Kawohl W influencing patients preferences and perceived ing. London: Sage, 2015.
et al. Participation in medical decision- involvement in shared decision-making in 104. Gordon O, Oades L. Social marketing of well-
making across Europe: an international longi- mental health care. J Ment Health 2015;24:24-8. being. In: Slade M, Oades L, Jarden A (eds).
tudinal multicenter study. Eur Psychiatry 2016; 88. Beauchamp T, Childress J. Principles of bio- Wellbeing, recovery and mental health. Cam-
35:39-46. medical ethics. Oxford: Oxford University bridge: Cambridge University Press 2017:311-
75. Coulter A, Jenkinson C. European patients Press, 2001. 23.
views on the responsiveness of health sys- 89. McGuire A, McCullough L, Weller S et al. 105. Mead S. Intentional peer support: an alterna-
tems and healthcare providers. Eur J Publ Missed expectations? Physicians views of tive approach. Plainfield: Shery Mead Con-
Health 2005;14:355-60. patients participation in medical decision- sulting, 2005.
76. Scholl I, Koelewijn-van Loon M, Sepucha K making. Med Care 2005;43:466-70. 106. McGregor J, Repper J, Brown H. The college
et al. Measurement of shared decision mak- 90. Dixon L, Holoshitz Y, Nossel I. Treatment is so different from anything I have done. A
ing a review of instruments. Z Evid Fortbild engagement of individuals experiencing mental study of the characteristics of Nottingham
Qual Gesundhwes 2011;105:313-24. illness: review and update. World Psychiatry Recovery College. J Ment Health Train Educ
77. Henderson C, Farrelly S, Moran P et al. Joint 2016;15:13-20. Pract 2014;9:3-15.
crisis planning in mental health care: the 91. Hamann J, Heres S. Adapting shared decision 107. Mahlke C, Kra mer U, Becker T et al. Peer sup-
challenge of implementation in randomized making for individuals with severe mental ill- port in mental health services. Curr Opin Psy-
trials and in routine care. World Psychiatry ness. Psychiatr Serv 2014;65:1483-6. chiatry 2014;27:276-81.
2015;14:281-3. 92. Goffman E. Asylums: essays on the social sit- 108. Dibb S, Simkin L. Implementation rules to
78. Campbell LA, Kisely SR. Advance treatment uation of mental patients and other inmates. bridge the theory/practice divide in market
directives for people with severe mental Harmondsworth: Penguin, 1968. segmentation. J Market Manag 2009;25:375-96.
illness. Cochrane Database Syst Rev 2009: 93. McCabe R, Heath C, Burns T et al. Engage- 109. Slade M. Personal recovery and mental illness.
CD005963. ment of patients with psychosis in the con- Cambridge: Cambridge University Press, 2009.
79. Henderson C, Flood C, Leese M et al. Effect of sultation. BMJ 2002;325:1148-51. 110. Galderisi S, Heinz A, Kastrup M et al. Toward
joint crisis plans on use of compulsory treat- 94. Kaminskiy E. The elephant in the room: a a new definition of mental health. World
ment in psychiatry: single blind randomised theoretical examination of power for shared Psychiatry 2015;14:231-3.
controlled trial. BMJ 2004;329:136-40. decision making in psychiatric medication
80. Thornicroft G, Farrelly S, Szmukler G et al. management. Intersectionalities 2015;4:19-38.
Clinical outcomes of Joint Crisis Plans to 95. Razzouk D, Gregorio G, Antunes R et al. Les-
reduce compulsory treatment for people with sons learned in developing community mental DOI:10.1002/wps.20412

COMMENTARIES
Shared decision making: a consideration of historical and political

contexts
M. Slade1 provides a thought-provok- is the CanMEDS framework. Here medi- medications or other physical treatments
ing and helpful review of ways in which cal training at all levels is seen as leading that are the special purview of doctors
shared decision making (SDM) may be towards development of connected com- are not particularly more powerful than
understood, its justifications generally petencies: that of medical expert is de- psychological or social interventions, yet
and the application of SDM principles scribed as an integrating role for the work delivery systems and power dynamics can
in mental health. He suggests adoption of a doctor as communicator, collabora- serve to support prioritization of those
of cross-sectoral practices, such as from tor, leader, health advocate and scholar7, interventions over others.
marketing and hospitality, in promoting which can be seen as equipping practi- A set of issues related to allocative and
SDM. This commentary takes a more tioners potentially to work in alignment distributional justice bring into the frame
medically oriented perspective and fur- with the chronic disease model, evidence- more formal kinds of politics. The breadth
ther explores some political implications based medicine or SDM. of existing services influences the options
raised by the challenges of SDM. Work cited by Slade demonstrates that available to service providers and so con-
SDM in mental health care may be people appreciate being invited to take ditions what may be considered within
formally a relatively recent arrival, but more autonomy than they may directly the SDM reasoning space. For instance, in
SDM does not exist in historical isolation ask for. Often the health care consulta- Australia at least, there are dramatic differ-
either within medicine or mental health tion starts from a point in which the ences in what is available in terms of psy-
care. Through the history of medicine, patient is situationally disadvantaged by chological treatments in different parts of
shifts in the power balance have occurred power dynamics, cultural expectations, the country10, so what is a primary care
in decision making between physician and the potentially disabling nature of physician for instance to do where psycho-
and patient2, and recent decades may interacting with an expert8. So, patients
logical treatments that could substantially
have seen some shifting of this locus of or clients appreciate efforts to promote
improve the lot of someone with a mental
control towards the patient. their empowerment even if they dont ask
health problem are not actually available?
Related concepts within modern medi- for it. In mental health care, there are also
Is it within the scope of SDM to advise
cine and medical sociology can be seen other important convergences: firstly with
patients with a current major depressive
as dating back at least half a century. the challenge of practice in the context of
the application or availability of compul- disorder that, if they lived in a more
Szasz and Hollander3 and Balint4, in the
sory treatment; then with the politics of affluent area or personally were better
mid-20th century, advocated for a shift
the recovery movement and with working financially resourced than they are, then
in management of chronic diseases from
towards recovery oriented practice. they should augment their antidepressants
activity-passivity and guidance-coopera-
tion models towards a model of mutual While there are multiple convergences with some well conducted cognitive be-
participation in doctor-patient relation- across the medical and mental health haviour therapy but, because of their loca-
ships. This approach shares features with care literature with the key propositions tion and personal circumstance, this isnt
what may now be termed SDM. of SDM, implementation often still faces an available option? Often this discussion
The chronic disease model elaborated great challenges. Promoting SDM can will not occur and the reasoning space of
by Wagner and others in the US5 empha- challenge clinicians in terms of the depth SDM will be constrained by what is practi-
sizes an activated and informed patient and breadth of their professional skills cally available, even where that does not
as a critical component. Sharing of de- and invite commitment through other conform to evidence-based practice or to
cisions may also be seen as critical to roles as citizens that may intertwine with
principles of justice. Here we encounter a
evidence-based medicine best practice, those of professional practice.
connection between SDM and what in the
which involves integrating individual Slade helpfully makes the point that
CanMEDS model7 is described broadly as
clinical expertise with the best available transitions towards SDM require changes
in prevailing service culture. Here we enter the health advocacy role and in practice
external clinical evidence. . . reflected in
many ways, but especially in more effec- a realm that in a broad sense is political, in may involve formal political engagement.
tive and efficient diagnosis and in the that it involves processes whereby power Implementation of SDM, as Slade sug-
more thoughtful identification and com- is assigned. For instance, by virtue of in- gests, may benefit from ideas from mar-
passionate use of individual patients come, specific legal powers and societal keting and hospitality. These can be seen
predicaments, rights, and preferences in status, not uncommonly psychiatrists can as just two examples among possible strat-
making clinical decisions about their assert influence in many service systems egies that stretch across all the CanMEDS
care6. that is not necessarily commensurate with competencies and can present clinicians
An influential model in medical train- the potency of the interventions they with political challenges. Even if we suc-
ing and practice, including in psychiatry, control9. For many psychiatric conditions, ceed in addressing many of the personal

politics that allow our practice to move services systems, and with the challenge 5. Bodenheimer T, Wagner EH, Grumbach K.
JAMA 2002;288:1775-9.
towards more effective use of the range of perhaps engaging in formal political 6. Sackett DL, Rosenberg WMC, Gray JAM et al.
of enablers to SDM that Slade has identi- activism where there are critical con- BMJ 1996;312:71-2.
fied, other critical factors may block pro- straints to implementation of SDM that 7. Royal College of Physicians and Surgeons of
Canada. CanMEDS 2015 physician competency
gress. Key impediments to SDM may relate to political policies on social justice. framework. Ottawa: Royal College of Physicians
arise because informal politics in insti- and Surgeons of Canada, 2015.
Graham Meadows 8. Illich I. Limits to medicine. Harmondsworth:
tutions constrain exploration of what is Department of Psychiatry, Monash University, and Pelican, 1977.
possible and because government formal University of Melbourne School of Population and 9. Frank RG, Glied SA. Better but not well: mental
Global Health, Melbourne, VIC, Australia
politics constrains the range of possible health policy in the United States since 1950.
Baltimore: Johns Hopkins University Press, 2006.
decisions in ways that may be palpably 1. Slade M. World Psychiatry 2017;16:146-53. 10. Meadows GN, Enticott JC, Inder B et al. Med J
unjust. 2. Kaba R, Sooriakumaran P. Int J Surg 2007;5:57- Aust 2015;202:190-4.
65.
Full engagement with SDM may chal-
3. Szasz T, Hollander MHA. Arch Intern Med
lenge practitioners with the prospect of DOI:10.1002/wps.20413
1956;97:585-92.
participation in processes that may ra- 4. Balint M. The doctor, his patient, and the illness.
London: Pitman, 1957.
dically shift existing power balances in
Involvement in decision making: the devil is in the detail

M. Slade1 highlights that there is super- increase their ability to participate in de- agree with. At all times, clinicians need to
ficial, not deep agreement between stake- cision making2. gauge the expectations, needs and pref-
holders about the idea of shared decision Secondly, different types of decisions erences of both parties in a fine balanc-
making (SDM) in mental health care. To are taken at specific junctures in an illness ing act.
begin to unpack why this may be so, it is trajectory, e.g. starting a psychotropic med- Understanding the extent to which
important to reflect on decision making ication is a much bigger decision than SDM is implemented is intricately linked
about what, when and with whom. changing the dose of an existing medica- to how it is measured. As Slade points
Firstly, what kinds of decisions are tion. Mental health problems tend to be out, decision making is a complex and
made in mental health care? They range episodic, so that the ability to process in- dynamic social interaction1. Most re-
from life-changing decisions, sometimes formation and motivation to participate search to date, with some notable excep-
when the person is deemed to no longer varies over time. For example, the presence tions4-7, is based on what people say
have capacity or to have reduced capacity of negative symptoms has been found to about SDM. Perhaps the most informa-
to make decisions (e.g., involuntary ad- be associated with less involvement in tive means of researching decision mak-
mission to hospital, medical treatment in decision making3. ing is to record and analyse what people
dementia, starting psychotropic medica- Thirdly, who is involved in decision do rather than what they say they do.
tion) to more routine decisions relating to, making in mental health care? In addi- This approach facilitates an understanding
for example, changing medication, ad- tion to the patient and clinician, carers of the dilemmas faced by both clinicians
dressing physical health, consent to share are frequently involved. Decision making and patients in situ and the resources they
information, and referral to other services in three-way communication, i.e., patient- deploy to deal with them. It offers a win-
(e.g., drug and alcohol, day opportunities). doctor-carer, is undoubtedly more nuanced dow on how clinicians and patients jointly
Some decisions require explicitly over- and delicate than two-way communica- construct the clinical encounter8.
riding the patients preferences in his/her tion. With three people, there is the po- In an observational study of decision
best interest. The danger is that this prac- tential for two people to become aligned making in outpatient clinics in the UK, in-
tice leaks into other decisions, due to a in support of a particular course of action. volving people with a diagnosis of schizo-
cognitive bias that people with mental This could be the patient and carer (e.g., phrenia or depression, there was striking
health problems are perceived to be less in a bid to reduce or stop medication), or variation in the extent to which different
worthy or capable of being involved in the doctor and carer (e.g., in a bid to in- psychiatrists involved patients in decision
decisions when capacity is intact. Hence, crease medication or admit the patient), making across their consultations. Out of
it is important to question and safeguard or the doctor and patient (e.g., in a bid to a total possible score of 48 using the
our practice against these assumptions. keep the patient out of hospital). Some- Observing Patient Involvement (OPTION)
Specific interventions may be required times, patients feel that carers are acting scale, scores ranged from 0 to 38. The
when decisional capacity is reduced, e.g., as advocates. At other times, they feel that differences in how psychiatrists commu-
preparing patients in an acute ward for carers are working against them with clini- nicated were overwhelmingly explained
planning talks with their psychiatrist to cians to make decisions that they do not more by their individual style than by

socio-demographic, structural or clinical have some questions? is more likely to in identifying what needs to be addressed.
factors3. elicit further discussion. Asking about medi- This step is likely to circumvent some of
This shows that there is a widely vary- cation with questions such as No prob- the difficulties that currently undermine
ing practice but also some good practice, lems with the medication? invites the SDM in mental health care.
which can be identified and disseminat- patient to confirm that there are no prob-
Rose McCabe
ed. It would be interesting to explore what lems, making it very difficult for the pa- College House, University of Exeter Medical School, St.
attitudes are associated with communica- tient to discuss concerns he/she may have Lukes Campus, Exeter, UK
tion practices that involve patients more and influence subsequent treatment pro-
1. Slade M. World Psychiatry 2017;16:146-53.
in decision making. posals. Clinicians need to be aware of how
2. Hamann J, Langer B, Winkler V et al. Acta
Decision aids are helpful in drawing question design shapes patient responses, Psychiatr Scand 2006;114:265-73.
attention to and focusing discussion on in order to involve patients in a meaningful 3. McCabe R, Khanom H, Bailey P et al. Patient
various aspects that need to be consid- rather than a superficial way. Educ Couns 2013;91:326-8.
4. Quirk A, Chaplin R, Lelliott P et al. Sociol
ered in making a decision. However, they Training clinicians so that they are
Health Illn 2012;34:95-113.
should not detract from the clinician- aware of the subtle differences in how 5. Seale C, Chaplin R, Lelliott P et al. Soc Sci Med
patient interaction, as how decision aids they communicate with patients gener- 2007;65:698-711.
are actually used in interactions is impor- ally, and in decision making specifically, 6. Goossensen A, Zijlstra P, Koopmanschap M.
Patient Educ Couns 2007;67:50-6.
tant in determining whether they are was found to improve clinician-patient
7. Goss C, Moretti F, Mazzi MA et al. Br J Psychia-
effective. For example, subtle differences communication and the therapeutic re- try 2008;193:416-21.
in how clinicians ask questions have con- lationship10. Eliciting the patients expe- 8. Heritage J, Maynard DW. Ann Rev Soc 2006;32:
351-74.
sequences for what patients say9. This is riences and listening are fundamental:
9. Heritage J, Robinson JD, Elliott MN et al. J Gen
critical for decision making. For example, they are the starting point for identifying Intern Med 2007;22:1429-33.
asking a patient if he/she has questions what decisions are to be made and wheth- 10. McCabe R, John P, Dooley J et al. Br J Psychia-
try 2016;209:517-24.
with the commonly deployed Any ques- er these reflect the patients concerns.
tions? is designed not to elicit any fur- Working with patients to reach a shared DOI:10.1002/wps.20414
ther information, whereas asking Do you understanding of concerns is the first step
Psychiatric practice: caring for patients, collaborating with partners,

or marketing to consumers?
Slades paper1 usefully articulates the data to support this view. One potentially Indeed, a potentially important issue is
clinical and ethical arguments in support important consideration is that, in psychi- that, in medical and psychiatric practice,
of shared decision making (SDM); empha- atry, the link between scientific knowledge disorders range from more typical con-
sizes that, despite widespread superficial and patient outcomes is not always as ditions (where, say, the disorder can be
agreement that SDM is important, there tight as we would ideally like; so that clini- conceptualized as caused by an external
is significant potential for contradiction cal decision making, whether shared or agent that both clinician and patient are
between these arguments; and puts for- not, is unable to predict fully which indi- committed to eradicating) to more atypi-
ward a number of approaches which may vidual will respond to which intervention. cal ones (where, for example, it is harder
inform responses to the shift in the mental Furthermore, the data included in the to differentiate the self from the illness,
health system brought about by more Cochrane review of the clinical value of which may itself impact negatively on
empowered patients. SDM for people with psychiatric disor- decision making)3.
Here I briefly comment on each of ders2 are from work done in Germany. It It is also noteworthy that the desire
these components of Slades review. While is theoretically possible that in other set- to participate in SDM appears higher in
I admire his elegant synthesis of the litera- tings, where patients may have different some patient groups. Good ethical argu-
ture and concur with the thrust of many of expectations of the clinical encounter, ments can be put forward to support dif-
his arguments, I do also wish to express the data may be even less supportive of ferent sorts of decision-making models
some cautions. the clinical argument for SDM. for different sorts of patients and differ-
First, Slade notes that while there is a Second, Slade notes that, despite the ent sorts of disorders.
clinical argument that patients who are ethical argument that SDM is a human Third, Slade emphasizes that, although
active decision makers will be more engag- right, clinical practice occurs in a range SDM is widely endorsed in official policies
ed, with consequent improved decision of different contexts, and it is less clear and by active clinicians, in theory it entails
making, increased adherence, and supe- that SDM is the best approach to deci- potential contradictions (for example, there
rior outcomes there are in fact limited sion making in non-capacitous patients. is a potential for empowered patients to

choose to be less adherent to treatment potentially useful approaches for address- one that is deserving of particular empha-
recommendations), and that in practice ing the shift in the mental health system sis and pride.
SDM is in fact often not implemented. that will be brought about by more em-
Dan J. Stein
Slade argues that data from the Clin- powered patients. Metaphors play a cru- Department of Psychiatry and Mental Health, University
ical decision making and outcome in cial role in framing our views of physical of Cape Town and Medical Research Council Unit on
routine care for people with severe men- and mental disorders, and of the clinician- Anxiety and Stress Disorders, Cape Town, South Africa
tal illness (CEDAR) study4 show that both patient relationship5. A collaborative model
patient outcomes and experiences are im- of the clinician-patient relationship has 2. Duncan E, Best C, Hagen S. Cochrane Database
proved by SDM, so that there is an align- been particularly useful in cognitive behav- Syst Rev 2010;1:CD007297.
3. Stein DJ. Can J Psychiatry 2013;58:656-62.
ment between clinical and ethical jus- ioural therapy, for example. And viewing
4. Puschner B, Becker T, Mayer B et al. Epidemiol
tifications. However, although the CEDAR the patient as a consumer does have some Psychiatr Sci 2016;25:69-79.
study is multi-national, it is based on a rel- advantages, perhaps particularly in the 5. Stein DJ. Philosophy of psychopharmacology.
Cambridge: Cambridge University Press, 2008.
atively restricted population (outpatients context of empowerment or activism6.
6. Stein DJ, Phillips KA. BMC Med 2013;11:133.
in Europe), and the statistical analysis cited However, we should be careful not to
by Slade does not focus on clinical symp- entirely jettison metaphors of the doctor-
toms (but rather on patient-rated unmet patient relationship that emphasize car- DOI:10.1002/wps.20415
needs). ing (rather than only collaborating or con-
Finally, Slade suggests that social mar- suming). Caring is a core aspect of the
keting and the hospital industry provide work of mental health professionals, and
Common sense alone is not enough

Slades paper1 suggests implementing happen. Decisions about taking medi- nicians. Preferences differ associated with
shared decision making (SDM) in mental cation or attending a group therapy need their personality, background and experi-
health care. This sounds desirable. SDM is to be implemented by the patient, and, if ences, and may vary even for the same
characterized by collaboration, and who the patient is not happy with the deci- patient depending on the given health
would disagree with a call for more col- sion, the treatment is unlikely to materi- problem, the context, the specific con-
laboration between psychiatrists and pa- alize. Reaching an explicit agreement with tent of the consultation, and the mood on
tients? A more detailed look, however, patients about any decision is, therefore, the day.
raises at least two major concerns. a matter of both professional obligation Slade cites a patient survey in the UK
Firstly, what exactly is SDM? Accord- and clinical necessity. No new concept of National Health Service, seemingly sug-
ing to Slade, SDM is an intermediate po- SDM is needed to reflect this. gesting that patients in mental health
sition between two extremes, one in which So, is there anything more specific? The care in the UK are not involved in treat-
the clinician decides, having consulted referenced review of conceptual models of ment decisions although they would like
the patient, and one in which the patient SDM identifies several types of clinician to be3. A closer look at the data shows,
decides, having received information from behaviour that characterize SDM. They however, that the question Were you in-
the clinician. SDM is in between those include: explaining the problem and op- volved as much as you wanted to be in
two and involves collaboration. Consult- tions; discussing benefits and risks, pa- agreeing what care you will receive? was
ing a patient and providing information, tient preferences and abilities; presenting answered with no by only 7% of pa-
however, also require collaboration, possi- evidence; making recommendations; clar- tients. Although it should be a challenge
bly with extensive questions, explanations ifying patients understanding; and mak- to reduce this figure even further, it is
and clarifications. So, the unique char- ing decisions explicit. All this has been hardly a reason to call for a radical change
acteristic of what would need to be im- part of good clinical practice for decades in the current approach.
plemented as SDM becomes unclear. and may be regarded as common sense. Furthermore, it should be noted that
Unless the decision is about coercive It remains unclear how a basic under- this survey was conducted in patients
treatment, clinicians are not entitled to standing of good communication bene- with severe mental illness in community
make decisions that patients do not agree fits from being relabelled as SDM2. mental health care in the UK, who can-
to. Obtaining informed consent to any Secondly, there is a claim that pa- not easily go to a different clinician in
treatment is a professional duty. Patients tients want SDM. Patients are not one case they are unhappy with their current
must explicitly agree to whatever is being homogenous block of people who would one. In other systems and other patient
decided in a consultation. Beyond this all want the same thing and all of the groups, e.g. when patients pay for their
formal requirement, patients agreement time. Patients have different preferences services directly or through insurance,
is needed anyway for making a treatment for the communication style of their cli- SDM is presumably even less of a prob-

lem, as patients can simply go to a dif- can learn and successfully apply new skills Terms like recovery, SDM and co-pro-
ferent psychiatrist, if they do not like the in addressing psychotic symptoms of their duction give the illusion of novelty, of
communication style or the treatment patients8. Such research still faces a range new ideas and new insights, when in
suggestions of their psychiatrists. of significant conceptual and methodolog- fact there have been none. I wish such
Do these concerns suggest that there ical challenges, and the results are hardly terms would be recognized as what they
is no problem in how clinicians and pa- conclusive. Yet, the findings may be seen are, a combination of common sense,
tients communicate in mental health as encouraging to pursue and advance simplification and fashionable jargon,
care? Far from it! On the contrary, I would both rigorous research in patient-clinician without much new substance that would
argue that the communication between communication and teaching of relevant help mental health care move forward.
patients and clinicians is at the heart of skills. They resemble the the emperors new
mental health care, and that improving The call for SDM may reflect a wider clothes in Andersens tale9. Facing a lack
this communication is the single most im- and fundamental problem of current psy- of novelty may be uncomfortable, but it
portant and promising route to more chiatry. The last decades have seen the is likely to be a necessary step towards
effective treatment. For achieving this, rise of appealing terms that arose in the real innovation.
however, general and vague terms such public arena and with lay audiences,
Stefan Priebe
as SDM are not very helpful. Research on where the absence of a precise definition Unit for Social and Community Psychiatry, WHO Collabo-
patient-clinician communication requires is often an advantage. Recent examples rating Centre for Mental Health Services Development,
precise theories, specified models and include SDM, but also recovery and co- Queen Mary University of London, London, UK
detailed analyses. production. Such terms are intuitively
Examples from our own group to il- appealing to various stakeholders, who 2. Priebe S, Dimic S, Wildgrube C et al. Eur Psy-
lustrate such research on different aspects are free to understand them in any way chiatry 2011;26:403-7.
3. Care Quality Commission. 2015 community
of patient-clinician communication in- they like. These terms may have their
mental health survey. Statistical release. Lon-
clude: in-depth analyses of the difficulty of value in a political debate, but less so in don: Care Quality Commission, 2015.
psychiatrists to address patients concerns a professional discourse. 4. McCabe R, Heath C, Burns T et al. BMJ 2002;
325:1148-51.
about their delusions4,5; a non-clinical One might argue that, for psychiatry
5. Zangrilli A, Ducci G, Bandinelli PL et al. BMC
experiment about how psychiatrists should as a scientific discipline, these terms Psychiatry 2014;14:178.
introduce themselves in the first encoun- are even harmful. All scientific progress 6. Priebe S, Palumbo C, Ahmed S et al. Br J Psy-
chiatry 2013;202:459-62.
ter6; and randomized controlled trials requires intellectual honesty as the start-
7. Priebe S, Kelley L, Omer S et al. Psychother
on how to improve communication and, ing point. One needs an unbiased account Psychosom 2015;84:304-3.
through that, treatment outcomes. of what a discipline has already discovered 8. McCabe R, John P, Dooley J et al. Br J Psychia-
try 2016;209:517-24.
One trial showed that a new interven- and achieved, but also of what the limita-
9. Anderson HC. The emperors new clothes.
tion structuring and focusing the routine tions are. If there has been limited pro- Copenhagen: Reitzel, 1837.
communication between patients and cli- gress in treatment concepts in psychiatry
nicians in community mental health care over the last three decades, then this is no DOI:10.1002/wps.20416
(DIALOG1) leads to substantially more disgrace. It has to be considered appropri-
favourable clinical and social outcomes7. ately, so that as a discipline we can learn
Another trial demonstrated that clinicians from failure and hopefully move on.
Shared decision making in mental health care settings: perspective,

purpose and practice
Rates of chronic illness are growing informed preferences and values through this commentary, (re)examining theoret-
rapidly, as are health care costs. These shared decision making1. In his review ical perspectives on SDM and proposing
phenomena and the burdens that they on shared decision making (SDM) in men- a fidelity framework to support the prac-
present demand not only biomedical and tal health care, M. Slade2 provides an ex- tice of SDM at the service level. Without
care-delivery advancements but also inno- cellent examination of existing research the commitment of practitioners and the
vations in patient engagement, defined as on the topic and offers innovative recom- transformation of the entire workplace3,
the process of engaging patients and their mendations such as involving social SDM is rhetoric, not a reality.
caregivers in effective self-care, behavior marketing and the hospitality industry Examining the theoretical perspective
change, and chronic disease management; to move the field forward. While I agree that underpins SDM is not purposeless. It
and [addressing] the need to better align with what Slade has written about SDM helps us better understand the essence and
treatment choices with patients well- tools, I would like to go one step back in values behind the practices. The closure of

asylums and psychiatric institutions across tions. It is of paramount importance that teams to provide for the ongoing and reg-
the globe beginning in the 1970s, along health knowledge construction moves ular supervision of practitioners. The five
with the introduction of community care, from a process that is hierarchical to one essential steps of SDM stipulated in the
brought about a paradigm shift in mental that is horizontal, or defined by consen- SHARE approach form part of the pro-
health care, moving the field from one sus3,8. Such a shift would allow all parties, posed fidelity framework9. These steps
that was traditional and professionally- if willing, to participate in the construc- are: a) to seek the participation of service
led to one that is service-centered and in tion of knowledge that forms the basis for users in the SDM process; b) to help ser-
which patients participate (patient-led care decision-making in the search for per- vice users explore and compare interven-
remains rare in practice). The authors of a sonal recovery from mental illness. The tion options based on their recovery goals;
seminal work explain that at present, proposed fidelity framework consists of c) to assess service users values and pref-
there are broadly three strands to the pro- the following elements9,10: erences; d) to reach decisions jointly with
ject of critical psychiatry: the development External environment. Health care serv- service users; and e) to evaluate the out-
of a critique of the influence of the phar- ices would deliver systematic and specific comes of SDM. Services would provide
maceutical industry on the theory and programs to promote social inclusion and physical settings that assist service users
practice of psychiatry, the establishment of equality and to reduce the stigma and dis- in participating in the SDM process, with
a medical discourse about mental suffering crimination associated with mental illness
features such as SDM corners and gadgets
that is sensitive to the issue of meaning, (e.g., targeting people recovering from psy-
(e.g., tablets, laptops). Moreover, senior
and the promotion of a partnership with chosis whose employment is at risk). These
management would invest in building up
the emerging user/survivor movement4. programs would form a solid base from
reference resources (e.g., evaluation instru-
In other words, the traditional medical which to implement SDM.
ments, SDM aids) and evaluating SDM
paradigm has been challenged, and an Leadership commitment. Organizations
implementation.
alternative discourse has been offered, would adopt a recovery approach in their
The current body of literature on
one that acknowledges the existence of overarching philosophies and put that con-
SDM has three major limitations. First,
both professional knowledge and lived- cept into practice (e.g., interactions with
although the outcome measures used by
experience knowledge that lay the often service users would be directed toward
the majority of trial studies are fairly
forgotten foundation for SDM. nurturing the service users autonomy
comprehensive in terms of covering
The sociological analysis of medical and choices). The position adopted in
clinical status, service satisfaction and
versus lived-experience knowledge dates SDM is that both professional and lived-
intervention adherence, no studies have
back to the early 1970s, with Freidsons experience knowledge are valid and that
assessed the innovation and creativity of
landmark publication5. Knowledge is cre- knowledge is most powerful when pro-
service users in problem solving, service
ated by people working individually or fessional and service user collaborate on
users decision making, or service users
interdependently, and it is often bounded terms of mutual understanding, respect
insight into and understanding about
by what society believes to be legitimate3,6. and equality. There should be clear pub-
their conditions. Second, we should not
Knowledge is never a neutral or objective licity in the form of brochures, posters
undermine the increasingly sophisti-
phenomenon but a matter of positional- and web resources about using SDM as a
cated capability of smartphones in the
ity, that is, of the place from which one tool to support service users in achieving
era of Web 3.011. It provides new options
speaks, to whom and for what purpose7. their recovery goals. Senior personnel of
and groundbreaking social media inter-
Knowledge construction (e.g., recovery- a given organization would join forces
faces that could support the applica-
oriented use of medication, illness man- with everyday practitioners to ensure that
tion of SDM. Third, there is a void in
agement strategies) and communication SDM be implemented in a multidisciplin-
the existing literature on the application
the essence of SDM are human activi- ary manner across different clinical serv-
of SDM in non-Western, non-Nordic ju-
ties and therefore subject to human vices ices (from acute inpatient facilities to
risdictions.
and virtues. There is a difference in pow- community-based services) throughout the
SDM is largely a psycholinguistic pro-
er between professional and service user entire organization. Similarly, the prac-
cess. The use of words, metaphors and
in an SDM session, and that difference is tice of SDM would be incorporated into
only exaggerated by time pressure and by the day-to-day operation of organizations. non-verbal communication; the art of
the meaning and suffering associated Finally, a hallmark of leadership commit- involving families and caregivers in the
with a mental health condition. ment to SDM would be the employment intervention process; and the level of ser-
Combining professional and lived- of people with personal experiences of vice users participation in mental health
experience knowledge in the search for mental illness in positions of senior man- services more broadly are crucial factors
personal recovery is not always a straight- agement, in committees monitoring SDM that need to be considered when it comes
forward process. The two forms of knowl- execution, and as peer support specialists to delivering SDM in mental health care.
edge sometimes work in a complementary to assist SDM programs.
Samson Tse
manner, but at other times their interac- Implementation of SDM. Organizations Department of Social Work and Social Administration,
tion causes more tension and raises ques- would establish SDM implementation University of Hong Kong, Hong Kong

5. Freidson E. Profession of medicine: a study of 9. Agency for Healthcare Research and Quality.
1. Fisher ES, Shortell SM, Savitz LA. JAMA 2016; the sociology of applied knowledge. Chicago: The SHARE approach. Rockville: Agency for
315:339-40. University of Chicago Press, 1988. Healthcare Research and Quality, 2015.
2. Slade M. World Psychiatry 2017;16:146-53. 6. Shaw I, Kauppinen K. Constructions of health 10. Centre for Mental Health Research and Inno-
3. Crepaz-Keay D, Fulford K, van Staden W. In: and illness. Hants: Ashgate, 2004. vation. Strengths model fidelity scale: instruc-
Sadler JZ, Van Staden W, Fulford K (eds). 7. Baker C. Cultural studies: theory and practice. tions for reviewers. Kansas: University of Kansas,
Oxford handbook of psychiatric ethics, Vol. London: Sage, 2008. School of Social Welfare, 2014.
1. New York: Oxford University Press, 2015:60- 8. Jordan B. In: Davis-Floyd RE, Sargent CF (eds). 11. Korsbek L, Tnder ES. Psychiatr Rehab J 2016;
87. Childbirth and authoritative knowledge: cross- 39:167-72.
4. Bracken P, Thomas P. Philos Psychiatry Psy- cultural perspectives. Berkeley: University of
chol 2010;17:219-28. California Press, 1997:55-79. DOI:10.1002/wps.20417
Incorporating shared decision making in mental health care requires

translating knowledge from implementation science
M. Slade1 provides a broad overview essary communication skills can appropri- to an individuals decision making capac-
of the literature on shared decision making ately use a decision aid during the consul- ity. While individuals with mental illness
(SDM) with a focus on mental health care. tation. The use of decision aids can indeed may have impaired cognitive abilities,
The overview is timely and pertinent, as promote the engagement of patients in the most desire and have the capacity to be
SDM is considered a central component decision making process, but there are involved in treatment decision making,
of the widely accepted recovery model of also other ways of fostering SDM, includ- including those with severe conditions
mental health services2. We are encour- ing patient-mediated interventions that such as schizophrenia and major depres-
aged by Slades focus on implementation, prompt patients to ask questions4. sion6. Similar to patients with other cog-
which is the current challenge facing SDM We agree with Slades second challenge nitive disabilities, strategies are available
practice across all settings and countries. that SDM implementation endeavors could to increase participation in decision mak-
Slade highlights significant challenges potentially be more successful if better ing among individuals with severe mental
to decision aid uptake, including quality integrated into other innovations in men- illness, such as the use of multiple display
control and the overwhelming number of tal health care. This argument is especially formats when communicating treatment
those aids. The movement toward quality compelling from a clinicians perspective. options and risks.
control of decision aids is over ten years By branding SDM as the most important Of course, these individuals are not
old. The International Patient Decision singular new intervention that clinicians always capable of becoming involved in a
Aids Standards Collaboration (http://ipdas. must adhere to in their portfolio of skills decision making process; this ability may
ohri.ca) has provided criteria to judge the and interventions, we undermine its po- vary over the course of their illness. In
quality of patient decision aids. Certifica- tential and may cause resistance. More such cases, joint crisis plans may be use-
tion is also underway and has the poten- work is needed to integrate SDM with ful. For example, when a patients decision
tial to improve the quality of the growing other health care innovations in particu- making capacity is reduced, a clinician or
number of those aids. lar fields of health care. Thus, the mental family member can draw on the patients
However, we agree that the current health field has the potential to take the stated preferences that were gathered
model of decision aid development and lead, for example, through the integra- when the patient was capable of making
maintenance is unsustainable. The use tion of SDM and advance directives and a decision. Such plans could be beneficial
of technology is being harnessed to ad- joint crisis plans5. in institutional settings where patient au-
dress this challenge. For example, the Slade highlights the important ethical tonomy is even more restricted.
SHARing Evidence to Inform Treatment tension between beneficence and patient Nevertheless, research has shown that
decisions (SHARE-IT) project is an ini- autonomy to make decisions. An over- most people diagnosed with a mental ill-
tiative designed to automate decision aid emphasis on beneficence-focused treat- ness have a similar level of decision making
production based on guideline updates3. ment at the expense of patient autonomy capacity as a healthy comparison group
While decision aids are useful adjuncts can result in treatment decisions that re- from the general population6. Increased
to SDM, it is important to clarify that present the clinicians values imposed on awareness of this ability would be an
the practice of SDM does not require a the patient. This is particularly concern- important step toward increasing patient
decision aid. Informing patients of their ing in mental health care, where the effec- engagement in SDM.
options, eliciting their preferences and tiveness of treatments is often overstated, This appeal for reducing the stigma
integrating these patient preferences into despite only modest gains and significant towards mentally ill patients by not deny-
the health care decision is a practice that potential side effects. ing them their decision making capacities
requires communication skills, not just As Slade indicates, the question most is related to the prominent and broader
tools. Only a clinician who has the nec- often raised in mental health care relates call for a culture change in health care

practice. In order to achieve this culture social marketing can only be one piece mentation science to effectively amplify
change in the clinical world and move of the jigsaw. the voice of those with mental illness in
away from paternalism, we need to do We recommend the Consolidated Frame- making treatment decisions through an
more than just change attitudes and work for Implementation Research7 to SDM process.
norms of individual health care profes- develop a theoretically based implemen-
Isabelle Scholl1, Paul J. Barr2
sionals. Change is needed at all levels, tation strategy. This stresses the need to 1
Department of Medical Psychology, University Medical
from individual to organizational and in- foster implementation at different levels Center Hamburg-Eppendorf, Hamburg, Germany; 2Dart-
stitutional. (e.g., individual, organizational, policy) mouth Institute for Health Policy and Clinical Practice,
Dartmouth College, Hanover, NH, USA
Slade correctly points out that when and describes social marketing as one
considering how to transform mental among a range of other activities (e.g., 1. Slade M. World Psychiatry 2017;16:146-53.
health care systems both regarding SDM education, role modeling, training) to en- 2. Storm M, Edwards A. Psychiatr Q 2013;84:313-27.
3. Agoritsas T, Heen AF, Brandt L et al. BMJ 2015;
and other possible upcoming changes it gage stakeholders at the individual level.
350:g7624.
could be helpful to use language and Another seminal model is the Behavior 4. Shepherd HL, Barratt A, Trevena LJ et al.
constructs from other sectors to inform Change Wheel8, which can be used to Patient Educ Couns 2011;83:379-85.
5. Henderson C, Farrelly S, Moran P et al. World
this transformation1. When discussing design behavior change interventions
Psychiatry 2015;14:281-3.
the implementation of SDM, whether in to foster routine implementation of SDM. 6. Wong JG, Clare CH, Holland AJ et al. Psychol
mental health care or in other clinical In summary, we applaud Slade for his Med 2000;30:295-306.
7. Damschroder LJ, Aron D, Keith RE et al. Imple-
areas, we should carefully consider trans- effort to push forward the SDM agenda
ment Sci 2009;4:50.
lating knowledge from the field of imple- in the mental health field. We agree 8. Michie S, van Stralen M, West R. Implement
mentation science to influence clinical with his conclusion that implementation Sci 2011;6:42.
care. For successful implementation we challenges are the key concern. Social
DOI:10.1002/wps.20418
need to take a range of basic sciences marketing and insights from the hospi-
(e.g., behavioral science, psychology, com- tality industry are unique and helpful,
munication, economics) into account; thus, but they must be combined with imple-
Mental health shared decision making in the US

M. Slades paper1 presents the most away from hospitals, expensive medica- decreasing self-esteem and opportunities,
accurate, balanced and up-to-date sum- tions, specialists, facility-based rehabilita- harm society by increasing stigma and
mary of shared decision making in men- tion, and other profit-generating services, segregation, and harm government by
tal health care that is currently available. even though studies show that patients wasteful spending.
Because his review takes a decidedly UK would prefer other services such as safe The crux of the US problem is that
perspective, I will address some of the housing, employment, peer supports, and prevention and social safety net services,
related issues in the US. help with general functioning3,4. People though preferred by people with mental
The US health care system (more ac- with mental illness recognize the need to health challenges, do not generate prof-
curately, the US health care non-system) address the social issues that cause and its. Effective interventions for primary, sec-
continues to be extraordinarily expen- exacerbate mental disorders. But shared ondary and tertiary prevention in mental
sive and ineffective. Health care services decision making may not include the health exist, but in the US we spend mini-
in the US have been created by vested services they want and need. mally in these areas. Northern European
interest groups: private hospitals, phar- Medical solutions to social problems are countries, by contrast, spend less on health
maceutical companies, insurance agen- very expensive and ineffective. Yet social care but more on the social safety net: pre-
cies, device makers, professional guilds, factors often determine exacerbations of natal services, early childhood care, mater-
specialty care groups, large health con- mental illness and cause excessive, un- nal leave, family support, early education,
glomerates, for-profit nursing homes, and necessary mental health treatment. Con- nutrition, early behavioral health interven-
so on. All of these entities prosper in the sider the current trends to increase mental tions, safe housing, and psychosocial sup-
US by providing services that maximize hospital beds and to incarcerate people ports for people with disabilities5.
profits rather than patient outcomes. with mental illness. The erosion of low- Consider the examples of supported
Although patient-centered care is widely cost housing and the absence of employ- housing6, supported employment7, and
endorsed as a principle in the US2, it is ment opportunities, rather than true in- supported medication management8. These
more honored in the breach than the creases in the prevalence or severity of interventions are highly effective, strongly
observance. In mental health, the call for mental illness, underlie these misguided desired by people with mental illness, and
patient-centered care and shared deci- initiatives. In fact, hospitals and prisons clearly helpful for recovery. But they are
sion making seems unlikely to shift care often harm people with mental illness by rarely available because social services

are not considered medical necessities. What if they involve listening to patients Robert E. Drake
Shared decision making cannot address rather than to vested interest groups?9 Dartmouth Institute on Health Policy and Clinical Prac-
tice, Geisel School of Medicine at Dartmouth, Lebanon,
unavailable services. More money, more clinical trials, and NH, USA
As health care costs in the US spiral more professionalization may not solve
out of control, policy makers and health problems that are related to social, educa- 1. Slade M. World Psychiatry 2017;16:146-53.
care leaders have pursued economic tional, economic and health inequities10. 2. Institute of Medicine. Shared decision-making
strategies for best care: patient decision aids.
outcomes such as lower hospital and Thus, shared decision making, to drive Washington: National Academy of Medicine, 2014.
emergency use rather than increased ineffective change in the US, must address 3. Shumway M, Saunders T, Shern D et al. Psy-
volvement or satisfaction with the health more than traditional medical interven- chiatr Serv 2003;54:1124-8.
4. Woltmann E, McHugo GJ, Drake RE. Psychiatr
encounter. Adoption of decision aids tions. People with mental health problems Serv 2011;62:54-60.
and shared decision making has been need and want safe neighborhoods, decent 5. Squires D, Anderson C. U.S. health care from a
largely ignored. Instead, policy makers housing, and opportunities for education, global perspective: spending, use of services,
prices, and health in 13 countries. New York:
continue to try to change incentives and employment and community integration. Commonwealth Fund, 2015.
risk adjustments within the health care Yet they are getting more and more medi- 6. Tsemberis S, Kent D, Respress C. Am J Publ
system in order to reduce costs. Thus far, cines, forced treatments, hospitals and Health 2012;102:13-6.
7. Drake RE, Bond GR, Goldman HH et al. Health
managed care, accountable care organi- psychiatric specialists. As inequality, preju- Affairs 2016;35:1098-105.
zations, paying for performance, behav- dice and health disparities expand in the 8. Deegan PE, Drake RE. Psychiatr Serv 2006;57:
ioral health integration, and other popular US, we must listen to people in a broader 1636-9.
9. Drake RE, Binagwaho A, Martell HC et al. BMJ
approaches to reform have not succeeded sense let people choose the services and 2014;349:7086.
as though we do not quite have the incen- outcomes that matter to them. People with 10. Mulley A, Richards T, Abbasi K. BMJ 2015;351:
tives and adjustments right! But what if mental health disabilities deserve access to h4448.
11. Americans with Disabilities Act of 1990. Pub.
the solutions are outside of the traditional housing, schools, jobs, family supports and L. No. 101-336, 104 Stat. 328 (1990).
health care system? What if they do not safety11. But will they be allowed to share
generate profits for medical industries? in these decisions? DOI:10.1002/wps.20419

RESEARCH REPORT
Prevalence, incidence and mortality from cardiovascular disease in

patients with pooled and specific severe mental illness: a large-scale
meta-analysis of 3,211,768 patients and 113,383,368 controls
Christoph U. Correll1-5, Marco Solmi5-7, Nicola Veronese5, Beatrice Bortolato5,8, Stella Rosson6, Paolo Santonastaso6, Nita Thapa-Chhetri9,
Michele Fornaro10, Davide Gallicchio6, Enrico Collantoni6, Giorgio Pigato6, Angela Favaro6, Francesco Monaco5, Cristiano Kohler11,
Davy Vancampfort12,13, Philip B. Ward14, Fiona Gaughran15, Andre F. Carvalho5,11, Brendon Stubbs5,15-17
1
Psychiatry Research, Zucker Hillside Hospital, Northwell Health, Glen Oaks, NY, USA; 2Department of Psychiatry and Molecular Medicine, Hofstra Northwell School of Medi-
cine, Hempstead, NY, USA; 3Center for Psychiatric Neuroscience, Feinstein Institute for Medical Research, Manhasset, NY, USA; 4Department of Psychiatry and Behavioral
Medicine, Albert Einstein College of Medicine, Bronx, NY, USA; 5Institute for Clinical Research and Education in Medicine, Padua, Italy; 6Department of Neurosciences, Univer-
sity of Padua, Padua, Italy; 7Mental Health Department, Local Health Unit 17, Padua, Italy; 8Mental Health Department, Local Health Unit 10, Portogruaro, Italy; 9University of
Connecticut Health Center, Farmington, CT, USA; 10New York Psychiatric Institute, Columbia University, New York, NY, USA; 11Department of Clinical Medicine and Transla-
tional Psychiatry Research Group, Federal University of Ceara, Fortaleza, Brazil; 12KU Leuven Department of Rehabilitation Sciences, Leuven, Belgium; 13KU Leuven University
Psychiatric Center, Leuven-Kortenberg, Belgium; 14School of Psychiatry, University of New South Wales, Sydney, Australia; 15South London and Maudsley, NHS Foundation
Trust, London, UK; 16Health Service and Population Research Department, Institute of Psychiatry, Psychology and Neuroscience, Kings College London, London, UK; 17Depart-
ment of Psychosis Studies, Institute of Psychiatry, Psychology and Neuroscience, Kings College London, London, UK
People with severe mental illness (SMI) schizophrenia, bipolar disorder and major depressive disorder appear at risk for cardiovascular dis-
ease (CVD), but a comprehensive meta-analysis is lacking. We conducted a large-scale meta-analysis assessing the prevalence and incidence of
CVD; coronary heart disease; stroke, transient ischemic attack or cerebrovascular disease; congestive heart failure; peripheral vascular disease;
and CVD-related death in SMI patients (N53,211,768) versus controls (N5113,383,368) (92 studies). The pooled CVD prevalence in SMI
patients (mean age 50 years) was 9.9% (95% CI: 7.4-13.3). Adjusting for a median of seven confounders, patients had significantly higher odds
of CVD versus controls in cross-sectional studies (odds ratio, OR51.53, 95% CI: 1.27-1.83; 11 studies), and higher odds of coronary heart dis-
ease (OR51.51, 95% CI: 1.47-1.55) and cerebrovascular disease (OR51.42, 95% CI: 1.21-1.66). People with major depressive disorder were at
increased risk for coronary heart disease, while those with schizophrenia were at increased risk for coronary heart disease, cerebrovascular dis-
ease and congestive heart failure. Cumulative CVD incidence in SMI patients was 3.6% (95% CI: 2.7-5.3) during a median follow-up of 8.4
years (range 1.8-30.0). Adjusting for a median of six confounders, SMI patients had significantly higher CVD incidence than controls in longi-
tudinal studies (hazard ratio, HR51.78, 95% CI: 1.60-1.98; 31 studies). The incidence was also higher for coronary heart disease (HR51.54,
95% CI: 1.30-1.82), cerebrovascular disease (HR51.64, 95% CI: 1.26-2.14), congestive heart failure (HR52.10, 95% CI: 1.64-2.70), and CVD-
related death (HR51.85, 95% CI: 1.53-2.24). People with major depressive disorder, bipolar disorder and schizophrenia were all at increased
risk of CVD-related death versus controls. CVD incidence increased with antipsychotic use (p50.008), higher body mass index (p50.008) and
higher baseline CVD prevalence (p50.03) in patients vs. controls. Moreover, CVD prevalence (p50.007), but not CVD incidence (p50.21),
increased in more recently conducted studies. This large-scale meta-analysis confirms that SMI patients have significantly increased risk of
CVD and CVD-related mortality, and that elevated body mass index, antipsychotic use, and CVD screening and management require urgent
clinical attention.
Key words: Cardiovascular disease, severe mental illness, schizophrenia, bipolar disorder, major depression, coronary heart disease, cere-
brovascular disease, congestive heart failure, premature mortality
People with severe mental illness (SMI) including schizo- CVD risk6,7, including stroke (risk ratio, RR51.34, 95% CI:
phrenia, bipolar disorder, major depressive disorder, and their 1.17-1.54), myocardial infarction (hazard ratio, HR51.31, 95%
related spectrum disorders have a life expectancy shortened CI: 1.09-1.57), coronary heart disease (RR51.36, 95% CI:
of 10-17.5 years compared to the general population1,2. While 1.24-1.49) and coronary heart disease-related death (HR51.36,
suicide explains some of this reduced life expectancy3, it is 95% CI: 1.14-1.63)6-8. While clearly informative, results concern-
now established that physical diseases account for the over- ing CVD were not specific for major depressive disorder defined
whelming majority of premature mortality4,5. Among physical according to established diagnostic criteria, possibly biasing
conditions, cardiovascular disease (CVD) is the main potentially such observed association towards a lower risk9. Another
avoidable contributor to early deaths in patients with SMI4. meta-analysis of longitudinal studies, which utilized standard-
Given the importance of understanding the magnitude, ized criteria to define bipolar disorder, reported mixed results,
contributors to and relative distribution of CVD risk in people since people with that disorder were actually not at increased
with SMI, a number of disease-specific meta-analyses investi- risk of myocardial infarction (RR51.09, 95% CI: 0.96-1.24),
gated if people with major depressive disorder, bipolar disor- whereas the risk of stroke was higher compared to controls
der or schizophrenia are at an increased risk of CVD compared (RR51.74, 95% CI: 1.29-2.35)10. Among individuals with schiz-
to controls. These meta-analyses reported that people with ophrenia, previous meta-analyses11,12 reported an overall
depression (defined by the presence of depressive symptoms increased risk of CVD compared to controls (RR51.53, 95% CI:
or a diagnosis of major depressive disorder) are at increased 1.27-1.86). This risk increase included stroke (up to RR51.71,

95% CI: 1.19-2.46) and heart failure (RR51.81, 95% CI: 1.42-2.29), Inclusion and exclusion criteria
but not coronary heart disease (RR51.20, 95% CI: 0.93-1.53).
While the existing literature has provided relevant insights, We included studies with the following characteristics: a)
several limitations are to be highlighted and important questions reporting on patients with schizophrenia, schizophrenia spec-
remain unanswered. First, some of the previous meta-analyses trum or schizoaffective disorder, bipolar disorder or bipolar
did not use standardized clinical assessments to identify and spectrum disorders, major depressive disorder or depressive
categorize SMI and/or cardiovascular events. Second, the exact episodes, or SMI (defined as at least two among major depres-
prevalence and incidence of each type of CVD among people sive spectrum, bipolar spectrum and schizophrenia spectrum
with SMI, both within and across major diagnostic SMI sub- disorders) according to DSM-III, DSM-IV, DSM-5, ICD-8, ICD-
groups, remains unclear. Third, the magnitude of premature 9 or ICD-10, or a medical record diagnosis based on a clinical
CVD-related mortality risk in people with SMI versus controls interview; b) having a cross-sectional or a retrospective/
is to be specified. Fourth, potential risk factors for increased prospective longitudinal design, either with or without a control
CVD and related mortality risk across the SMI groups have not group; c) using a standardized definition of CVD; d) reporting RR,
been elucidated via meta-analytic techniques, which could HR or odds ratio (OR) comparing patients with region-specific
help identify targets for treatment guidelines, clinical stand- controls, percentage or number of events at baseline (data used
ards and development of preventive and therapeutic pro- for cross-sectional analysis 5 prevalence) and/or follow-up (data
grams. In this regard, large-scale pooled analyses in the SMI used for longitudinal analysis 5 cumulative incidence).
population can provide relevant information, allowing the We excluded studies that investigated cardiovascular risk
investigation of potentially shared risk factors across many
estimates and/or factors, subclinical CVD, or SMI rates in pop-
studies and participants, thus dissecting CVD risk factors asso-
ulations with CVD. In case of multiple publications from the
ciated with SMI and/or treatments for these disorders from
same study, only the most recent paper or the article with the
factors which are non-specific or shared with the general pop-
longest follow-up was included. When required, we contacted
ulation13. Additionally, pooling of data allows for the investiga-
the primary/corresponding authors of potential studies to
tion of demographic, regional and treatment variables, both
confirm eligibility or acquire unpublished variables of interest.
within and across major diagnostic categories.
Given the caveats mentioned above, the current gaps within
the literature and the need to better understand CVD risk among Data extraction
people with SMI, we conducted a large scale meta-analysis
investigating the prevalence, incidence and mortality attributed Seven authors divided in four pairs independently extracted
to CVD and their correlates among people with SMI, both data in a standardized Microsoft Excel sheet, with reciprocal
within and across major diagnostic groups. validation of data extraction results. The extracted data includ-
ed: authors, year and country; geographic region; study design;
data source; period of data collection; SMI diagnostic criteria;
METHODS CVD diagnostic criteria; specific SMI and CVD diagnosis; case
and control inclusion criteria; number of cases and controls;
This systematic review and meta-analysis adhered to the percentage or number with CVD, coronary heart disease, cere-
PRISMA statement14, following a predetermined, but unpub- brovascular disease and congestive heart failure at baseline;
lished protocol. number of events at follow-up; follow-up duration; number
and type of covariates considered in the analyses; OR, RR, rate
ratio and HR with their respective 95% upper and lower CIs;
Search strategy mean age with standard deviation; mean body mass index
with standard deviation; proportion of males; co-occurring
An electronic literature search was conducted in PubMed, obesity, alcohol and substance related disorders, diabetes,
Embase and Scopus from database inception until August 2, hypertension, and hyperlipidemia; married status; employ-
2016 by two independent reviewers, using the search terms ment status; percentage of patients with poorest income and
(bipolar disorder OR mania OR schizophrenia OR schizoaffec- least urbanized; and percentage of patients taking antipsy-
tive OR psychosis OR major depression OR serious mental chotics. Rate ratios calculated with Cox regression models
illness) AND (cardiovascular OR stroke OR cerebrovascular OR were included in HR analyses. When authors did not specify
transient ischemic attack OR transient ischaemic attack OR
whether or not a rate ratio had been calculated with Cox
peripheral vascular OR myocardial infarction OR coronary
regression models, we contacted them seeking clarification.
heart disease OR coronary artery disease OR ischemic heart
disease OR ischaemic heart disease OR hypertensive heart
disease OR angina OR cardiac failure OR heart failure OR Outcomes
congestive heart failure OR atrial fibrillation OR pulmonary
embolism OR cardiovascular mortality). Furthermore, bibli- Primary outcomes were CVD prevalence and cumulative
ographies of included papers were reviewed. incidence plus CVD-related mortality in people with SMI, as

well as adjusted OR for prevalence and HR for incidence rates status: geographical region of the sample; time of data collec-
in SMI versus controls. Secondary outcomes were the same tion; percentage of patients taking antipsychotics; and the dif-
measures for specific CVDs (i.e., coronary heart disease, cere- ference between patient and control samples regarding age,
brovascular disease, congestive heart failure) in SMI patients, body mass index, proportion of males and of those with married
as well as adjusted OR and HR versus controls. status, unemployed, with poorest income, least urbanized, and
Prevalence and OR were calculated from cross-sectional having co-occurring obesity, alcohol and substance-related dis-
studies and from baseline results of longitudinal studies. orders, diabetes, hypertension or hyperlipidemia.
Where available, incidence, RR and HR were calculated from
longitudinal studies.
RESULTS
Quality assessment
Search results
For the purpose of this meta-analysis, a checklist (yes versus
Out of 18,064 initial hits across the searched electronic
no) was used to assess the methodological quality of included
databases, 11,878 unduplicated hits were screened, and 11,576
studies. The evaluation of methodological quality across stud-
were excluded through title/abstract reading. Altogether, 302
ies was based on the following factors: clear diagnostic criteria,
full texts were reviewed, and 210 were excluded with specific
presence of a control group, matching of the control group,
reasons. Among 92 studies meeting inclusion criteria, 27 had a
covariate-adjusted outcomes, reported cardiovascular risk fac-
cross-sectional design22-48 and 65 studies had a retrospective
tors at baseline, and follow-up 5 years.
or prospective longitudinal design49-113 (Figure 1).
Data analysis
Characteristics of included studies
This meta-analysis was performed using Comprehensive
Meta-Analysis V315. All outcomes were meta-analyzed when at We included 92 studies, with a total population of 3,211,768
least two studies provided data. A random effects model16,17 patients (mean age 50 years, 49% male) with SMI and
was used to account for between-study heterogeneity. We cal- 113,383,368 controls (mean age 51 years, 49% male), with a total
culated pooled CVD prevalences and pooled CVD cumulative of 116,595,136 subjects when summing those studies where
incidences, each with SMI subgrouping. For dichotomous pri- patient and control sample sizes were not separately reported.
Altogether, 27 studies (N527,037,943) were cross-sectional and
mary and secondary outcomes comparing pooled SMI and
65 studies (N589,557,193) were longitudinal. Overall, 38 studies
SMI subgroups with controls, we calculated unadjusted as
included patients with schizophrenia (of which 29 were longitu-
well as adjusted pooled OR for cross-sectional data, and unad-
dinal), 30 with bipolar disorder (21 longitudinal), 30 with major
justed pooled RR, as well as adjusted pooled HR, for longitudi-
depressive disorder (22 longitudinal), and 14 with SMI (8 longitu-
nal data. Funnel plots were visually inspected, and Eggers
dinal). Taken together, six studies included only patients with
test18 and Begg-Mazumdar Kendalls tau19 were used to deter-
SMI (N5884,412), 16 studies included only patients with bipolar
mine if publication bias was likely. When publication bias was
disorder (N571,832), 20 studies included only patients with
present, the trim and fill20 procedure was run to evaluate if the
major depressive disorder (N5111,360), and 29 studies includ-
results changed after imputing potentially missing studies.
ed only patients with schizophrenia (N51,591,106), while 19
Between-study heterogeneity was measured using the chi-
studies included different subgroups of SMI, providing data for
squared and I-squared statistics, with chi-squared p<0.05 and
each of them separately (some studies included more than
I-squared 50% indicating significant heterogeneity21. To iden-
one diagnostic group, see Tables 1 and 2 for details).
tify potential moderators, meta-regression was run with Com-
prehensive Meta-Analysis V3 for unadjusted outcomes where
heterogeneity was significant. Meta-analysis: cross-sectional results
Since CVD rates in the general population vary across the
world, we also performed a stratified analysis across geograph- The pooled CVD prevalence in SMI was 9.9% (95% CI: 7.4-13.3;
ic regions (Asia, Europe, North America, Oceania) regarding 38 studies). Individual rates were 8.4% for people with bipolar
raw CVD prevalence and incidence in SMI populations, and disorder (95% CI: 5.4-12.6, 12 studies, N566,911); 11.7% for
compared patients to their respective region-specific general those with major depressive disorder (95% CI: 3.6-32.2, 7 stud-
population controls (calculating RRs as well as adjusted ORs ies, N583,965); 11.8% for those with schizophrenia (95% CI:
and HRs for the four regional strata and comparing them 7.1-19.0, 13 studies, N5191,982), and 11.8% for those with SMI
across the different regions whenever at least two studies pro- (95% CI: 4.1-29.4, 6 studies, N517,286) (p<0.001 for SMI diag-
vided data per each region). nostic subgroup comparisons).
The following study and patient characteristics were explor- Adjusting for a median of seven potential confounders, the
ed as potential moderators and mediators in addition to SMI adjusted pooled OR for CVD in SMI compared to controls was

Figure 1 PRISMA flow chart
1.53 (95% CI: 1.27-1.83, p<0.001, 11 studies). For specific pooled previously significant ORs remained statistically signifi-
CVDs, pooled together, people with SMI had an increased risk cant, confirming the association of CVD, coronary heart disease
of coronary heart disease (OR51.51, 95% CI: 1.47-1.55, and cerebrovascular disease with SMI, while the OR for conges-
p<0.001, 5 studies) and cerebrovascular disease (OR51.42, tive heart failure became marginally significant (p50.05).
95% CI: 1.21-1.66, p<0.001, 6 studies), with a strong statistical
trend for congestive heart failure (OR51.28, 95% CI: 0.99-1.65,
Meta-analysis: longitudinal adjusted results
p50.06, 4 studies). Considering separately single types of SMI
and CVD, in adjusted OR analyses, bipolar disorder was not Among patients with SMI, 3.6% (95% CI: 2.7-5.3%) experi-
significantly associated with CVD or its subtypes; major enced a CVD event during a median follow-up period of 8.4
depressive disorder was significantly associated with CVD and years (range 1.8-30.0) (65 studies). After adjusting for a median
coronary heart disease; and schizophrenia was significantly of six potential confounders, people with SMI were at signifi-
associated with coronary heart disease, cerebrovascular dis- cantly increased risk across longitudinal studies for CVD
ease and congestive heart failure (Table 3). No adjusted ORs (HR51.78, 95% CI: 1.60-1.98) (31 studies, N5671,384 cases vs.
were available for mixed SMI groups. N514,335,203 controls) as well as for specific CVDs, including
All significant results were significantly heterogeneous. After coronary heart disease (HR51.54, 95% CI: 1.30-1.82, 18 studies,
adjusting for publication bias with the trim-and-fill method, all N5194,017 cases vs. N513,530,858 controls), cerebrovascular

Table 1 Cross-sectional studies: characteristics of included studies and samples
No. No. Period of No.

Study Country cases controls data collection SMI definition Inclusion criteria for cases covariates
Beyer et al22 USA 1,379 - 2001-2002 Medical records Bipolar disorder -

23
Bresee et al Canada 28,775 2,281,636 1995-2006 ICD-9,10 Schizophrenia 4
Bresee et al24 Canada 399 120,044 2005 Medical records Schizophrenia 11
Chen et al25 Taiwan 80 - 2015 DSM-IV Bipolar disorder, >60 years -
Curkendall et al26 Canada 3,022 12,088 1994-1999 ICD-9 Schizophrenia 7
Devantier et al27 Denmark 28 27 2009-2011 ICD-10 Major depressive disorder, late onset -
Hagg et al28 Sweden 269 - 2000-2003 DSM-IV Schizophrenia, 20-69 years -
Herbst et al29 USA 10,573 total population 2001-2002 DSM-IV Major depressive disorder, >60 years 11
Huang et al30 Taiwan 117,987 21,356,304 2000-2003 ICD-9 Bipolar disorder or major depressive disorder 1
31
Hyde et al Australia 355 - 2008-2012 Medical records Severe mental illness, prescribed clozapine -
Kilbourne et al32 USA 8,083 - 2001 ICD-9 Severe mental illness, >60 years -
Kilbourne et al33 USA 9,705 5,353 2000-2001 ICD-9 Bipolar disorder or severe mental illness, male 3
Lindegard34 Sweden 368 87,176 1966-1979 ICD-9, DSM-III Major depressive disorder or bipolar disorder -
35
Maina et al Italy 185 - 2006-2008 DSM-IV Severe mental illness -
Morden et al36 Canada 65,362 65,362 2000-2007 ICD-9 Schizophrenia 4
Munoli et al37 India 120 - 2011 ICD-10 Bipolar disorder -
Nielsen et al38 Denmark 937 - 1969-2014 ICD-10 Schizophrenia -

39
Niranjan et al USA 5,695 34,979 2007 DSM-IV Major depressive disorder 6
Oreski et al40 Croatia 289 192 2011 ICD-10 Bipolar disorder or schizophrenia -
41
Prieto et al USA 988 - 2009-2013 DSM-IV Severe mental illness -
Scherrer et al42 USA 628 6,903 1990-1992 DSM-III Major depressive disorder, male twins -
43
Scott et al Multicenter 52,095 total population 2001-2011 DSM-IV Bipolar disorder or major depressive disorder 6
Shen et al44 Taiwan 203 2,036 2005-2007 ICD-9 Schizophrenia, in intensive care unit 6
Smith et al45 UK 9,677 1,414,701 2007 Medical records Schizophrenia 3
Smith et al46 UK 2,582 1,421,796 2007 Medical records Bipolar disorder 2
Swain et al47 Multicenter 45,288 total population 2001-2011 DSM-IV Bipolar disorder or major depressive disorder 7
Zilkens et al48 Australia 656 349 2000-2009 ICD-8,9,10 Major depressive disorder, 65-84 years, -
developing dementia
SMI severe mental illness
disease (HR51.64, 95% CI: 1.26-2.14, 11 studies, N5188,841 (HR51.65, 95% CI: 1.10-2.47, 3 studies), with a trend toward a
cases vs. N513,113,564 controls), congestive heart failure significant association with cerebrovascular disease (HR51.60,
(HR52.10, 95% CI: 1.64-2.70, 2 studies, N5409 cases vs. 95% CI: 0.99-2.57, 4 studies), but no significant association
N541,678 controls), peripheral vascular disease (only unad- with coronary heart disease (HR51.16, 95% CI: 0.76-1.78, 4
justed RR53.11, 95% CI: 2.46-3.91, three studies), and CVD- studies). One study reported a significant association with
related death (HR51.85, 95% CI: 1.53-2.24, 16 studies, N5353,407 congestive heart failure (HR 5 2.27, 95% CI: 1.49-3.45).
cases vs. N57,317,053 controls). According to adjusted HRs, major depressive disorder was
According to adjusted HRs, schizophrenia was significantly significantly associated with CVD in longitudinal studies
associated with CVD in longitudinal studies (HR51.95, 95% (HR51.72, 95% CI: 1.48-2.00, 18 studies) as well as with coro-
CI: 1.41-2.70, 14 studies), as well as with coronary heart dis- nary heart disease (HR51.63, 95% CI: 1.33-2.00, 9 studies),
ease (HR51.59, 95% CI: 1.08-2.35, 5 studies), cerebrovascular cerebrovascular disease (HR52.04, 95% CI: 1.05-3.96, 3 stud-
disease (HR51.57, 95% CI: 1.09-2.25, 5 studies), and CVD- ies), congestive heart failure (HR52.02, 95% CI: 1.48-2.75, 2
related death (HR52.45, 95% CI: 1.64-3.65, 9 studies). studies), and CVD-related death (HR51.63, 95% CI: 1.25-2.13,
According to adjusted HRs, bipolar disorder was significant- 7 studies).
ly associated with CVD in longitudinal studies (HR51.57, 95% According to adjusted HRs, mixed SMIs were significantly
CI: 1.28-1.93, 10 studies) as well as with CVD-related death associated with CVD in longitudinal studies (HR53.24, 95%

Table 2 Longitudinal studies: characteristics of included studies and samples
No. No. Period of SMI No.

Study Country cases controls data collection definition Inclusion criteria for cases covariates
Almeida et al49 Australia 1,503 35,691 1996-2010 ICD-9 Schizophrenia, bipolar disorder or major 8
depressive disorder, 65-85 years, male
Bremmer et al50 The Netherlands 41 2,080 1992-2000 DSM-III Major depressive disorder, >55 years 13
51
Butnoriene et al Lithuania 184 369 2003-2004 DSM-IV Major depressive disorder, >45 years 4
Callaghan et al52 Canada 5,999 5,999 2002-2006 Medical Bipolar disorder 6

records
Callaghan et al53 Canada 9,815 9,815 2002-2006 ICD-10 Bipolar disorder 8

54
Carney et al USA 1,074 726,262 1996-2001 ICD-9 Schizophrenia or schizoaffective disorder 4
Chen et al55 Taiwan 63,913 63,913 2002-2008 ICD-9 Schizophrenia 8
Clouse et al56 USA 16 60 1982-1992 DSM-III Major depressive disorder with diabetes 7
Coryell et al57 USA 903 - 1998-1999 RDC Severe mental illness -

58
Crump et al Sweden 6,618 6,580,418 2003-2009 ICD-10 Bipolar disorder 6
Crump et al59 Sweden 8,277 6,097,834 2003-2009 ICD-10 Schizophrenia, >25 years 6
Davis et al60 Hawaii 280 39,000 1999-2005 Medical Major depressive disorder 5
records
Davydow et al61 Denmark 68,137 5,912,158 1999-2013 ICD-9 Schizophrenia, schizoaffective disorder or 5
bipolar disorder
Enger et al62 USA 1,920 9,600 1995-1999 ICD-9 Schizophrenia, on antipsychotic treatment, -
15-64 years
Fiedorowicz et al63 USA 288 147 1978-1981 RDC Bipolar disorder 8

64
Filik et al UK 482 1,998 1999-2002 DSM-IV Schizophrenia, schizophreniform or 6
schizoaffective disorder
Fors et al65 Sweden 255 1,275 1981-1991 DSM-II Schizophrenia 3
Gasse et al66 Denmark 873,898 52,693,301 1995-2009 ICD-8,10 Severe mental illness (affective psychoses) 22
Goldstein et al67 USA 5,835 26,266 2001-2005 DSM-IV Bipolar disorder or major depressive disorder 8
Healy et al68 UK 1,429 - 1875-1924; Medical Schizophrenia -

1994-2010 records
Hendrie et al69 USA 757 30,831 1999-2008 ICD-9 Schizophrenia, >65 years -
Hou et al70 Taiwan 8,264 - 1985-2008 DSM-III or Schizophrenia -

IV, ICD-9
Hsieh et al71 Taiwan 9,715 - 2001-2009 ICD-9 Schizophrenia 10

72
Huang et al Taiwan 7,937 31,748 1996-2006 ICD-9 Major depressive disorder 9
Ifteni et al73 Romania 7,189 - 1989-2011 DSM-IV Schizophrenia, inpatients -

74
Jakobsen et al Denmark 74,759 338,747 1977-2000 ICD-8,10 Schizophrenia or major depressive disorder 2
Janszky et al75 Sweden 646 48,675 1969-2007 ICD-8 Major depressive disorder, 18-20 years 7
Jokinen & Sweden 346 - 1980-2005 DSM-IV Major depressive disorder or bipolar disorder -
Nordstrom76
Joukamaa et al77 Finland 606 8,000 1977-1994 Medical Schizophrenia, mood disorder or severe 1
records mental illness
Kendler et al78 Sweden 5,647 24,727 1998-2003 ICD-10 Major depressive disorder, twins -
Kiviniemi et al79 Finland 6,987 - 1998-2003 ICD-9 Schizophrenia, first onset -

Lahti et al80 Finland 204 11,880 1969-2004 ICD-8,9,10 Schizophrenia 5
Lan et al81 Taiwan 3,681 - 2001-2006 ICD-9 Bipolar disorder -

82
Laursen et al Denmark 22,294 2,411,852 1995-2007 ICD-8,10 Schizophrenia or bipolar disorder, 15-52 3
years
Laursen et al83 Denmark 1,454 59,256 1995-2006 ICD-8,10 Schizophrenia or bipolar disorder 4

Table 2 Longitudinal studies: characteristics of included studies and samples (continued)
No. No. Period of SMI No.
Study Country cases controls data collection definition Inclusion criteria for cases covariates
Lemogne et al84 France 4,336 16,621 1990-2010 ICD-9,10 Depression or severe mental illness 6
(bipolar disorder, psychosis)
Li et al85 Taiwan 1,003 4,012 1996-2009 ICD-9 Major depressive disorder 6
Lin et al86 Taiwan 7,353 22,059 2000-2006 ICD-9 Schizophrenia 8

87
Lin et al Taiwan 2,289 16,413 1998-2003 ICD-9 Bipolar disorder 10
Lin et al88 Taiwan 5,001 10,002 1998-2003 ICD-9 Schizophrenia, <45 years 9
89
Maina et al Italy 309 - 2003-2011 DSM-IV Bipolar disorder -
McDermott et al90 USA 503 2,083 1990-2003 ICD-9 Schizophrenia or severe mental illness 9
Murray-Thomas et al91 UK 232,132 193,920 1997-2001 ICD-10 Schizophrenia, bipolar disorder or major 2
depressive disorder
Olfson et al92 USA 1,138,853 - 2001-2007 ICD-10 Schizophrenia, 20-64 years 4
Osborn et al93 UK 38,824 - 1995-2010 Medical Bipolar disorder or severe mental illness, -
records 30-90 years
Pratt et al94 USA 73 1,107 1981-1994 DSM-III Major depressive disorder 11

95
Prieto et al USA 334 334 1966-1996 DSM-IV Bipolar disorder 4
Rahman et al96 Sweden 6,822 29,832 1998-2002 ICD-7,8,9,10 Major depressive disorder, twin population 7
study
Ramsey et al97 USA 129 1,339 1981-1982 DSM-III Bipolar disorder or major depressive disorder 6
Saint Onge et al98 USA 548 10,821 1999-2006 ICD Major depressive disorder 11
Scherrer et al99 USA 77,568 214,749 1999-2007 ICD-9 Major depressive disorder, 25-80 years 4
Schoepf & Heun100 UK 1,418 14,180 2000-2012 ICD-10 Schizophrenia, inpatients -
101
Schoepf et al UK 621 6,210 2000-2012 ICD-10 Bipolar disorder -
Shah et al102 USA 538 7,103 1988-2006 DSM-III Major depressive disorder or bipolar 14
disorder, 17-39 years
Stewart et al103 USA 235 - NA ICD-9 Major depressive disorder -

104
Surtees et al UK 3,057 16,592 1996-2008 DSM-IV Major depressive disorder, 45-80 years 11
Ting et al105 China 153 7,682 1996-2008 DSM-IV Major depressive disorder with diabetes 18
Torniainen et al106 Sweden 21,492 214,920 2006-2015 ICD-10 Schizophrenia, 17-65 years 2
Tsai et al107 Taiwan 80,569 241,707 1999-2003 ICD-9 Schizophrenia 8
Tsan et al108 USA 49,173 - 2002-2009 ICD-9 Schizophrenia -
van Marwijk et al109 The Netherlands 143 139 2002-2003 DSM-IV Major depressive disorder, >55 years -
Weeke et al110 Denmark 3,795 - 1950-1957; ICD-8 Bipolar disorder -
1969-1977
Westman et al111 Sweden 17,101 10,631,208 1987-2006 ICD-10 Bipolar disorder 3
Wu et al112 Taiwan 16,821 67,284 1999-2010 ICD-9 Bipolar disorder 9
Wu et al113 Taiwan 70,225 207,592 1996-2007 ICD-9 Schizophrenia or bipolar disorder 8
SMI severe mental illness, RDC Research Diagnostic Criteria
CI: 2.15-4.88, 3 studies) as well as with CVD-related death Quality assessment of included studies
(HR52.75, 95% CI: 1.32-5.73, 3 studies).
All significant results were significantly heterogeneous, Quality ratings of single studies are presented in Table 5. All
except for mixed SMI and CVD risk, as well as all the conges- studies used clear diagnostic criteria, by design. Among the 27
tive heart failure results. After trim and fill procedure, all cross-sectional studies, all except 9 studies had a control group,
results remained unchanged, and Egger test did not show any 5 studies used a matched control sample, 13 studies adjusted
evidence of publication bias influencing the results (see Table analyses for relevant covariates, and all except 6 studies reported
4 for details). cardiovascular risk factors. Among the 65 longitudinal studies, all

170
Table 3 Meta-analysis of cross-sectional studies: unadjusted and adjusted odds ratios
Meta-analysis of unadjusted odds ratios Meta-analysis of covariate adjusted odds ratios
Hetero- Hetero-
No. participants Unadjusted odds ratios No. participants Adjusted odds ratios
No. geneity No. geneity
Disorder studies Patients Controls OR 95% CI p I2 studies Patients Controls OR 95% CI p I2
Cardiovascular disease
Bipolar disorder 4 19,562 1,526,110 1.73 1.11 2.71 0.02 91 4 2,640 1,423,135 1.28 0.90 1.80 0.17 52
Major depressive disorder 3 1,577 47,851 2.08 1.51 2.88 <0.001 58 7 7,050 43,570 1.75 1.36 2.26 <0.001 69
Schizophrenia 10 190,584 4,100,315 1.23 0.92 1.65 0.16 99 5 42,076 3,860,505 1.38 0.93 2.05 0.11 96
Severe mental illnesses 1 146 2,083 1.59 0.87 2.88 0.13 - - - - - - - - -
Pooled 14 211,869 7,808,603 1.59a 1.32 1.91 <0.001 99 11 51,766 5,325,871 1.53e 1.27 1.83 <0.001 94
Coronary heart disease
Bipolar disorder 3 19,504 1,524,771 1.75 1.11 2.77 0.02 94 1 2,582 1,421,796 0.94 0.79 1.11 0.49 -
Major depressive disorder 1 958 35,691 2.44 2.13 2.79 <0.0001 - 3 6,323 41,882 2.52 1.81 3.52 <0.001 93
Schizophrenia 8 187,359 4,086,191 1.03 0.85 1.25 0.76 98 1 399 120,044 1.52 1.48 1.56 <0.001 -
Severe mental illnesses 1 146 2,083 1.02 0.56 1.83 0.96 - - - - - - - -
Pooled 8 207,967 4,160,030 1.80b 1.62 2.00 <0.001 98 5 9,304 1,583,722 1.51f 1.47 1.55 <0.001 90
Cerebrovascular disease
Major depressive disorder 3 1,577 47,851 2.24 1.33 3.79 0.003 81 2 656 349 1.64 0.96 2.78 0.07 72
Schizophrenia 5 41,071 37,77,039 1.63 1.19 2.24 0.003 96 3 32,196 2,413,768 2.05 1.59 2.64 <0.001 61
Severe mental illnesses 1 146 2,083 1.02 0.56 1.83 0.96 - - - - - - - -
Pooled 10 45,535 5,454,785 1.63c 1.31 2.02 <0.0001 93 6 35,434 3,835,913 1.42g 1.21 1.66 <0.001 90
Congestive heart failure
Bipolar disorder 1 2,582 1,421,796 1.38 1.03 1.84 0.03 - 1 2,582 1,421,796 1.11 0.80 1.54 0.53 0
Major depressive disorder - - - - - - - - - - - - - - - -
Schizophrenia 5 40,984 3,743,431 1.71 1.36 2.15 <0.001 92 3 41,474 5,708,425 1.60 1.06 2.40 0.02 97
Severe mental illnesses 1 146 2,083 1.59 0.87 2.88 0.13 - - - - - - - - -

d h
Pooled 6 43,712 5,167,189 1.57 1.32 1.87 <0.001 88 4 44,056 7,130,221 1.28 0.99 1.65 0.06 96
Bold values represent significant results

Egger test for bias: a2.24, p50.53; b6.41, p50.03 (Duval & Tweedie trim-and-fill procedure adjusted OR: 1.35, 95% CI: 0.98-1.83); c21.69, p50.24; d22.39, p50.15; e21.73, p50.18; f0.17, p50.93; g22.59,
p50.07; h25.26, p50.20
World Psychiatry 16:2 - June 2017

Table 4 Meta-analysis of longitudinal studies with publication bias assessment

Meta-analysis of unadjusted relative risk Meta-analysis of covariate adjusted hazard ratio
Hetero- Hetero-
No. participants Unadjusted relative risk No. participants Adjusted hazard ratio
Disorder studies Patients Controls RR 95% CI p I2 studies Patients Controls HR 95% CI p I2
Cardiovascular disease
Bipolar disorder 12 66,549 9,606,575 1.50 1.28 1.75 <0.0001 76 10 91,187 6,967,728 1.57 1.28 1.93 <0.0001 91
Major depressive disorder 13 328,431 800,718 1.29 0.92 1.81 0.14 99 18 282,621 682,045 1.72 1.48 2.00 <0.0001 67
Schizophrenia 16 361294 16,096,125 1.21 1.006 1.45 0.04 98 14 296,778 7,176,374 1.95 1.41 2.70 <0.0001 99
Severe mental illnesses 2 874022 52,709,922 2.44 1.13 5.25 0.02 74 3 798 31,724 3.24 2.15 4.88 <0.0001 0
a g
Pooled 33 1,630,296 76,031,192 1.38 1.23 1.54 <0.0001 98 31 671,384 14,335,203 1.78 1.60 1.98 <0.0001 95
Coronary heart disease
Schizophrenia 8 169,507 15,446,625 0.93 0.81 1.08 0.33 87 5 75,860 6,348,965 1.59 1.08 2.35 0.02 95
Severe mental illnesses 1 873,898 52,693,301 1.80 1.74 1.86 <0.0001 - - - - - - - - -
Pooled 17 1,212,362 75,235,865 1.75b 1.69 1.80 <0.0001 99 18 194,017 13,530,858 1.54h 1.30 1.82 <0.0001 92
Cerebrovascular disease
Major depressive disorder 4 8,121 41,665 1.55 1.02 2.35 0.04 77 3 7,046 38,853 2.04 1.05 3.96 0.04 74
Schizophrenia 8 243,254 15,475,608 1.48 1.21 1.81 <0.0001 96 5 157,964 6,425,336 1.57 1.09 2.25 0.02 95
Severe mental illnesses - - - - - - - - - - - - - - - -

c i
Pooled 17 284,273 22,187,932 1.53 1.29 1.82 <0.0001 96 11 188,841 13,113,564 1.64 1.26 2.14 <0.0001 90
171
172
Table 4 Meta-analysis of longitudinal studies with publication bias assessment (continued)
Meta-analysis of unadjusted relative risk Meta-analysis of covariate adjusted hazard ratio
Hetero- Hetero-
No. participants Unadjusted relative risk No. participants Adjusted hazard ratio
Disorder studies Patients Controls RR 95% CI p I2 studies Patients Controls HR 95% CI p I2
Congestive heart failure
Bipolar disorder 1 6,215 2,411,852 11.52 9.37 23.14 <0.0001 - 1 58 1,339 2.27 1.49 3.45 <0.0001 0
Major depressive disorder - - - - - - - - 2 351 40,339 2.02 1.48 2.75 <0.0001 0
Schizophrenia 3 85,290 9,050,272 1.80 1.15 2.79 0.009 84 - - - - - - - -

d
Pooled 4 91,505 11,459,059 8.24 6.84 9.94 <0.0001 99 2 409 41,678 2.10 1.64 2.70 <0.0001 0
Peripheral vascular disease
Bipolar disorder 1 6,215 2,411,852 3.44 2.70 4.38 <0.0001 - - - - - - - - -
Major depressive disorder - - - - - - - - - - - - - - - -
Schizophrenia 3 85,290 9,050,272 0.96 0.43 2.17 0.92 93 - - - - - - - -

e
Pooled 3 91,505 11,402,868 3.11 2.46 3.91 <0.0001 98 - - - - - - - -
Death due to cardiovascular disease
Bipolar disorder 5 37,144 356,298 1.31 0.94 1.83 0.11 75 3 17,420 162,231 1.65 1.10 2.47 0.02 88
Schizophrenia 9 53,779 7,179,454 1.26 0.84 1.90 0.27 96 9 152,690 6,872,808 2.45 1.64 3.65 <0.0001 96
Severe mental illnesses 3 874,146 52,714,134 2.99 2.84 3.13 <0.0001 0 3 798 31,724 2.75 1.32 5.73 0.007 75
f j
Pooled 18 1,151,181 60,287,400 2.89 2.75 3.03 <0.0001 99 16 353,407 7,317,053 1.85 1.53 2.24 <0.0001 95
Egger test for bias: a20.44, p50.80; b21.24, p50.71; c0.03, p50.96; d8.07, p50.37; e3.08, p50.60; f23.66, p50.20; g1.16, p50.31; h20.13, p50.92; i2.57, p50.07; j21.19, p50.43

Table 5 Quality assessment of included studies
Clear diagnostic Control Matched Coavariate adjusted Reported cardiovascular Follow-up at

Study criteria group controls analyses risk factors at baseline least 5 years
Cross-sectional studies
Beyer et al22 Y N N N Y N
Bresee et al23 Y Y N Y Y N
Bresee et al24 Y Y N Y Y N
Chen et al25 Y N N N Y N
26
Curkendall et al Y Y Y Y Y N
Devantier et al27 Y Y Y N Y N
Hagg et al28 Y N N N Y N
Herbst et al29 Y Y N Y N N
30
Huang et al Y Y N Y Y N
Hyde et al31 Y N N N Y N
32
Kilbourne et al Y N N N Y N
Kilbourne et al33 Y Y N Y Y N
Lindegard34 Y Y N N N N
Maina et al35 Y N N N Y N
Morden et al36 Y Y Y Y Y N
Munoli et al37 Y N N N Y N
Nielsen et al38 Y N N N Y N
Niranjan et al39 Y Y N Y Y N
Oreski et al40 Y Y N N Y N
Prieto et al41 Y N N N Y N
42
Scherrer et al Y Y N N N N
Scott et al43 Y Y N Y N N
Shen et al44 Y Y Y Y Y N
Smith et al45 Y Y N Y Y N
Smith et al46 Y Y N Y Y N
Swain et al47 Y Y N Y N N
48
Zilkens et al Y Y Y N N N
Longitudinal studies
Almeida et al49 Y Y Y N Y Y
50
Bremmer et al Y Y Y N Y Y
Butnoriene et al51 Y Y Y N N Y
52
Callaghan et al Y Y Y Y Y N
Callaghan et al53 Y Y Y Y Y N
54
Carney et al Y Y Y N Y N
Chen et al55 Y Y Y Y Y Y
56
Clouse et al Y Y Y N Y Y
Coryell et al57 Y Y N N N Y
58
Crump et al Y Y Y N N Y
Crump et al59 Y Y Y N Y Y
Davis et al60 Y Y Y N Y N
Davydow et al61 Y Y Y N N Y

Table 5 Quality assessment of included studies (continued)
Enger et al62 Y Y Y Y Y N
Fiedorowicz et al63 Y Y Y N Y Y
64
Filik et al Y Y Y N Y N
Fors et al65 Y Y Y Y N Y
66
Gasse et al Y Y Y N N Y
Goldstein et al67 Y Y Y N Y N
Healy et al68 Y Y N N N Y
Hendrie et al69 Y Y Y N Y Y
Hou et al70 Y Y N N N Y
Hsieh et al71 Y Y N N N N
Huang et al72 Y Y Y Y Y Y
Ifteni et al73 Y Y N N N Y
Jakobsen et al74 Y Y Y Y N Y
Janszky et al75 Y Y Y N Y Y
Jokinen & Nordstrom et al76 Y Y N N N Y
Joukamaa et al77 Y Y Y N N Y
78
Kendler et al Y Y Y N N Y
Kiviniemi et al79 Y Y N N N Y
80
Lahti et al Y Y Y N Y Y
Lan et al81 Y Y N N Y Y
82
Laursen et al Y Y Y N N Y
Laursen et al83 Y Y Y N N Y
84
Lemogne et al Y Y Y N Y Y
Li et al85 Y Y Y Y Y Y
86
Lin et al Y Y Y Y Y Y
Lin et al87 Y Y Y Y Y Y
88
Lin et al Y Y Y Y Y Y
Maina et al89 Y Y N N N Y
McDermott et al90 Y Y Y N N Y
Murray-Thomas et al91 Y Y Y N N N
Olfson et al92 Y Y N N N Y
Osborn et al93 Y Y Y N Y Y
Pratt et al94 Y Y Y N Y Y
Prieto et al95 Y Y Y Y Y Y
96
Rahman et al Y Y Y Y Y N
Ramsey et al97 Y Y Y N Y Y
98
Saint Onge et al Y Y Y N Y Y
Scherrer et al99 Y Y Y N Y Y
100
Schoepf & Heun Y Y Y Y N Y
Schoepf et al101 Y Y Y Y Y Y
102
Shah et al Y Y Y N Y Y
Stewart et al103 Y Y N N N Y
104
Surtees et al Y Y Y N N Y

Table 5 Quality assessment of included studies (continued)
Ting et al105 Y Y Y N Y Y
Torniainen et al106 Y Y Y Y N Y
107
Tsai et al Y Y Y Y Y Y
Tsan et al108 Y Y N N Y Y
van Marwijk et al109 Y Y Y Y Y N
Weeke et al110 Y Y N N N N
111
Westman et al Y Y Y N N Y
Wu et al112 Y Y Y Y Y Y
113
Wu et al Y Y Y N Y Y
N no, Y yes
Table 6 Prevalence and incidence of cardiovascular disease (CVD) in severe mental illness stratified by region
Risk ratios
Prevalence Incidence for incident Adjusted hazard ratios
Regional strata Analysis details of CVD of CVD CVD for incident CVD
Asia Pooled estimate, % (95% CI) 5.4 (4.3-6.7) 2.6 (1.9-3.6) 1.63 (1.31-2.04) 1.75 (1.38-2.22)
p value <0.0001 <0.0001 <0.0001 <0.0001
Heterogeneity, I2 (p value) 98 (<0.0001) 100 (<0.0001) 99 (<0.0001) 96 (<0.0001)
No. comparisons 8 12 9 10
Europe Pooled estimate, % (95% CI) 9.7 (6.5-14.2) 3.4 (2.2-5.3) 1.17 (0.96-1.42) 1.88 (1.44-2.46)
p value <0.0001 <0.0001 0.11 <0.0001
Heterogeneity, I2 (p value) 97 (<0.0001) 100 (<0.0001) 97 (<0.0001) 96 (<0.0001)
North America Pooled estimate, % (95% CI) 14.6 (12.0-17.7) 4.6 (3.4-6.2) 1.39 (0.91-2.12) 1.88 (1.62-2.19)
p value <0.0001 <0.0001 0.13 <0.0001
Heterogeneity, I2 (p value) 97 (<0.0001) 100 (<0.0001) 97 (<0.0001) 62 (0.003)
Oceania Pooled estimate, % (95% CI) 20.6 (10.9-35.4) 26.3 (24.1-28.6) 1.52 (1.40-1.66) 1.58 (1.41-1.78)
p value <0.0001 <0.0001 <0.0001 <0.0001
Heterogeneity, I2 (p value) 97 (<0.0001) 100 (<0.0001) 0 (0.72) 0 (0.84)
p (difference between regions) <0.0001 <0.0001 0.08 0.29
had a control group, which was matched in all but 12 studies, Oceania (Asia: 5.4% and 2.6%; Europe: 9.7% and 3.4%; North
only 19 studies adjusted for covariates, 38 studies reported on America: 14.6% and 4.6%; Oceania: 20.6% and 26.3%; p<0.0001
cardiovascular risk factors, and all except 12 studies had a for both prevalence and incidence). However, when comparing
follow-up of at least 5 years. CVD risk in SMI patients in each region with their respective
control groups, there was no statistically significant difference
anymore across regions, with both RRs and adjusted HRs show-
Regional CVD prevalence, incidence and ing comparably increased CVD incidence risk in the SMI popu-
longitudinal risk lation (RRs ranging from 1.17 in Europe to 1.63 in Asia, p50.08;
and HRs ranging from 1.58 in Oceania to 1.88 in both Europe
Raw CVD prevalence and incidence rates consistently in- and North America, p50.29) (Table 6). There were insufficient
creased from Asia, through Europe and North America, to numbers of studies to perform this analysis for adjusted ORs

regarding prevalence rates across regions, or for adjusted ORs, use, elevated body mass index and elevated baseline CVD.
RRs or HRs pertaining to specific CVD subgroups. Based on these results, it is imperative that clinicians: a) only
utilize antipsychotics, particularly for non-psychotic condi-
tions, when alternative treatment options with lower CVD risk
Meta-regression potential have been tried sufficiently; and b) screen for and
manage emerging and existing CVDs as well as their risk fac-
Due to heterogeneous or partial reporting of possible mod- tors, including weight gain and elevated body mass index. Our
erator variables in the included studies, all meta-regression data, adding to research demonstrating a significantly higher
analyses were based on a much reduced number of studies. prevalence of metabolic syndrome in people with SMI com-
Hence, all analyses were less powered in comparison with the pared to controls114, clearly suggest there is an urgent need to
large sets of data used for cross-sectional prevalence and lon- prevent and manage CVD risk in this population.
gitudinal incidence analyses. Nonetheless, CVD incidence Our results demonstrating a higher CVD prevalence in SMI
increased significantly with a higher percentage of patients populations versus controls in more recent studies are also
using antipsychotics (12 studies; b50.04, 95% CI: 0.01-0.08, concerning, as they support accumulating data indicating that
p50.008), higher baseline body mass index in patients vs. con- secondary prevention has been much less successful in the
trols (6 studies; b50.24, 95% CI: 0.06-0.42, p50.008), and SMI population that in the general population, leading to a
higher CVD prevalence at baseline in patients vs. controls (7 widening of the mortality gap in recent years49,115,116. Our
studies; b50.07, 95% CI: 0.01-0.14, p50.03). CVD prevalence findings confirm prior reports that antipsychotic medication
increased in more recent studies (38 studies; b50.07, 95% CI: use is associated with higher CVD risk13,117,118. However, due
0.02-0.12, p50.007), whereas the same was not true for CVD
to limitations in the published data, we were unable to explore
incidence (65 studies; b50.02, 95% CI50.07 to 0.01, p50.21).
variations in CVD risk profiles between different antipsychotic
medications13,117-120. Previous research has suggested that the
highest cardio-metabolic risks are associated with clozapine
DISCUSSION and olanzapine, whilst the lowest risk is with aripiprazole,
ziprasidone, lurasidone, amisulpride and high potency typical
To our knowledge, this is the first large scale meta-analysis antipsychotics13,117-122. However, in this context it is also
providing comprehensive quantitative data on the prevalence important to note that antipsychotic medications can decrease
and incidence of CVD in people with SMI, including both CVD-related mortality, as reported for example in Finnish79
pooled data and comparisons across CVD and SMI diagnostic and Swedish123 national database studies, that are highly gen-
subgroups. Our results establish that approximately 10% of eralizable. These data underscore that symptom control and
people with SMI with a mean age of 50 years have at least one functional improvement benefit both psychiatric and overall
comorbid CVD. Moreover, our longitudinal analysis docu- health, as severe psychiatric illness negatively affects lifestyle
ments a 3.6% incidence rate of CVD during a median of 8.4 behaviors, medical care seeking and adherence to medical
years of follow-up. Patients with SMI show a 53% higher risk treatments. Thus, benefits of improved psychiatric status with
for having CVD, a 78% higher risk for developing CVD, and an antipsychotics and other psychotropic agents need to be care-
85% higher risk of death from CVD compared to the regionally fully weighed against their potential for elevated cardiometa-
matched general population. bolic risk, which differs across available agents13,117.
This study provides a worldwide epidemiologic representa- Since antipsychotic medication use moderates CVD risk and
tion of CVD prevalence and incidence rates in SMI, reporting since antipsychotics are increasingly used as first line treatments
the lowest absolute prevalence and incidence in Asia, increas- for much more prevalent non-psychotic conditions, including
ing through Europe and North America, and reaching the bipolar disorder124 and major depressive disorder with subopti-
highest levels in Oceania. However, in analyses with sufficient mal response to antidepressant treatment125, the pool of people
numbers of available studies, neither RRs nor adjusted HRs at an increased CVD risk is greatly enlarged. Therefore, research
indicated significantly different CVD incidence risk across on the underlying mechanisms for the increased CVD risk after
regions, meaning that SMI patients are at an increased risk pharmacotherapy initiation is even more urgently needed to
across the world and that CVD risk-reducing interventions in develop more effective and targeted preventive and intervention-
SMI are needed with the same urgency across all regions of al treatments. Studies should also examine whether different
the world. Moreover, while the prevalence and incidence of clinical subtypes of depression (i.e., melancholic, psychotic,
each CVD in people with SMI show some minor variations, peo- atypical or undifferentiated) and bipolar disorder (e.g., type 1 or
ple with major depressive disorder, bipolar disorder and schizo- 2, cyclothymic disorder), certain mood states (manic, depressive,
phrenia are clearly all at an increased risk of CVD-related deaths mixed or euthymic), or different antipsychotics, antidepressants
compared to population-stratified controls, calling for urgent or mood stabilizers13 significantly moderate CVD risk.
action. Furthermore, the pathophysiology underlying the associa-
We were able to identify some important and actionable tion between SMI and CVD risk is complex and not well
moderators of increased CVD risk, including antipsychotic understood, clearly requiring further investigation. Emerging

evidence suggests that SMI and CVD share pathophysiological adjustment for potential publication bias with the trim and fill
features, including hypothalamic-pituitary-adrenal and mito- analysis.
chondrial dysfunction, peripheral immune activation, neuro- In conclusion, SMIs pooled together were significantly
inflammation, oxidative and nitrosative stress, as well as com- associated in cross-sectional studies with CVD, coronary heart
mon genetic links and epigenetic interactions126. However, disease, cerebrovascular disease and CVD-related death. Addi-
since these different mechanisms probably interact, research tionally, in longitudinal studies, each specific diagnostic SMI
that integrates these pathways is urgently needed. Beyond group was significantly associated with CVD and CVD-related
mechanistic evaluations, such studies also need to investigate death. Furthermore, schizophrenia was associated with coro-
the general and specific effects of physical health improve- nary heart disease and cerebrovascular disease, while bipolar dis-
ments on SMI outcomes. order was associated with congestive heart failure, and major
Future research should also investigate optimal monitoring depressive disorder was associated with coronary heart disease,
regimens across stratified patient subgroups as well as the most cerebrovascular disease, and congestive heart failure.
effective timing and efficacy of primary, secondary and tertiary Importantly, our data confirm that CVDs are associated with
preventive interventions120,127. In this regard, studies should an increased risk of mortality in people with SMI, which to a
comprehensively assess relevant moderator and mediator varia- large part explains the shortened life expectancy of people with
bles of CVD risk, including type and duration of specific psy- SMI compared to the general population2,4,5. Furthermore, we
chotropic medications use, physical activity (including using showed geographical variations in raw CVD prevalence and inci-
passive monitoring via actimetry or mobile phone technology), dence risk in SMI populations, but no significant regional vari-
diet, smoking, body mass index, personal and family history of ance in the difference in CVD risk compared to the region-
CVD, in order to identify subgroups of patients who may require specific general population. Finally, the fact that antipsychotic
different monitoring and or interventions schemes. Long-term use, higher body mass index and baseline CVD significantly
follow-up studies are also required to accurately document the increased the risk for CVD morbidity and mortality underscores
emergence of more distal physical and mental health as well as the urgent need to limit antipsychotic use to those populations
health economic outcomes in relationship to the early identifi- truly requiring them, choosing the lowest risk antipsychotic
cation and management of CVD risk factors and manifest CVD agents first in the treatment algorithm, screening all SMI patients
conditions in people with SMI. regularly for CVD risk factors and conditions, and addressing any
Finally, since people with SMI engage in unhealthy lifestyle identified abnormalities aggressively.
and often take psychotropic medication for extensive periods,
long-term follow-up studies are needed that assess whether
current predictor models based on the magnitude of tradition- ACKNOWLEDGEMENTS
al CVD risk factor effects observed in the general population B. Stubbs and F. Gaughran receive support from the National Institute for Health
apply or need to be adjusted for the SMI population93, in Research (NIHR) Collaboration for Leadership in Applied Health Research and
Care South London at Kings College Hospital NHS Foundation Trust. F. Gaughran
whom CVD risk factors also emerge at a far earlier age117,128. is also funded by the National Institute for Health Research Collaboration for
While this is the most comprehensive meta-analysis of CVD Leadership in Applied Health Research and Care Funding scheme and by the
Stanley Medical Research Institute. The views expressed in this publication are
risk in people with SMI conducted to date, we acknowledge those of the authors and not necessarily those of the funding institutions. C.U.
several limitations that are largely related to factors in the pri- Correll, M. Solmi and N. Veronese are joint first authors of the paper.
mary data. First, lifestyle behavior information (e.g., physical

activity) was inadequately reported, precluding meta-analytic REFERENCES
assessment of these important factors as moderating or medi-
1. Chang CK, Hayes RD, Perera G et al. Life expectancy at birth for people
ating variables. People with SMI are less likely than the general with serious mental illness and other major disorders from a secondary
population to engage in physical activity and have higher lev- mental health care case register in London. PLoS One 2011;6:e19590.
2. Lawrence D, Hancock KJ, Kisely S. The gap in life expectancy from pre-
els of sedentary behaviour129, smoke more130, consume diets
ventable physical illness in psychiatric patients in Western Australia: ret-
that are high in saturated fats and refined sugars, while being rospective analysis of population based registers. BMJ 2013;346:f2539.
low in fruit and vegetables131, all factors relevant for CVD risk. 3. Popovic D, Benabarre A, Crespo JM et al. Risk factors for suicide in
schizophrenia: systematic review and clinical recommendations. Acta
Second, variables such as clinical subtypes of major depressive Psychiatr Scand 2014;130:418-26.
disorder and bipolar disorder, negative symptom severity in 4. Hoang U, Goldacre MJ, Stewart R. Avoidable mortality in people with
people with schizophrenia, and concomitant or previous use schizophrenia or bipolar disorder in England. Acta Psychiatr Scand 2013;
127:195-201.
of specific antipsychotics, antidepressants and mood stabil- 5. Walker ER, McGee RE, Druss BG. Mortality in mental disorders and global
izers were not reported or were insufficiently reported or con- disease burden implications: a systematic review and meta-analysis.
trolled for in most available studies. Third, as expected when JAMA Psychiatry 2015;72:334-41.
6. Wu Q, Kling JM. Depression and the risk of myocardial infarction and
combining observational data132, many of the results were coronary death: a meta-analysis of prospective cohort studies. Medicine
moderately to highly heterogeneous. However, in accordance 2016;95:e2815.
with the MOOSE guidelines133, we conducted meta-regression 7. Gan Y, Gong Y, Tong X et al. Depression and the risk of coronary heart disease:
a meta-analysis of prospective cohort studies. BMC Psychiatry 2014;14:371.
analyses and were able to explain some of the observed het- 8. Dong JY, Zhang YH, Tong J et al. Depression and risk of stroke: a meta-
erogeneity. In addition, all of our results remained robust after analysis of prospective studies. Stroke 2012;43:32-7.

9. Van der Kooy K, van Hout H, Marwijk H et al. Depression and the risk for 36. Morden NE, Lai Z, Goodrich DE et al. Eight-year trends of cardiometa-
cardiovascular diseases: systematic review and meta analysis. Int J Geri- bolic morbidity and mortality in patients with schizophrenia. Gen Hosp
atr Psychiatry 2007;22:613-26. Psychiatry 2012;34:368-79.
10. Prieto ML, Cuellar-Barboza AB, Bobo WV et al. Risk of myocardial infarc- 37. Munoli RN, Praharaj SK, Sharma PS. Co-morbidity in bipolar disorder: a
tion and stroke in bipolar disorder: a systematic review and exploratory retrospective study. Indian J Psychol Med 2014;36:270-5.
meta-analysis. Acta Psychiatr Scand 2014;130:342-53. 38. Nielsen J, Juel J, Alzuhairi KS et al. Unrecognised myocardial infarction in
11. Fan Z, Wu Y, Shen J et al. Schizophrenia and the risk of cardiovascular patients with schizophrenia. Acta Neuropsychiatr 2015;27:106-12.
diseases: a meta-analysis of thirteen cohort studies. J Psychiatr Res 2013; 39. Niranjan A, Corujo A, Ziegelstein RC et al. Depression and heart disease
47:1549-56. in US adults. Gen Hosp Psychiatry 2012;34:254-61.
12. Li M, Fan YL, Tang ZY et al. Schizophrenia and risk of stroke: a meta- 40. Oreski I, Jakovljevic M, Aukst-Margetic B et al. Comorbidity and multi-
analysis of cohort studies. Int J Cardiol 2014;173:588-90. morbidity in patients with schizophrenia and bipolar disorder: similari-
13. Correll CU, Detraux J, De Lepeleire J et al. Effects of antipsychotics, anti- ties and differencies. Psychiatr Danub 2012;24:80-5.
depressants and mood stabilizers on risk for physical diseases in people 41. Prieto ML, McElroy SL, Hayes SN et al. Association between history of
with schizophrenia, depression and bipolar disorder. World Psychiatry psychosis and cardiovascular disease in bipolar disorder. Bipolar Disord
2015;14:119-36. 2015;17:518-27.
14. Moher D, Liberati A, Tetzlaff J et al. Preferred reporting items for system- 42. Scherrer JF, Xian H, Bucholz KK et al. A twin study of depression symp-
atic reviews and meta-analyses: the PRISMA statement. Int J Surg 2010;8: toms, hypertension, and heart disease in middle-aged men. Psychosom
336-41. Med 2003;65:548-57.
15. Biostat. Comprehensive Meta-Analysis. https://www.meta-analysis.com. 43. Scott KM, de Jonge P, Alonso J et al. Associations between DSM-IV mental
16. DerSimonian R, Kacker R. Random-effects model for meta-analysis of disorders and subsequent heart disease onset: beyond depression. Int J
clinical trials: an update. Contemp Clin Trials 2007;28:105-14. Cardiol 2013;168:5293-9.
17. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin 44. Shen HN, Lu CL, Yang HH. Increased risks of acute organ dysfunction
Trials 1986;7:177-88. and mortality in intensive care unit patients with schizophrenia: a nation-
18. Egger M, Davey Smith G, Schneider M et al. Bias in meta-analysis wide population-based study. Psychosom Med 2011;73:620-6.
detected by a simple, graphical test. BMJ 1997;315:629-34. 45. Smith DJ, Langan J, McLean G et al. Schizophrenia is associated with
19. Begg CB, Mazumdar M. Operating characteristics of a rank correlation excess multiple physical-health comorbidities but low levels of recorded
test for publication bias. Biometrics 1994;50:1088-101. cardiovascular disease in primary care: cross-sectional study. BMJ Open
20. Duval S, Tweedie R. Trim and fill: a simple funnel-plot-based method of 2013;3.
testing and adjusting for publication bias in meta-analysis. Biometrics 46. Smith DJ, Martin D, McLean G et al. Multimorbidity in bipolar disorder
2000;56:455-63. and undertreatment of cardiovascular disease: a cross sectional study.
21. Higgins JP, Thompson SG, Deeks JJ et al. Measuring inconsistency in BMC Med 2013;11:263.
meta-analyses. BMJ 2003;327:557-60. 47. Swain NR, Lim CC, Levinson D et al. Associations between DSM-IV men-
22. Beyer J, Kuchibhatla M, Gersing K et al. Medical comorbidity in a bipolar tal disorders and subsequent non-fatal, self-reported stroke. J Psychosom
outpatient clinical population. Neuropsychopharmacology 2005;30:401-4. Res 2015;79:130-6.
23. Bresee LC, Majumdar SR, Patten SB et al. Prevalence of cardiovascular 48. Zilkens RR, Bruce DG, Duke J et al. Severe psychiatric disorders in mid-
risk factors and disease in people with schizophrenia: a population- life and risk of dementia in late-life (age 65-84 years): a population based
based study. Schizophr Res 2010;117:75-82. case-control study. Curr Alzheimer Res 2014;11:681-93.
24. Bresee LC, Majumdar SR, Patten SB et al. Diabetes, cardiovascular dis- 49. Almeida OP, Hankey GJ, Yeap BB et al. Mortality among people with
ease, and health care use in people with and without schizophrenia. Eur severe mental disorders who reach old age: a longitudinal study of a
Psychiatry 2011;26:327-32. community-representative sample of 37,892 men. PLoS One 2014;9:
25. Chen PH, Gildengers AG, Lee CH et al. High serum sodium level in affec- e111882.
tive episode associated with coronary heart disease in old adults with 50. Bremmer MA, Hoogendijk WJ, Deeg DJ et al. Depression in older age is a
bipolar disorder. Int J Psychiatry Med 2015;50:422-33. risk factor for first ischemic cardiac events. Am J Geriatr Psychiatry 2006;
26. Curkendall SM, Mo J, Glasser DB et al. Cardiovascular disease in patients 14:523-30.
with schizophrenia in Saskatchewan, Canada. J Clin Psychiatry 2004;65: 51. Butnoriene J, Bunevicius A, Saudargiene A et al. Metabolic syndrome,
715-20. major depression, generalized anxiety disorder, and ten-year all-cause
27. Devantier TA, Norgaard BL, Ovrehus KA et al. Coronary plaque volume and cardiovascular mortality in middle aged and elderly patients. Int J
and composition assessed by computed tomography angiography in Cardiol 2015;190:360-6.
patients with late-onset major depression. Psychosomatics 2014;55:243- 52. Callaghan RC, Boire MD, Lazo RG et al. Schizophrenia and the incidence
51. of cardiovascular morbidity: a population-based longitudinal study in
28. Hagg S, Lindblom Y, Mjorndal T et al. High prevalence of the metabolic Ontario, Canada. Schizophr Res 2009;115:325-32.
syndrome among a Swedish cohort of patients with schizophrenia. Int 53. Callaghan RC, Khizar A. The incidence of cardiovascular morbidity
Clin Psychopharmacol 2006;21:93-8. among patients with bipolar disorder: a population-based longitudinal
29. Herbst S, Pietrzak RH, Wagner J et al. Lifetime major depression is associ- study in Ontario, Canada. J Affect Disord 2010;122:118-23.
ated with coronary heart disease in older adults: results from the National 54. Carney CP, Jones L, Woolson RF. Medical comorbidity in women and men
Epidemiologic Survey on Alcohol and Related Conditions. Psychosom with schizophrenia: a population-based controlled study. J Gen Intern
Med 2007;69:729-34. Med 2006;21:1133-7.
30. Huang KL, Su TP, Chen TJ et al. Comorbidity of cardiovascular diseases 55. Chen MH, Li CT, Hsu JW et al. Atopic diseases and subsequent ischemic
with mood and anxiety disorder: a population based 4-year study. Psychi- stroke among patients with schizophrenia: a nationwide longitudinal
atry Clin Neurosci 2009;63:401-9. study. Psychosom Med 2015;77:1031-8.
31. Hyde N, Dodd S, Venugopal K et al. Prevalence of cardiovascular and 56. Clouse RE, Lustman PJ, Freedland KE et al. Depression and coronary
metabolic events in patients prescribed clozapine: a retrospective obser- heart disease in women with diabetes. Psychosom Med 2003;65:376-83.
vational, clinical cohort study. Curr Drug Saf 2015;10:125-31. 57. Coryell W, Turvey C, Leon A et al. Persistence of depressive symptoms
32. Kilbourne AM, Cornelius JR, Han X et al. General-medical conditions in and cardiovascular death among patients with affective disorder. Psycho-
older patients with serious mental illness. Am J Geriatr Psychiatry 2005; som Med 1999;61:755-61.
13:250-4. 58. Crump C, Sundquist K, Winkleby MA et al. Comorbidities and mortality
33. Kilbourne AM, Post EP, Bauer MS et al. Therapeutic drug and cardiovas- in bipolar disorder: a Swedish national cohort study. JAMA Psychiatry
cular disease risk monitoring in patients with bipolar disorder. J Affect 2013;70:931-9.
Disord 2007;102:145-51. 59. Crump C, Winkleby MA, Sundquist K et al. Comorbidities and mortality
34. Lindegard B. Physical illness in severe depressives and psychiatric alco- in persons with schizophrenia: a Swedish national cohort study. Am J
holics in Gothenburg, Sweden. J Affect Disord 1982;4:383-93. Psychiatry 2013;170:324-33.
35. Maina G, DAmbrosio V, Aguglia A et al. Bipolar disorders and metabolic 60. Davis J, Fujimoto RY, Juarez DT et al. Major depression associated with rates
syndrome: a clinical study in 185 patients. Riv Psichiatr 2010;45:34-40. of cardiovascular disease state transitions. Am J Manag Care 2008;14:125-8.

61. Davydow DS, Ribe AR, Pedersen HS et al. Serious mental illness and risk 85. Li CT, Bai YM, Tu PC et al. Major depressive disorder and stroke risks: a 9-
for hospitalizations and rehospitalizations for ambulatory care-sensitive year follow-up population-based, matched cohort study. PLoS One 2012;
conditions in Denmark: a nationwide population-based cohort study. 7:e46818.
Med Care 2016;54:90-7. 86. Lin HC, Chen YH, Lee HC et al. Increased risk of acute myocardial infarc-
62. Enger C, Weatherby L, Reynolds RF et al. Serious cardiovascular events tion after acute episode of schizophrenia: 6 year follow-up study. Aust N
and mortality among patients with schizophrenia. J Nerv Ment Dis 2004; Z J Psychiatry 2010;44:273-9.
192:19-27. 87. Lin HC, Tsai SY, Lee HC. Increased risk of developing stroke among
63. Fiedorowicz JG, Solomon DA, Endicott J et al. Manic/hypomanic symp- patients with bipolar disorder after an acute mood episode: a six-year fol-
tom burden and cardiovascular mortality in bipolar disorder. Psychosom low-up study. J Affect Disord 2007;100:49-54.
Med 2009;71:598-606. 88. Lin HC, Hsiao FH, Pfeiffer S et al. An increased risk of stroke among
64. Filik R, Sipos A, Kehoe PG et al. The cardiovascular and respiratory young schizophrenia patients. Schizophr Res 2008;101:234-41.
health of people with schizophrenia. Acta Psychiatr Scand 2006;113:298- 89. Maina G, Bechon E, Rigardetto S et al. General medical conditions are
305. associated with delay to treatment in patients with bipolar disorder. Psy-
65. Fors BM, Isacson D, Bingefors K et al. Mortality among persons with chosomatics 2013;54:437-42.
schizophrenia in Sweden: an epidemiological study. Nord J Psychiatry 90. McDermott S, Moran R, Platt T et al. Heart disease, schizophrenia, and
2007;61:252-9. affective psychoses: epidemiology of risk in primary care. Community
66. Gasse C, Laursen TM, Baune BT. Major depression and first-time hospi- Ment Health J 2005;41:747-55.
talization with ischemic heart disease, cardiac procedures and mortality 91. Murray-Thomas T, Jones ME, Patel D et al. Risk of mortality (including
in the general population: a retrospective Danish population-based sudden cardiac death) and major cardiovascular events in atypical and
cohort study. Eur J Prev Cardiol 2014;21:532-40. typical antipsychotic users: a study with the general practice research
67. Goldstein BI, Schaffer A, Wang S et al. Excessive and premature new- database. Cardiovasc Psychiatry Neurol 2013;2013:247486.
onset cardiovascular disease among adults with bipolar disorder in the 92. Olfson M, Gerhard T, Huang C et al. Premature mortality among
US NESARC cohort. J Clin Psychiatry 2015;76:163-9. adults with schizophrenia in the United States. JAMA Psychiatry 2015;72:
68. Healy D, Le Noury J, Harris M et al. Mortality in schizophrenia and relat- 1172-81.
ed psychoses: data from two cohorts, 1875-1924 and 1994-2010. BMJ 93. Osborn DP, Hardoon S, Omar RZ et al. Cardiovascular risk prediction
Open 2012;2. models for people with severe mental illness: results from the prediction
69. Hendrie HC, Tu W, Tabbey R et al. Health outcomes and cost of care and management of cardiovascular risk in people with severe mental ill-
among older adults with schizophrenia: a 10-year study using medical nesses (PRIMROSE) research program. JAMA Psychiatry 2015;72:143-51.
records across the continuum of care. Am J Geriatr Psychiatry 2014;22: 94. Pratt LA, Ford DE, Crum RM et al. Depression, psychotropic medication,
427-36. and risk of myocardial infarction. Prospective data from the Baltimore
70. Hou PY, Hung GC, Jhong JR et al. Risk factors for sudden cardiac death ECA follow-up. Circulation 1996;94:3123-9.
among patients with schizophrenia. Schizophr Res 2015;168:395-401. 95. Prieto ML, Schenck LA, Kruse JL et al. Long-term risk of myocardial
71. Hsieh PH, Hsiao FY, Gau SS et al. Use of antipsychotics and risk of cere- infarction and stroke in bipolar I disorder: a population-based Cohort
brovascular events in schizophrenic patients: a nested case-control study. Study. J Affect Disord 2016;194:120-7.
J Clin Psychopharmacol 2013;33:299-305. 96. Rahman I, Humphreys K, Bennet AM et al. Clinical depression, antide-
72. Huang CJ, Hsieh MH, Hou WH et al. Depression, antidepressants, and pressant use and risk of future cardiovascular disease. Eur J Epidemiol
the risk of coronary heart disease: a population-based cohort study. Int J 2013;28:589-95.
Cardiol 2013;168:4711-6. 97. Ramsey CM, Leoutsakos JM, Mayer LS et al. History of manic and hypo-
73. Ifteni P, Correll CU, Burtea V et al. Sudden unexpected death in schizo- manic episodes and risk of incident cardiovascular disease: 11.5 year
phrenia: autopsy findings in psychiatric inpatients. Schizophr Res 2014; follow-up from the Baltimore Epidemiologic Catchment Area Study.
155:72-6. J Affect Disord 2010;125:35-41.
74. Jakobsen AH, Foldager L, Parker G et al. Quantifying links between acute 98. Saint Onge JM, Krueger PM, Rogers RG. The relationship between major
myocardial infarction and depression, anxiety and schizophrenia using depression and nonsuicide mortality for U.S. adults: the importance of
case register databases. J Affect Disord 2008;109:177-81. health behaviors. J Gerontol B Psychol Sci Soc Sci 2014;69:622-32.
75. Janszky I, Ahnve S, Lundberg I et al. Early-onset depression, anxiety, and 99. Scherrer JF, Garfield LD, Chrusciel T et al. Increased risk of myocardial
risk of subsequent coronary heart disease: 37-year follow-up of 49,321 infarction in depressed patients with type 2 diabetes. Diabetes Care 2011;
young Swedish men. J Am Coll Cardiol 2010;56:31-7. 34:1729-34.
76. Jokinen J, Nordstrom P. HPA axis hyperactivity and cardiovascular mor- 100. Schoepf D, Heun R. Bipolar disorder and comorbidity: increased preva-
tality in mood disorder inpatients. J Affect Disord 2009;116:88-92. lence and increased relevance of comorbidity for hospital-based mortali-
77. Joukamaa M, Heliovaara M, Knekt P et al. Mental disorders and cause- ty during a 12.5-year observation period in general hospital admissions.
specific mortality. Br J Psychiatry 2001;179:498-502. J Affect Disord 2014;169:170-8.
78. Kendler KS, Gardner CO, Fiske A et al. Major depression and coronary 101. Schoepf D, Uppal H, Potluri R et al. Physical comorbidity and its rele-
artery disease in the Swedish twin registry: phenotypic, genetic, and vance on mortality in schizophrenia: a naturalistic 12-year follow-up in
environmental sources of comorbidity. Arch Gen Psychiatry 2009;66:857- general hospital admissions. Eur Arch Psychiatry Clin Neurosci 2014;264:
63. 3-28.
79. Kiviniemi M, Suvisaari J, Koivumaa-Honkanen H et al. Antipsychotics 102. Shah AJ, Veledar E, Hong Y et al. Depression and history of attempted sui-
and mortality in first-onset schizophrenia: prospective Finnish register cide as risk factors for heart disease mortality in young individuals. Arch
study with 5-year follow-up. Schizophr Res 2013;150:274-80. Gen Psychiatry 2011;68:1135-42.
80. Lahti M, Tiihonen J, Wildgust H et al. Cardiovascular morbidity, mortality 103. Stewart JC, Perkins AJ, Callahan CM. Effect of collaborative care for
and pharmacotherapy in patients with schizophrenia. Psychol Med 2012; depression on risk of cardiovascular events: data from the IMPACT ran-
42:2275-85. domized controlled trial. Psychosom Med 2014;76:29-37.
81. Lan CC, Liu CC, Lin CH et al. A reduced risk of stroke with lithium expo- 104. Surtees PG, Wainwright NW, Luben RN et al. Depression and ischemic
sure in bipolar disorder: a population-based retrospective cohort study. heart disease mortality: evidence from the EPIC-Norfolk United Kingdom
Bipolar Disord 2015;17:705-14. prospective cohort study. Am J Psychiatry 2008;165:515-23.
82. Laursen TM, Munk-Olsen T, Gasse C. Chronic somatic comorbidity and 105. Ting RZ, Lau ES, Ozaki R et al. High risk for cardiovascular disease in Chi-
excess mortality due to natural causes in persons with schizophrenia or nese type 2 diabetic patients with major depressiona 7-year prospective
bipolar affective disorder. PLoS One 2011;6:e24597. analysis of the Hong Kong Diabetes Registry. J Affect Disord 2013;149:
83. Laursen TM, Mortensen PB, MacCabe JH et al. Cardiovascular drug use 129-35.
and mortality in patients with schizophrenia or bipolar disorder: a Danish 106. Torniainen M, Mittendorfer-Rutz E, Tanskanen A et al. Antipsychotic
population-based study. Psychol Med 2014;44:1625-37. treatment and mortality in schizophrenia. Schizophr Bull 2015;41:656-63.
84. Lemogne C, Nabi H, Melchior M et al. Mortality associated with depres- 107. Tsai KY, Lee CC, Chou YM et al. The incidence and relative risk of stroke
sion as compared with other severe mental disorders: a 20-year follow- in patients with schizophrenia: a five-year follow-up study. Schizophr
up study of the GAZEL cohort. J Psychiatr Res 2013;47:851-7. Res 2012;138:41-7.

108. Tsan JY, Stock EM, Gonzalez JM et al. Mortality and guideline-concordant 122. Nielsen J, Skadhede S, Correll CU. Antipsychotics associated with the
care for older patients with schizophrenia: a retrospective longitudinal development of type 2 diabetes in antipsychotic-naive schizophrenia
study. BMC Med 2012;10:147. patients. Neuropsychopharmacology 2010;35:1997-2004.
109. van Marwijk HW, van der Kooy KG, Stehouwer CD et al. Depression 123. Tiihonen J, Mittendorfer-Rutz E, Torniainen M et al. Mortality and cumu-
increases the onset of cardiovascular disease over and above other deter- lative exposure to antipsychotics, antidepressants, and benzodiazepines
minants in older primary care patients, a cohort study. BMC Cardiovasc in patients with schizophrenia: an observational follow-up study. Am J
Disord 2015;15:40. Psychiatry 2016;173:600-6.
110. Weeke A, Juel K, Vaeth M. Cardiovascular death and manic-depressive 124. Pillarella J, Higashi A, Alexander GC et al. Trends in use of second-
psychosis. J Affect Disord 1987;13:287-92. generation antipsychotics for treatment of bipolar disorder in the United
111. Westman J, Hallgren J, Wahlbeck K et al. Cardiovascular mortality in bipolar States, 1998-2009. Psychiatr Serv 2012;63:83-6.
disorder: a population-based cohort study in Sweden. BMJ Open 2013;3. 125. Davidson JR. Major depressive disorder treatment guidelines in America
112. Wu HC, Chou FH, Tsai KY et al. The incidence and relative risk of stroke and Europe. J Clin Psychiatry 2010;71(Suppl. E1):e04.
among patients with bipolar disorder: a seven-year follow-up study. PLoS 126. Manu P, Correll CU, Wampers M et al. Markers of inflammation in
One 2013;8:e73037. schizophrenia: association vs. causation. World Psychiatry 2014;13:189-
113. Wu SI, Chen SC, Liu SI et al. Relative risk of acute myocardial infarction 92.
in people with schizophrenia and bipolar disorder: a population-based 127. De Hert M, Cohen D, Bobes J et al. Physical illness in patients with severe
cohort study. PLoS One 2015;10:e0134763. mental disorders. II. Barriers to care, monitoring and treatment guide-
114. Vancampfort D, Stubbs B, Mitchell AJ et al. Risk of metabolic syndrome lines, plus recommendations at the system and individual level. World
and its components in people with schizophrenia and related psychotic Psychiatry 2011;10:138-51.
disorders, bipolar disorder and major depressive disorder: a systematic 128. Correll CU, Lencz T, Malhotra AK. Antipsychotic drugs and obesity.
review and meta-analysis. World Psychiatry 2015;14:339-47. Trends Mol Med 2011;17:97-107.
115. Laursen TM, Nordentoft M, Mortensen PB. Excess early mortality in 129. Stubbs B, Firth J, Berry A et al. How much physical activity do people
schizophrenia. Annu Rev Clin Psychol 2014;10:425-48. with schizophrenia engage in? A systematic review, comparative meta-
116. Nielsen RE, Uggerby AS, Jensen SO et al. Increasing mortality gap for analysis and meta-regression. Schizophr Res 2016;176:431-40.
patients diagnosed with schizophrenia over the last three decades a Dan- 130. Dickerson F, Stallings CR, Origoni AE et al. Cigarette smoking among per-
ish nationwide study from 1980 to 2010. Schizophr Res 2013;146:22-7. sons with schizophrenia or bipolar disorder in routine clinical settings,
117. De Hert M, Detraux J, van Winkel R et al. Metabolic and cardiovascular 1999-2011. Psychiatr Serv 2013;64:44-50.
adverse effects associated with antipsychotic drugs. Nat Rev Endocrinol 131. Bly MJ, Taylor SF, Dalack G et al. Metabolic syndrome in bipolar disorder
2011;8:114-26. and schizophrenia: dietary and lifestyle factors compared to the general
118. Correll CU, Joffe BI, Rosen LM et al. Cardiovascular and cerebrovascular risk population. Bipolar Disord 2014;16:277-88.
factors and events associated with second-generation antipsychotic com- 132. Speyer H, Norgaard HCB, Birk M et al. The CHANGE trial: no superiority
pared to antidepressant use in a non-elderly adult sample: results from a of lifestyle coaching plus care coordination plus treatment as usual com-
claims-based inception cohort study. World Psychiatry 2015;14:56-63. pared to treatment as usual alone in reducing risk of cardiovascular dis-
119. Vancampfort D, Correll CU, Wampers M et al. Metabolic syndrome and ease in adults with schizophrenia spectrum disorders and abdominal
metabolic abnormalities in patients with major depressive disorder: a obesity. World Psychiatry 2016;15:155-65.
meta-analysis of prevalences and moderating variables. Psychol Med 133. Stroup DF, Berlin JA, Morton SC et al. Meta-analysis of observational
2014;44:2017-28. studies in epidemiology: a proposal for reporting. Meta-analysis Of
120. De Hert M, Correll CU, Bobes J et al. Physical illness in patients with Observational Studies in Epidemiology (MOOSE) group. JAMA 2000;283:
severe mental disorders. I. Prevalence, impact of medications and dispar- 2008-12.
ities in health care. World Psychiatry 2011;10:52-77.
121. Correll CU, Robinson DG, Schooler NR et al. Cardiometabolic risk in
patients with first-episode schizophrenia spectrum disorders: baseline
results from the RAISE-ETP study. JAMA Psychiatry 2014;71:1350-63. DOI:10.1002/wps.20420

RESEARCH REPORT
Has the rising placebo response impacted antidepressant clinical

trial outcome? Data from the US Food and Drug
Administration 1987-2013
Arif Khan1,2, Kaysee Fahl Mar1, Jim Faucett1, Shirin Khan Schilling1,3, Walter A. Brown4
1
Northwest Clinical Research Center, Bellevue, WA, USA; 2Department of Psychiatry, Duke University School of Medicine, Durham, NC, USA; 3Department of Psychiatry, Uni-
versity of Connecticut, Hartford, CT, USA; 4Department of Psychiatry and Human Behavior, Brown University, Providence, RI, USA
More than fifteen years ago, it was noted that the failure rate of antidepressant clinical trials was high, and such negative outcomes were
thought to be related to the increasing magnitude of placebo response. However, there is considerable debate regarding this phenomenon and
its relationship to outcomes in more recent antidepressant clinical trials. To investigate this, we accessed the US Food and Drug Administra-
tion (FDA) reviews for sixteen antidepressants (85 trials, 115 trial arms, 23,109 patients) approved between 1987 and 2013. We calculated the
magnitude of placebo and antidepressant responses, antidepressant-placebo differences, as well as the effect sizes and success rates, and com-
pared these measures over time. Exploratory analysis investigated potential changes in trial design and conduct over time. As expected, the
magnitude of placebo response has steadily grown in the past 30 years, increasing since 2000 by 6.4% (r50.46, p<0.001). Contrary to expecta-
tions, a similar increase has occurred in the magnitude of antidepressant response (6.0%, r50.37, p<0.001). Thus, the effect sizes (0.30 vs. 0.29,
p50.42) and the magnitude of antidepressant-placebo differences (10.5% vs. 10.3%, p50.37) have remained statistically equivalent. Further-
more, the frequency of positive trial arms has gone up in the past 15 years (from 47.8% to 63.8%), but this difference in frequency has not
reached statistical significance. Trial design features that were previously associated with a possible lower magnitude of placebo response were
not implemented, and their relationship to the magnitude of placebo response could not be replicated. Of the 34 recent trials, two imple-
mented enhanced interview techniques, but both of them were unsuccessful. The results of this study suggest that the relationship between the
magnitude of placebo response and the outcome of antidepressant clinical trials is weak at best. These data further indicate that anti-
depressant-placebo differences are about the same for all of the sixteen antidepressants approved by the FDA in the past thirty years.
Key words: Antidepressants, clinical trials, placebo response, antidepressant-placebo difference, effect size, success rate, enhanced inter-
view techniques
Fifteen years following the advent of several new antide- responders, defined as patients with 50% or greater reduction
pressants in the mid-1980s, it became evident that the in depressive symptoms, had remained the same after 1991.
success rate of antidepressant clinical trials was low; less However, no mechanism was offered to explain this shift from
than 50% of trials demonstrated statistical superiority for anti- a growing placebo response to a steady one11, nor did the
depressants over placebo1,2. Following Walsh et als finding3 of authors evaluate the effect of such a phenomenon on the out-
a rising placebo response, it was assumed that the clinical trial come of antidepressant clinical trials.
failure rate was related to this phenomenon4. Concern over the impact of increasing placebo response on
Investigators have attempted to determine if the increasing antidepressant clinical trials has fueled a line of inquiry look-
placebo response in antidepressant clinical trials observed by ing for variables predicting higher rates of placebo response,
Walsh et al3 continues to this day. Meta-analytic reviews of based on post-hoc analyses12,13. Several hypotheses, such as
antidepressant clinical trials5,6, or psychotropic trials in general7, the idea that more severely depressed patients might be rela-
as well as patient-level data in trials for major depression8 tively non-responsive to placebo, have been proposed on the
have converged in showing that the placebo response has con- basis of associative observations from these analyses14,15. How-
tinued to grow over the past 15 years. Furthermore, Khin et al9 ever, prospectively selecting more severely depressed patients
conducted an internal review for the US Food and Drug for antidepressant clinical trials has neither resulted in a reduc-
Administration (FDA), which seemed to confirm that the mag- tion in magnitude of the placebo response nor in enhanced
nitude of placebo response was continuing to increase. antidepressant-placebo differences16.
Although this group of investigators had access to specific Research has illuminated other possible variables, such as
data, they did not identify the antidepressant trials that they the flexible dosing of the investigational antidepressant, poten-
reviewed. tially showing a relationship to reduction of placebo response17.
One discordant voice is a study published by Furukawa This flexible dosing schedule has been suggested for use in anti-
et al10, which contradicts the observation of an increase in pla- depressant clinical trials but, as of now, not fully implemented.
cebo response rate in more recent trials. These investigators Furthermore, retrospective analysis of earlier trials has found
conducted a review of 252 depression studies, examining the that placebo response is higher in trials of longer duration18
rate of therapeutic response to placebo using various depen- compared to shorter ones, although this phenomenon has not
dent measures. They surmised that the proportion of placebo been tested prospectively.

Another hypothesis has been that the magnitude of placebo magnitude of antidepressant response as well as the magnitude
response and its variability was related to the low reliability of placebo response, and clearly report the statistical analysis
among clinicians assessing depressed patients19-21. It was used for efficacy approval of the antidepressant. Last, this data-
then recommended that patient sessions should be audio- or base is very large, with patient numbers in the tens of thou-
video-taped and audited by a centralized group of specifically sands, allowing to observe patterns with more confidence.
trained raters to increase reliability. This type of enhanced We hypothesized that the magnitude of placebo response
interviewing technique has been implemented, although its has continued to increase in more recent antidepressant clini-
effects on the outcome of more recent antidepressant trials cal trials, and that such an increase in placebo response may
remain questionable22,23. have reduced the frequency of successful trials. Also, we theo-
What stands out from these studies aiming to elucidate fac- rized that an increase in placebo response would correspond
tors possibly mitigating placebo response in antidepressant to a decrease in the antidepressant-placebo differences and
clinical trials is that such factors are elusive and complex, and observed effect sizes of more recent antidepressant clinical tri-
that their predictive ability varies across different contexts24. als. Lastly, we explored if any of the research design features or
This lack of fruitfulness in pinpointing what may moderate enhanced interview techniques proposed to help contain pla-
placebo response in antidepressant clinical trials has led to a cebo response have been implemented, and if so, with what
form of therapeutic nihilism. results.
In fact, following the observation that antidepressant effica-
cy in clinical trials appears more robust when severely de-
pressed patients are included, and that antidepressants do not
METHODS
reliably perform better than placebo, criticism has been raised
regarding antidepressants overall therapeutic efficacy and
Selection of trials
ability to treat the more mildly depressed population25-28.
However, other investigators do not agree with this view, con-
For the purpose of determining if the pattern of increasing
tending that the magnitude of the antidepressant-placebo
placebo response continued in antidepressant clinical trials fol-
response in clinical trials does not reflect the actual therapeu-
lowing Walsh et als observation3, we formulated groups based
tic efficacy of antidepressants in ordinary clinical practice29-32.
on this point in time. We assigned each trial for an investiga-
However, in the midst of this investigative history, it has
tional antidepressant to the year that the antidepressant was
become obvious that expectations for antidepressant effect
approved, and grouped trials into pre-2000 and post-2000 ones.
have changed as use of psychiatric medications has increased
We included only acute, parallel-group, double-blind, placebo-
exponentially in the past 30 years33. For example, currently
controlled trials for investigational antidepressants approved
one in six adults in the US are reported to have taken a psychi-
after registering a new drug application (NDA) program with
atric medication (primarily antidepressants) in the past year34,
the FDA. Trials were included if they enrolled adult patients
potentially indicating high regards for antidepressant efficacy.
with a primary diagnosis of major depressive disorder.
This observation of a potential increase in expectations for
Data from treatment arms evaluating active comparator
antidepressants has given credence to the theory that placebo
antidepressants (approved antidepressants not under investi-
response has increased due to the heightened expectations of
gation) were excluded from this analysis, due to the fact that
clinicians and patients. Specifically, studies investigating this
the focus of this examination was to characterize new antide-
theory35-37 showed that the higher the risk of receiving placebo
pressants in the process of gaining approval, not performance
in an antidepressant clinical trial, the lower was the magnitude
of established antidepressants.
of placebo response. The caveat is that this theory has not
In addition, we excluded data from treatment arms of inves-
been fully tested prospectively.
tigational antidepressants at dosing levels not approved by the
Given the possibility that the magnitude of placebo re-
FDA, as shown in product labeling. Therefore, we examined
sponse continued to increase in recent antidepressant clinical
only the clinical trial data from arms with doses expected to
trials and may have impacted the outcome of these trials, we
guide approved use of the investigational antidepressant.
conducted the present study. We evaluated data from the med-
We excluded depression trials enrolling only geriatric (>65
ical and statistical reviews of sixteen antidepressant programs
years old) patients, children (<18 years old) and inpatients, as
approved by the FDA from 1987 to 2013, comparing the earlier
well as relapse prevention or maintenance studies, as it is not pos-
antidepressant clinical trials to more recent ones.
sible to draw comparisons between trials studying unique popu-
We decided to conduct this analysis using the FDA clinical
lations or with confounding differences in experimental design.
trial database38 for several reasons. First, these data are not
influenced by publication/investigator/analysis bias, while
these selectivity biases are common in the published litera- Trials included and excluded in this analysis
ture39,40. Second, findings are verified at the source by the FDA
staff in order to authenticate them. Third, reviews conducted After review of the FDA database for NDA registrations
by the FDA more often provide an analysis that includes the approved between 1987 and 2013, we identified a total of

sixteen adult depression programs for inclusion in the analy- in previous analyses of antidepressant clinical trials39,41. As not-
sis. The investigational antidepressants (with year of approval) ed in Turner et als paper39, the formula for calculation of
were: fluoxetine hydrochloride (1987), sertraline hydrochloride Hedges g requires baseline scores, change scores and confi-
(1991), paroxetine hydrochloride (1992), venlafaxine hydro- dence intervals, as well as number of patients to generate t
chloride (1993), nefazodone hydrochloride (1994), mirtazapine scores. NDA packets do not reliably report these data in full, and
(1996), bupropion hydrochloride SR (1996), venlafaxine hydro- therefore we followed the statistical workaround method out-
chloride ER (1997), citalopram (1998), escitalopram oxalate lined in the supplement to Turners paper42, using the inverse t
(2002), duloxetine hydrochloride (2002), desvenlafaxine succi- score function in Microsoft Excel. Precise p values and degrees
nate (2008), trazodone hydrochloride ER (2010), vilazodone of freedom are imputed into the function to calculate a t score,
hydrochloride (2011), levomilnacipran hydrochloride (2013) which can be transformed to Hedges g using a specific equa-
and vortioxetine hydrobromide (2013). tion. Hedges g effect size relies on number of patients and
These programs comprised a total of 125 efficacy evaluation therefore is susceptible to sample size error. We used an appro-
trials. We excluded 40 trials after applying our selection criteria: priate correction to mitigate this risk.
six were conducted in a geriatric population, 22 were uncon- Corrected Hedges g scores were calculated for each trial
trolled, four were carried out in inpatients, four had a relapse arm. We examined effect sizes for trial arms as opposed to
prevention design, and four used doses not approved by the means for the trial overall, because the FDA evaluates efficacy
FDA. Thus, 85 registration trials were included in this analysis. for trial arms separately and uses these individual comparisons
These 85 trials had 172 treatment arms: 33 were active com- to support efficacy claims. Since the FDA approval process con-
parators and 24 utilized a dose of the investigational antide- siders these individual comparisons, we wanted to examine
pressant not approved by the FDA. After excluding these 57 individual treatment arm effects sizes both to retain the vari-
arms, 115 active treatment arms of investigational antidepres- ability of signal detection among differing dose levels as well as
sants at approved doses remained for analysis. to replicate the data handling of the FDA approval process. To
generate a mean effect size for the two groups of pre-2000 and
post-2000, we weighted the corrected effect size by the degrees
Data analysis
of freedom to further account for sample size error.
All statistics were performed with IBM Statistical Package
The medical and statistical reviews conducted by the FDA
for Social Sciences (SPSS). Independent sample t tests were
contain the published results of efficacy analysis along with
used to compare means from older antidepressant trial arms
the treatment group raw baseline and change scores on the
to the more recent ones, to evaluate if any significant changes
primary efficacy measure when available. We encountered
had occurred in the distribution of scores from outcome meas-
alternative statistical methods for handling missing data from
ures. Correlations between year of new drug approval and per-
patient dropout in the reporting and analysis of these efficacy
cent symptom reduction, and between year of new drug
data. These methods included observed cases analysis, analy-
approval and mean program effect size, were calculated using
sis of covariance, and last observation carried forward (LOCF).
Pearsons coefficient.
Since data from LOCF analysis were available for all of the tri-
We calculated frequency of trial design characteristics,
als, we decided to use data (primary efficacy measure scores, p
including duration (8 weeks and <8 weeks), number of trial
values, and patient numbers) from these LOCF statistical com- arms (2 arms or 3 arms), and dosing schedule (fixed or flexi-
putation tables. ble). We computed percentages of trials using either category
We decided to calculate percent symptom reduction as our of design feature and used chi-square analysis of proportions
measure of response magnitude. We divided the mean change to explore any evidence of systematic implementation.
score reported in the FDA reviews by the mean baseline score Statistical analysis of the results of trials using enhanced
and multiplied by 2100 to get a percent symptom reduction interview assessment techniques was not possible because
that takes into account variation in baseline and different only two (vortioxetine 317 and levomilnacipran MD-02) recent
measurement scales. This measure was calculated for placebo
trials out of 34 used such techniques.
and antidepressant treatment groups separately.
We calculated the average antidepressant-placebo differ-
ence, taken by subtracting the placebo percent symptom
reduction from the antidepressant percent symptom reduc- RESULTS
tion for each trial arm. In instances where placebo had a great-
er percent symptom reduction than antidepressant, this meas- Tables 1 and 2 report the program/trial essential character-
ure would be negative. istics. Prior to 2000, there were nine antidepressant NDA pro-
Success of a treatment arm was defined as it is in the FDA grams, contributing 51 trials and 67 active treatment arms
reviews, with a p value threshold of 0.05 for endpoint analysis of from efficacy tables that met our inclusion/exclusion criteria.
the primary efficacy measure. The seven programs approved after 2000 supplied 34 trials and
We calculated effect sizes for individual treatment arm com- 48 active treatment arms for analysis. Four (6.0%) of the treat-
parisons using Hedges g formula. This procedure has been used ment arms in pre-2000 trials and 13 (27.1%) of the treatment

Table 1 Characteristics of 51 clinical trials for the approval of nine antidepressants from 1987 to 2000
184
Placebo Investigational antidepressant
Primary Baseline/change Baseline/change p value for
Protocol Dosing Duration efficacy No. score on primary Percent No. score on primary Percent efficacy Effect size
number schedule (weeks) measure patients efficacy measure response patients efficacy measure response calculation (Hedges g)
Fluoxetine (1987)
19 Flexible 4 HAM-D 24 28.2/25.5 19.5 22 28.6/212.5 43.7 0.011 0.77
27 Flexible 6 HAM-D 163 28.2/28.4 29.8 181 27.5/211.0 40.0 0.012 0.27
25 Flexible 4 HAM-D 24 25.8/28.8 34.1 18 26.2/27.2 27.5 0.50 0.21

62-A Fixed 6 HAM-D 56 4.0/21.21 30.2 105 3.9/21.33 34.1 0.46 0.12
103 3.9/21.38 35.4 0.34 0.16
100 3.9/21.19 30.5 0.50 0.11
62-B Fixed 6 HAM-D 48 24.3/25.7 23.4 97 24.7/29.8 39.6 0.007 0.48
97 24.1/29.6 39.8 0.01 0.46
103 24.2/27.2 29.8 0.34 0.17
Sertraline (1991)
103 Fixed 6 HAM-D 86 25.3/27.6 30.0 90 24.8/210.6 42.7 0.34 0.32
89 24.9/29.8 39.4 0.102 0.25
82 25.7/29.9 38.5 0.252 0.23
104 Flexible 8 HAM-D 141 23.4/28.2 35.0 142 23.3/211.7 50.2 0.001 0.40
315 Flexible 8 HAM-D 73 22.2/26.7 29.6 76 23.1/28.8 38.1 0.46 0.12
Paroxetine (1992)
01-001 Flexible 6 HAM-D 24 27.4/210.5 38.3 24 28.0/213.5 48.2 0.204 0.37
02-001 Flexible 6 HAM-D 53 25.9/26.8 26.3 51 26.6/212.3 46.2 0.0042 0.57
02/002 Flexible 6 HAM-D 34 24.9/25.8 23.3 36 25.0/210.9 43.6 0.0146 0.60
02/003 Flexible 6 HAM-D 33 28.9/27.2 24.9 33 28.6/29.7 33.9 0.3092 0.25
02-004 Flexible 4 HAM-D 38 27.3/27.2 26.4 36 28.9/212.2 42.2 0.0018 0.75

03-005 Flexible 6 HAM-D 42 26.8/24.0 14.9 40 26.1/210.0 38.3 0.0076 0.60
03-006 Flexible 6 HAM-D 38 28.7/23.0 10.5 39 29.7/29.1 30.6 0.0014 0.75
03-001 Flexible 4 HAM-D 38 24.8/24.7 19.0 40 24.9/210.8 42.4 0.006 0.63
03-002 Flexible 4 HAM-D 40 25.6/26.2 24.2 40 24.9/28.0 32.1 0.0004 0.61
03-003 Flexible 4 HAM-D 42 27.0/29.2 34.1 39 25.7/29.3 36.2 0.98 0.01
03-004 Flexible 4 HAM-D 37 27.0/26.7 24.8 37 27.6/210.4 37.7 0.04 0.48
09 Fixed 12 HAM-D 51 NR/28.2 NR 104 NR/210.6 NR 0.34 0.16
99 NR/29.0 NR 0.34 0.16
100 NR/29.4 NR 0.34 0.16

Table 1 Characteristics of 51 clinical trials for the approval of nine antidepressants from 1987 to 2000 (continued)
Venlafaxine (1993)
600A-206 Flexible 4 HAM-D 47 28.6/24.8 16.8 46 28.2/214.2 50.4 0.006 0.58

600A-301 Flexible 6 HAM-D 78 24.6/29.5 38.6 64 25.4/213.9 54.7 0.0004 0.61
600A-302 Flexible 6 HAM-D 75 24.4/28.9 36.5 65 25.0/211.9 47.6 0.008 0.45
600A-303 Flexible 6 HAM-D 79 24.6/29.9 40.2 70 23.6/210.1 42.8 0.493 0.11
600A-203 Fixed 6 HAM-D 92 25.3/26.7 26.5 77 26.0/211.1 42.7 0.004 0.45
79 26.0/211.9 45.8 0.001 0.51
75 24.9/210.5 42.2 0.003 0.47
600A-313 Fixed 6 HAM-D 75 25.4/29.5 37.4 72 25.6/210.9 42.6 0.193 0.21
77 25.6/211.8 46.1 0.142 0.24
Nefazodone (1994)
030A2-0007 Fixed 6 HAM-D 47 26.4/29.8 37.1 47 25.4/210.7 42.1 0.60 0.11
03A0A-003 Fixed 6 HAM-D 45 25.9/26.8 26.3 44 25.4/211.0 43.3 0.03 0.46
03A0A-004A Fixed 6 HAM-D 77 23.5/28.5 36.2 76 23.6/29.0 38.1 0.66 0.71
03A0A-004B Fixed 6 HAM-D 80 25.0/29.4 37.6 78 25.4/212.4 48.8 0.02 0.37

CN104-005 Flexible 8 HAM-D 90 23.5/28.0 34.0 86 24.4/212.0 49.2 0.01 0.39
CN104-006 Flexible 8 HAM-D 78 23.8/28.9 37.4 80 23.5/210.0 42.6 0.35 0.15
Mirtazapine (1996)
003-002 Flexible 6 HAM-D 44 24.7/25.4 21.9 44 24.2/211.7 48.3 0.0008 0.73
003-003 Flexible 6 HAM-D 45 25.5/28.8 34.5 45 25.4/210.4 40.9 0.49 0.14
003-008 Fixed 6 HAM-D 28 25.8/29.6 37.2 30 26.0/27.6 29.2 0.293 20.28
28 25.5/27.3 28.6 0.282 20.29
30 25.3/28.1 32.0 0.346 20.25
003-020/3220 Flexible 6 HAM-D 39 29.5/24.8 16.3 40 27.8/210.3 37.1 0.004 0.66
003-021/3220 Flexible 6 HAM-D 48 24.4/29.5 38.9 45 24.2/211.7 48.3 0.223 0.25
003-022/3220 Flexible 6 HAM-D 50 31.2/29.0 28.8 49 33.0/216.1 48.8 0.003 0.61
003-024/3220 Flexible 6 HAM-D 48 27.7/27.7 27.8 50 27.5/212.1 44.0 0.01 0.53
85027 Flexible 4 HAM-D 61 26.2/210.9 41.6 64 26.4/213.4 50.8 0.189 0.23
185
186
Table 1 Characteristics of 51 clinical trials for the approval of nine antidepressants from 1987 to 2000 (continued)
Bupropion SR (1996)
203 Fixed 8 HAM-D 117 23.2/28.1 34.9 113 23.4/210.2 43.6 0.04 0.27
205 Fixed 8 HAM-D 116 23.4/28.3 35.5 111 23.6/29.0 38.1 0.53 0.8
111 24.2/29.3 38.4 0.30 0.14
212 Fixed 8 HAM-D 148 23.9/29.8 41.0 144 24.4/211.1 45.5 0.16 0.16
Venlafaxine ER (1997)
208 Flexible 12 HAM-D 91 24.6/28.7 35.4 85 24.4/214.9 61.1 0.001 0.50
209 Flexible 8 HAM-D 100 23.6/26.8 28.8 91 24.5/211.7 47.8 0.0003 0.53
367 Flexible 8 HAM-D 81 26.6/213.1 49.2 83 26.5/215.6 58.9 0.37 0.14
85 NR/NR NR 0.14 0.23
Citalopram (1998)
85A Flexible 4 HAM-D 78 33.7/29.6 28.5 82 33.5/212.9 38.5 0.0344 0.33
86141 Flexible 6 HAM-D 50 21.0/24.9 23.3 97 22.2/26.3 28.4 0.316 0.17
89303 Fixed 6 HAM-D 64 23.7/210.6 44.7 61 23.0/213.3 57.8 0.12 0.28
91206 Fixed 6 HAM-D 124 24.6/29.3 37.8 120 24.4/212.2 50.0 0.0025 0.39
110 24.5/212.1 49.4 0.0053 0.37
89306 Fixed 6 MADRS 88 33.1/216.0 48.3 97 31.3/216.0 51.1 0.964 0.07
NR data not reported or censored in the Food and Drug Administration packet, HAM-D Hamilton Depression Rating Scale, MADRS Montgomery-Asberg Depression Rating Scale
Bold prints indicate a positive trial arm

Table 2 Characteristics of 34 clinical trials for the approval of seven antidepressants after 2000

Baseline/change
Primary score on primary Baseline/change p value for
Protocol Dosing Duration efficacy No. efficacy Percent No. score on primary Percent efficacy Effect size
number schedule (weeks) measure patients measure response patients efficacy measure response calculation (Hedges g)
Escitalopram (2002)

MD01 Fixed 8 MADRS 119 29.5/29.4 31.9 118 28.0/212.8 45.7 0.0007 0.45
125 28.9/213.9 48.1 0.0001 0.51
MD 02 Flexible 8 MADRS 125 28.8/211.2 38.9 124 28.7/212.9 45.0 0.251 0.15
99001 Fixed 8 MADRS 189 28.7/213.6 47.4 188 29.2/216.3 55.8 0.006 0.28
99003 Flexible 8 MADRS 154 28.7/212.5 43.6 155 29.0/215.3 52.8 0.0064 0.31
Duloxetine (2002)
HMAQa Flexible 8 HAM-D 57 20.6/26.5 31.6 56 19.6/28.5 43.4 0.15 0.27
HMAQb Flexible 8 HAM-D 55 20.0/25.7 28.5 61 19.9/26.2 31.2 0.95 0.01
HMATa Fixed 8 HAM-D 89 17.8/24.3 24.2 81 17.4/25.5 31.6 0.138 0.23
HMATb Fixed 8 HAM-D 88 17.2/24.2 24.2 86 18.1/27.7 42.7 0.003 0.45
HMBHa Fixed 9 HAM-D 115 21.1/25.2 24.5 121 21.5/29.3 43.0 0.001 0.43
HMBHb Fixed 9 HAM-D 136 20.5/27.2 35.3 123 20.3/28.9 43.8 0.048 0.25
Desvenlafaxine (2008)
332 Fixed 9 HAM-D 150 23.0/29.6 41.7 150 23.4/211.5 49.2 0.02 0.27
147 23.4/211.0 47.0 0.09 0.20
333 Fixed 8 HAM-D 161 24.3/210.8 44.4 164 24.3/213.2 54.3 0.004 0.32
158 24.4/213.7 56.2 0.001 0.37
223 Fixed 8 HAM-D 78 NR NR 63 NR NR 0.59 0.09
72 NR NR 0.52 0.11
116 NR/29.6 NR 0.076 0.23
113 NR/210.5 NR 0.002 0.41
124 NR/212.1 NR 0.008 0.34
304 Flexible 8 HAM-D 114 NR/NR NR 120 NR/NR NR 0.28 0.14
309 Flexible 8 HAM-D 120 NR/212.5 NR 117 NR/213.4 NR 0.381 0.11
187
Table 2 Characteristics of 34 clinical trials for the approval of seven antidepressants after 2000 (continued)
188
Baseline/change
Primary score on primary Baseline/change p value for
Protocol Dosing Duration efficacy No. efficacy Percent No. score on primary Percent efficacy Effect size
number schedule (weeks) measure patients measure response patients efficacy measure response calculation (Hedges g)
Trazodone ER (2010)
04ACL3-001 Flexible 8 HAM-D 206 22.4/29.25 41.3 206 23.2/211.2 48.2 0.0055 0.27
Vilazodone (2011)
GNSC-04-DP-02 Flexible 8 MADRS 199 30.7/29.7 31.6 198 30.8/212.9 41.9 0.001 0.33
CLDA-07-DP-02 Fixed 8 MADRS 231 32.0/210.8 33.8 232 31.9/213.3 41.7 0.009 0.24
Levomilnacipran (2013)
MD-01 Fixed 8 MADRS 175 35.6/211.6 32.6 176 36.0/214.8 41.1 0.0186 0.25
177 36.1/215.6 43.2 0.0038 0.31
176 36.0/216.5 45.8 0.0005 0.37
MD-03 Flexible 8 MADRS 214 35.2/212.2 33.8 215 35.0/215.3 43.7 0.0051 0.27
MD-10 Flexible 8 MADRS 185 31.0/211.3 36.5 185 30.8/214.6 47.4 0.0027 0.31
187 31.2/214.4 46.2 0.0043 0.30
MD-02 Flexible 8 MADRS 182 35.5/214.2 40.0 175 NR/NR NR NR NR
F02695 LP2 02 Flexible 10 MADRS 277 30.5/214.5 47.5 276 30.7/218.7 60.9 0.0001 0.55
Vortioxetine (2013)
11492A Fixed 6 MADRS 105 33.9/214.5 42.8 100 34.0/220.2 59.4 0.0001 0.55
305 Fixed 8 HAM-D 139 32.7/211.3 35.6 139 33.1/216.2 48.9 0.001 0.40
13267A Fixed 8 MADRS 158 31.5/211.7 37.1 149 31.8/217.2 54.1 0.0001 0.45
151 31.2/218.8 60.3 0.0001 0.45
315US Fixed 8 MADRS 153 31.5/212.8 40.6 145 31.9/214.3 44.8 0.224 0.14
147 32.0/215.6 48.8 0.023 0.26
316US Fixed 8 MADRS 155 32.0/210.8 33.8 154 32.2/213.0 40.4 0.058 0.19
148 32.5/214.4 44.3 0.002 0.36
11984A Fixed 8 MADRS 145 NR/214.8 NR 151 NR/216.3 NR 0.185 0.15
317 Fixed 8 MADRS 149 33.4/212.9 38.6 143 34.1/213.7 40.2 0.597 0.06
142 33.6/213.4 39.9 0.745 0.04
NR data not reported or censored in the Food and Drug Administration packet, HAM-D Hamilton Depression Rating Scale, MADRS Montgomery-Asberg Depression Rating Scale
Bold prints indicate a positive trial arm

Table 3 Evaluation of efficacy outcomes in antidepressant regis- 6.4% (t522.9, df574, p50.005). This represents a 21.5%
tration trials before and after 2000 change over 15 years (Table 3).
Before 2000 After 2000 p
Percent symptom reduction as a measure of response mag-
nitude increased by an almost identical 6.0% in the antide-
No. programs 9 7 pressant treatment arm, from pre-2000 trials at 40.6% (613.7)
No. trials 51 34 to post-2000 trials at 46.6% (67.0) (t522.9, df596, p50.005).
This represents a 14.8% change over 15 years (Table 3).
No. active treatment arms 67 48
Figure 1 shows placebo and antidepressant response rates
Successful treatment arms 47.8% (32/67) 63.8% (30/47) 0.09
over time. Growth rate was nearly parallel in placebo and anti-
% symptom reduction depressant treatments, with both treatment conditions having
Antidepressant 40.6% (613.7) 46.6% (67.0) 0.005 significant positive relationships (placebo: r50.46, p<0.001;
Placebo 29.8% (612.6) 36.2% (66.6) 0.005 antidepressants: r50.37, p<0.001) between time and percent
Mean antidepressant-placebo 10.5% (69.2) 10.3% (65.0) 0.37 symptom reduction.
difference The antidepressant-placebo differences have remained equi-
Effect size (Hedges g) 0.30 (60.24) 0.29 (60.12) 0.42 valent over the years, as a result of matching growth in both
treatment condition responses. The mean antidepressant-
placebo difference in trials from pre-2000 was 10.5% (69.2) as
compared to 10.3% (65.0) in the trials post-2000 (p50.37)
arms in post-2000 trials had missing baseline or change score (Table 3).
data. Treatment arms for antidepressant clinical trials conducted
Due to missing data, we calculated placebo response magni- prior to 2000 were successful in 47.8% of cases (32 out of 67
tude based on 76 out of 85 placebo arms (89.4%), antidepressant treatment arms), compared with a treatment arm success rate
response magnitude based on 98 out of 115 treatment arms of 63.8% (30 out of 47) in antidepressant trials post-2000. Chi-
(85.2%), antidepressant-placebo differences based on 98 out of square analysis of proportions determined that this difference
115 antidepressant-placebo group comparisons (85.2%), and was not statistically significant (p50.09).
effect sizes based on 114 out of 115 treatment arms (99.1%). Effect sizes based on number of patients and p values from
Prior to 2000, placebo reduced symptoms on average by individual treatment arm LOCF analysis revealed no signifi-
29.8% (612.6) compared to 36.2% (66.6) in programs post- cant change over the 31 years of antidepressant program data.
2000, resulting in a significant increase in placebo response by The average weighted effect size across trial arms conducted
Figure 1 Percent symptom reduction in 74 placebo and 92 antidepressant treatment arms from 85 clinical trials for 16 antidepressant approval
programs plotted with time. The correlation between year of new drug approval and percent symptom reduction was significant in both the
placebo (r50.46, p<0.001) and the antidepressant group (r50.37, p<0.001).

The study showed that the pattern of increase in placebo
response noted in 2001 by Walsh et al3 has continued. The
magnitude of symptom reduction with placebo has steadily
increased from 29.8% to 36.2% (p50.005). These results con-
verge with the findings by Khin et al9 and other investigators5-
8
that placebo symptom reduction has continued to increase
in more recent antidepressant clinical trials.
The increase in placebo response observed in recent antide-
pressant clinical trials is in contrast with a recent study by Fur-
ukawa et al10, reporting a stability in placebo response rate
after 1991. We attribute this discrepancy to differences in study
design. That study included data from published sources,
which have been shown to contain selection bias39,40 and fre-
quently use different statistical analyses from those performed
Figure 2 Mean effect size (Hedges g) of antidepressant clinical trials by the FDA reviewers. Therefore, use of published sources may
based on year of approval. There was no significant relationship have resulted in different datasets. Additionally, our study
between year of new drug approval and mean program effect size
(r520.06, p50.85).
used percent symptom reduction as a measure of placebo
response and this value is on a continuum, allowing for analy-
sis of more subtle changes than a binary measure such as
before 2000 was 0.30 (60.24), while for trials after 2000 it was number of patients meeting a therapeutic response threshold,
0.29 (60.12) (p50.42). as used by Furukawa et al10 and many others.
Figure 2 shows this trend of stability in effect size through- Contrary to expectations, given our finding of a continued
out the years simplified by averaging trial effect sizes to gener- increase in placebo response over time, the success rate of
ate overall values for each antidepressant program. Program antidepressant clinical trials has gone up over the past 15 years
effect sizes were not correlated with any kind of change over (from 47.8% to 63.8%, p50.09). This has occurred as the mag-
time (p50.85). nitude of the antidepressant response has also gone up con-
The trial design suggestions12, including enhanced rater siderably (from 40.6% to 46.6%, p50.005).
interview techniques19-21, put forth by investigators based on In essence, both the magnitude of placebo response and
post-hoc analyses of placebo response were not implemented
antidepressant response have steadily increased over the past
in recent clinical trials. Specifically, the trends examined were
thirty years among these sixteen new antidepressant pro-
opposite in direction to the modifications in trial design previ-
grams. The success rate of antidepressant trials has remained
ously suggested: trials were of longer duration, had a greater
about the same, showing a modest increase in recent years.
number of treatment arms, and rarely used flexible dosing
This is confirmed by the finding that treatment arm effect sizes
schedules; all elements previously corresponding to higher
have remained about the same, with a distribution around 0.30,
placebo response. There was no observed association between
and antidepressant-placebo differences continue to show a 10%
trial design features and trial outcomes in post-2000 trials (see
antidepressant advantage regardless of placebo response. In
Table 2 for trial design characteristics).
other words, the newer antidepressants appear about as effica-
Regarding enhanced interview techniques19-21, two out of
34 recent antidepressant clinical trials submitted for review by cious as the older ones.
the FDA used such techniques. Neither of these (trial 317 for Potential remedies that have been suggested in order to
vortioxetine43 and trial MD-02 for levomilnacepran44) was suc- mitigate placebo response, such as changes in study designs
cessful. (use of flexible dosing, shorter duration of trials, and fewer
number of treatment arms12) seem not to have been systemat-
ically implemented or to have had effect on the outcomes of
more recent antidepressant trials. Our exploration also sug-
DISCUSSION
gests that these trial design and conduct factors may not be
causally related to the magnitude of placebo response (see
Given the present state of uncertainty in the research sur-
Table 2), so that the prospective implementation of these sug-
rounding placebo response in antidepressant clinical trials
and the importance of this phenomenon, this study aimed to gestions may not have the effect expected based on theory or
evaluate if placebo response as measured by symptom reduc- observed from retrospective analysis. In particular, the two
tion has continued to rise over the past 15 years compared to antidepressant clinical trials that prospectively used enhanced
the earlier 15 years. The study also attempted to determine if interview techniques failed to show superiority over placebo
decreases in success rate and measures of antidepressant- in NDA programs for vortioxetine and levomilnacepran.
placebo differences accompanied the growth in symptom In this context, it is important to note that the current
reduction with placebo. results do not support earlier studies regarding the impact of

placebo response on trial outcomes, which found that the mag- 9. Khin NA, Chen Y, Yang Y et al. Exploratory analyses of efficacy data from
major depressive disorder trials submitted to the US Food and Drug
nitude of placebo response was inversely associated with the Administration in support of New Drug Applications. J Clin Psychiatry
frequency of positive outcomes in trials conducted between 2011;72:464-72.
1987 and 19994. This relationship holds true for those earlier 10. Furukawa TA, Cipriani A, Atkinson LZ et al. Placebo response rates in anti-
depressant trials: a systematic review of published and unpublished
trials, but has dissolved in the more recent post-2000 trials.
double-blind randomized controlled studies. Lancet Psychiatry 2016;3:
What these current data show is that, in spite of the con- 1059-66.
tinuing growth of placebo response, antidepressants appear to 11. Enck P. Placebo response in depression: is it rising? Lancet Psychiatry
2016;3:1005-6.
maintain an advantage of about 10% (effect size of 0.30, a 12. Khan A, Kolts RL, Thase ME et al. Research design features and patient
modest one), suggesting that acting to mitigate placebo re- characteristics associated with the outcome of antidepressant clinical tri-
sponse may not be a critical component of the success and als. Am J Psychiatry 2004;161:2045-9.
13. Papakostas GI, Ostergaard SD, Iovieno N. The nature of placebo response
outcomes of efficacy analysis in antidepressant clinical trials. in clinical studies of major depressive disorder. J Clin Psychiatry 2015;76:
Potential mechanisms explaining the growth in placebo 456-66.
response and relationship to trial outcomes were not fully 14. Khan A, Leventhal RM, Khan SR et al. Severity of depression and response
to antidepressants and placebo: an analysis of the Food and Drug Admin-
explored in this study. However, we noticed that there has istration database. J Clin Psychopharmacol 2002;22:40-5.
been a substantial increase in the sample size in both placebo 15. Khan A, Brodhead AE, Kolts RL et al. Severity of depressive symptoms and
and antidepressant treatment arms in recent years. As described response to antidepressants and placebo in antidepressant trials. J Psy-
chiatr Res 2005;29:145-50.
by Liu et al44, increased sample size has been associated with 16. Khan A, Schwartz K, Kolts RL et al. Relationship between depression sever-
clinical trial outcomes of investigational hypertension medica- ity entry criteria and antidepressant clinical trial outcomes. Biol Psychiatry
tions, and the relative mechanism calls for further exploration. 2007;62:65-71.
17. Khan A, Khan SR, Walens G et al. Frequency of positive studies among
A drawback to our study is that it was an observational fixed and flexible dose antidepressant clinical trials: an analysis of the
post-hoc analysis rather than prospective in design. More Food and Drug Administration Summary Basis of Approval reports. Neu-
important, FDA medical and statistical reports do not include ropsychopharmacology 2003;28:552-7.
18. Khan A, Khan SR, Leventhal RM et al. Symptom reduction and suicide risk
subject-level data. This summarization of data in FDA reviews in patients treated with placebo in antidepressant clinical trials: a replica-
of new investigational antidepressants does not allow a more tion analysis of the Food and Drug Administration database. Int J Neuro-
psychopharmacol 2001;4:113-8.
detailed analysis. However, the sponsoring pharmaceutical
19. Demitrack MA, Faries D, Herrera JM et al. The problem of measurement
companies or the FDA may undertake such an analysis to pro- error in multisite clinical trials. Psychopharmacol Bull 1998;34:19-24.
vide better insight into the relationship between placebo re- 20. Kobak K, Thase ME. Why do clinical trials fail? The problem of measure-
ment error in clinical trials: time to test new paradigms? J Clin Psycho-
sponse and antidepressant clinical trial outcomes. pharmacol 2007;27:1-5.
In conclusion, the results of this study suggest that the rela- 21. Kobak KA, Feiger AD, Lipsitz JD. Interview quality and signal detection in
tionship between the magnitude of placebo response and the clinical trials. Am J Psychiatry 2005;162:628.
22. Khan A, Faucett J, Brown WA. Magnitude of placebo response and
success of antidepressant clinical trials is weak at best. These response variance in antidepressant clinical trials using structured, taped,
data indicate that the antidepressant-placebo differences are and appraised rater interviews compared to traditional rating interviews.
about the same for all of the sixteen antidepressants approved J Psychiatr Res 2014;51:88-92.
23. Khan A, Faucett J, Brown WA. Magnitude of change with antidepres-
by the FDA in the past thirty years. This finding has implica- sants and placebo in antidepressant clinical trials using structured,
tions for guiding future clinical trials and warrants exploratory taped and appraised rater interview (SIGMA-RAPS) compared to trials
using traditional semi-structured interviews. Psychopharmacology
analysis of other potential factors that may influence the out-
2014;231:4301-7.
come of antidepressant trials. 24. Stein DJ, Baldwin DS, Dolberg OT et al. Which factors predict placebo
response in anxiety disorders and major depression? An analysis of placebo-
controlled studies of escitalopram. J Clin Psychiatry 2006;67:1741-6.
25. Kirsch K, Deacon BJ, Huedo-Medina TB et al. Initial severity and antide-
REFERENCES pressant benefits: a meta-analysis of data submitted to the Food and Drug
1. Khan A, Warner HA, Brown WA. Symptom reduction and suicide risk in Administration. PLoS Med 2008;5:e45.
patients treated with placebo in antidepressant clinical trials. Arch Gen 26. Fournier JC, DeRubeis RJ, Hollon SD. Antidepressant drug effects depres-
Psychiatry 2000;57:311-7. sion severity: a patient-level meta-analysis. JAMA 2010;303:47-53.
2. Khan A, Khan S, Brown WA. Are placebo controls necessary to test new anti- 27. Naudet F, Millet B, Charlier P et al. Which placebo to cure depression? A
depressants and anxiolytics? Int J Neuropsychopharmacol 2002;5:193-7. thought-provoking network meta-analysis. BMC Med 2013;11:230.
3. Walsh BT, Seidman SN, Sysko R et al. Placebo response in studies of major 28. Moncrieff J. Antidepressants: misnamed and misrepresented. World Psy-
depression: variable, substantial, and growing. JAMA 2001;287:1840-7. chiatry 2015;14:302-3.
4. Khan A, Detke M, Khan S et al. Placebo response and antidepressant clini- 29. Fountoulakis KN, Mo ller H. Efficacy of antidepressants: a re-analysis and
cal trial outcome. J Nerv Ment Dis 2003;191:211-8. re-interpretation of the Kirsch data. Int J Neuropsychopharmacol 2011;14:
5. Rief W, Nestoriuc Y, Weiss S et al. Meta-analysis of the placebo response in 405-12.
antidepressant trials. J Affect Disord 2009;118:1-8. 30. Cipriani A, Geddes JR. Placebo for depression: we need to improve the
6. Undurraga J, Baldessarini RJ. Randomized, placebo-controlled trials of quality of scientific information but also reject too simplistic approaches
antidepressants for acute major depression: thirty-year meta-analytic or ideological nihilism. BMC Med 2014;12:105.
review. Neuropsychopharmacology 2012;37:851-64. 31. Montgomery SA. Antidepressant or antidepressant plus placebo effect?
7. Weimer K, Colloca L, Enck P. Placebo effects in psychiatry: mediators and World Psychiatry 2015;14:303-4.
moderators. Lancet Psychiatry 2015;2:246-57. 32. Kasper S, Dold M. Factors contributing to the increasing placebo response
8. Mancini M, Wade AG, Perugi G et al. Impact of patient selection and study in antidepressant trials. World Psychiatry 2015;14:304-6.
characteristics on signal detection in placebo-controlled trials with anti- 33. Khan A, Brown WA. Antidepressants vs. placebo in major depression: an
depressants. J Psychiatr Res 2014;51:21-9. overview. World Psychiatry 2015;14:294-300.

34. Moore TJ, Mattison DR. Adult utilization of psychiatric drugs and differ- 40. Lee K, Bacchetti P, Sim I. Publication of clinical trials supporting success-
ences by sex, age, and race. JAMA Intern Med 2017;177:274-5. ful new drug applications: a literature analysis. PLoS Med 2008;5:1348-56.
35. Sinyor M, Levitt AJ, Cheung AH et al. Does inclusion of a placebo arm 41. Gibertini M, Nations K, Whitaker J. Obtained effect size as function of
influence response to active antidepressant treatment in randomized con- sample size in approved antidepressants: a real-world illustration in sup-
trolled trials? Results from pooled and meta-analysis. J Clin Psychiatry port of better trial design. Int Clin Psychopharmacol 2012;27:100-6.
2010;71:270-9. 42. Turner EH, Matthews AM, Linardatos E et al. Selective publication of anti-
36. Papakostas GI, Fava M. Does the probability of receiving placebo depressant trials and its influence on apparent efficacy. N Engl J Med
influence clinical trial outcome? A meta-regression of double-blind, 2008;358:252-60.
randomized clinical trials in MDD. Eur Neuropsychopharmacol 2009; 43. Mahableshwarkar A, Jacobsen P, Serenko M et al. A randomized, double-
19:34-40. blind, placebo-controlled study of the efficacy and safety of 2 doses of vor-
37. Rutherford BR, Roose SP. A model of placebo response in antidepressant tioxetine in adults with major depressive disorder. J Clin Psychiatry 2015;
clinical trials. Am J Psychiatry 2013;170:723-33. 76:583-91.
38. Food and Drug Administration. www.accessdata.fda.gov. 44. Liu KS, Snavely DB, Ball WA et al. Is bigger better for depression trials?
39. Turner EH, Matthews AM, Linardatos E et al. Selective publication of anti- J Psychiatr Res 2008;42:622-30.
depressant trials and its influence on apparent efficacy. N Engl J Med
2008;358:252-60. DOI:10.1002/wps.20421

RESEARCH REPORT
Risk of suicide, deliberate self-harm and psychiatric illness after

the loss of a close relative: a nationwide cohort study
Mai-Britt Guldin1, Maiken Ina Siegismund Kjaersgaard2, Morten Fenger-Grn1, Erik Thorlund Parner2, Jiong Li2, Anders Prior1,2,
Mogens Vestergaard1,2
1
Research Unit for General Practice, Aarhus University, Aarhus, Denmark; 2Department of Public Health, Aarhus University, Aarhus, Denmark
The loss of a close relative is a common event, yet it is associated with increased risk of serious mental health conditions. No large-scale study
has explored up to now the importance of the bereaved persons relation to the deceased while accounting for gender and age. We performed a
nationwide Danish cohort study using register information from 1995 through 2013 on four sub-cohorts including all persons aged 18 years
exposed to the loss of a child, spouse, sibling or parent. We identified 1,445,378 bereaved persons, and each was matched by gender, age and
family composition to five non-bereaved persons. Cumulative incidence proportions were calculated to estimate absolute differences in suicide,
deliberate self-harm and psychiatric illness. Cox proportional hazard regression was used to calculate hazard ratios while adjusting for poten-
tial confounders. Results revealed that the risk of suicide, deliberate self-harm and psychiatric illness was increased in the bereaved cohorts for
at least 10 years after the loss, particularly during the first year. During that year, the risk difference was 18.9 events in 1,000 persons after loss
of a child (95% CI: 17.6-20.1) and 16.0 events in 1,000 persons after loss of the spouse (95% CI: 15.4-16.6). Hazard ratios were generally highest
after loss of a child, in younger persons, and after sudden loss by suicide, homicide or accident. One in three persons with a previous psychiat-
ric diagnosis experienced suicide, deliberate self-harm or psychiatric illness within the first year of bereavement. In conclusion, this study
shows that the risk of suicide, deliberate self-harm and psychiatric illness is high after the loss of a close relative, especially in susceptible sub-
groups. This suggests the need for early identification of high-risk persons displaying adjustment problems after loss of a close family member,
in order to reduce the risk of serious mental health outcomes.
Key words: Bereavement, suicide, deliberate self-harm, psychiatric hospitalization, loss of a child, sudden loss
Death of a close relative is a common experience in adult- risk of serious mental health conditions in specific subgroups
hood. In the US, it has been estimated that more than 40,000 after the loss of different types of close relatives. Understanding
parents lose a child every year1, and more than half of the pop- the pattern of grief-related disorders and serious mental health
ulation over 65 years are widowed2. Although bereavement is a conditions is important to health care planning13,16,17.
natural life event, it is often followed by emotional suffering In a comprehensive population-based cohort, we investi-
and adjustment challenges. Studies have shown an association gated the absolute and relative risk of suicide, deliberate self-
between the loss of a loved one and a range of mental health harm and psychiatric illness in people exposed to the loss of a
complications, particularly depression and post-traumatic stress child, spouse, parent or sibling. We evaluated whether the
disorder3-9. association was modified by gender, age, urbanization, or pre-
Prolonged and complicated grief reactions have been fre- existing mental conditions or physical diseases.
quently studied, and prolonged grief disorder has recently been
suggested for inclusion in the ICD-1110-14. The relevant Working
Group has concluded that prolonged and complicated grief
reactions are significantly associated with serious psychosocial METHODS
and health problems, including suicidality, substance abuse
and cardiovascular disease11. Study population and design
A representative population-based survey has shown female
gender, old age, and loss of a child or the spouse to be risk fac- A population-based cohort was established by using a unique
tors in grief complications15, whereas epidemiological studies personal identification number which links individual-level
show male gender to be associated with an excess risk of sui- data between nationwide Danish registers18. The study cohort
cide and mortality after loss2. Yet, no studies have investigated comprised four sub-cohorts of persons aged 18 years or older
suicidal behaviour and psychiatric illness across different and residing in Denmark during the inclusion period from Jan-
types of loss, and considered previous history of mental and uary 1, 1995 to December 31, 2012 (N51,445,378). Each sub-
physical illness in the bereaved when interpreting the data15. cohort included all persons exposed to the loss of either a
The vast majority of individuals exposed to the loss of a loved child, spouse/registered partner, parent or sibling.
person exhibit time-limited disruptions to daily functioning, and Exposure to loss was assessed by identifying deceased per-
it has been argued that a mix of genetic, personality and environ- sons and linking them to their family members using informa-
mental determinants act as protective or risk factors16. However, tion from the Danish Civil Registration System19. Sudden and
to study this mix of determinants requires large-scale studies. Up unnatural loss was defined as suicide, accident or homicide in
to now, few investigations have had sufficient size to quantify the the Danish Register of Causes of Death20. A person could serve

Figure 1 Hazard ratios (Y-axis) for serious mental health outcomes based on time since bereavement (0-10 years)
as exposed in more than one sub-cohort if more than one type in accordance with the criteria of the Danish National Patient
of loss was experienced in the study period, but we included Register22 or the Psychiatric Central Research Register, which
only the first loss within each type of loss. has previously been used in Danish register studies23.
Each person experiencing loss was matched based on gen- Potential confounders or effect modifiers included in the
der and birthday (670 days) with five unexposed reference analyses were: gender, age, calendar period, degree of urbani-
persons. We ensured that each of the reference persons had a zation, history of psychiatric diagnosis, past inpatient psychiat-
relative of the same type as the one lost by the bereaved per- ric hospitalization, past deliberate self-harm, current use of
son. The matching algorithm was applied with replacements psychotropic medication, and history of selected somatic dis-
between strata. Each person was followed until one of the eases. Urbanization was classified according to the DEGURBA
studied outcomes, death, emigration, or end of study period, variable24 used by the European Union and Statistics Denmark
whichever came first. (densely, intermediately 40000, intermediately <40000, thin-
ly 15000, or thinly <15000 populated). Past psychiatric diag-
Outcome variables and data sources noses were categorized based on a five-year history in the
Psychiatric Central Research Register. Considered diagnoses
The main outcome was a serious mental health condition were: mood and anxiety disorders (ICD-10 codes F30-F48),
defined as suicide, deliberate self-harm, or psychiatric illness. schizophrenia and related disorders (F20-F29), and alcohol or
The three events were studied individually and as a composite drug abuse (F10-F19). History of psychiatric hospitalization
outcome for all four types of loss. Suicide was identified via the was coded by identifying any inpatient hospitalization in the
ICD-10 coding system (codes X60-X84) obtained from the Dan- Psychiatric Central Research Register during the five years pri-
ish Register of Causes of Death20. Psychiatric illness was defined or to entry date. Past deliberate self-harm was also considered
as any inpatient or outpatient hospitalization, or psychiatric during the five years prior to the entry date based on the earli-
emergency room contact registered in the Danish Psychiatric er defined criteria. The National Prescription Registry25 was
Central Research Register21. Deliberate self-harm was defined used to assess a one-year history of redeemed psychotropic

Table 1 Risk differences of suicide, deliberate self-harm and psychiatric illness in the first year after loss per 1,000 persons
Bereaved Non-bereaved
CIP CIP Difference

Events (95% CI) Events (95% CI) (95% CI)
Loss of child (N5501,954)
Composite 2,762 33.0 (31.8-34.2) 5,920 14.2 (13.8-14.5) 18.9 (17.6-20.1)
Suicide 30 0.36 (0.25-0.51) 66 0.16 (0.12-0.20) 0.20 (0.07-0.34)
Deliberate self-harm 1,007 12.0 (11.3-12.8) 2,641 6.3 (6.1-6.6) 5.7 (4.9-6.5)
Psychiatric illness 2,447 29.3 (28.1-30.4) 4,912 11.7 (11.4-12.1) 17.5 (16.3-18.7)
Loss of spouse (N52,242,464)
Composite 11,002 29.5 (28.9-30.0) 25,110 13.4 (13.3-13.6) 16.0 (15.4-16.6)
Suicide 279 0.74 (0.66-0.84) 204 0.11 (0.10-0.13) 0.64 (0.55-0.73)
Loss of parent (N55,312,274)
Composite 16,858 19.1 (18.8-19.3) 65,426 14.8 (14.7-14.9) 4.3 (4.0-4.6)
Suicide 187 0.21 (0.18-0.24) 533 0.16 (0.16-0.17) 0.09 (0.06-0.12)
Loss of sibling (N5615,576)
Composite 2,904 28.3 (27.3-29.3) 7,945 15.5 (15.2-15.8) 12.8 (11.8-13.9)

Suicide 21 0.20 (0.13-0.31) 75 0.15 (0.12-0.18) 0.06 (20.04 to 0.15)
CIP 2 cumulative incidence proportion
medication by the Anatomical Therapeutic Chemical codes for The hazard ratios were adjusted for degree of urbanization,
antipsychotics (N05A), anxiolytics (N05B), sedatives (N05C), or past psychiatric diagnoses, past psychiatric hospitalization, past
antidepressants (N06A). Data on somatic diseases were obtained deliberate self-harm, current use of psychotropic medication,
from the National Patient Register on the basis of diagnoses (ICD- and history of somatic diseases. In sub-analyses, hazard ratios
8/ICD-10 codes) of chronic obstructive pulmonary disease (491- were calculated on the basis of both sudden and unnatural
492/J41-J44), cancer (140-209/C00-C97), spine disorder (728/ losses and disease-related losses.
M40-M54), asthma (493/I60-I66), diabetes (249-250/E10-E14), and All data handling and statistical analyses were performed
ischemic heart disease (410-414/I20-I25). with SAS9 (SAS Institute Inc., Cary, NC, USA) and Stata 14 (Sta-
The study was approved by the Danish Data Protection taCorp LP, College Station, TX, USA).
Agency (2013-41-1719).
RESULTS
Statistical analysis
Within the study period, 83,659 persons experienced loss of
To assess the absolute risk of a serious mental health condi-
a child, 373,744 loss of the spouse, 885,379 loss of a parent,
tion, we calculated the cumulative incidence proportion for
and 102,596 loss of a sibling. The matched cohorts were five
the bereaved and non-bereaved cohorts while taking into
times larger (N5418,295, 1,868,720, 4,426,895, and 512,980,
account competing death. The hazard ratios for suicide, delib-
respectively). During 72,621,128 person-years of follow-up
erate self-harm and psychiatric illness comparing bereaved to (range 5 0-19 years), we identified 128,120 (8.9%) bereaved
non-bereaved persons were calculated by stratified Cox pro- persons and 530,026 (7.3%) non-bereaved persons (p<0.0001)
portional hazard regression with time since bereavement as who suffered from one of the three outcomes: suicide (0.1% vs.
the underlying time scale to allow for separate baseline haz- 0.06%, p<0.0001), deliberate self-harm (3.5% vs. 2.8%, p<0.0001)
ards in each matching group (one exposed and five matches). or psychiatric illness (5.3% vs. 4.5%, p<0.0001).

Figure 2 Adjusted hazard ratios (HR) and cumulative incidence proportion (CIP, %) within one year of the loss of a child or parent according
to demographic variables and health status at the time of the loss
The hazard ratio of a serious mental health condition was The cumulative incidence proportion was considerably higher
increased in the bereaved cohorts for at least 10 years after the in persons with a previous psychiatric diagnosis. In general,
loss, particularly during the first year (Figure 1). In this first about one third of these persons experienced serious mental
year, the risk difference was 18.9 events in 1,000 persons (95% health conditions during bereavement (i.e., 37% of persons
CI: 17.6-20.1) after loss of a child, 16.0 events in 1,000 persons previously diagnosed with alcohol or drug abuse who lost a
(95% CI: 15.4-16.6) after loss of the spouse, 4.3 events in 1,000 spouse; 44% of persons with a previous diagnosis of schizo-
persons (95% CI: 4.0-4.6) after loss of a parent, and 12.8 events phrenia who lost a parent). Sub-analyses revealed that sudden
in 1,000 persons (95% CI: 11.8-13.9) after loss of a sibling, unnatural loss resulted in a markedly higher risk of a serious
compared to non-bereaved persons (Table 1). Psychiatric ill-
mental health condition in the first year after bereavement
ness was the most frequent outcome.
(for all types of loss) compared to other losses (Figure 4).
When we compared bereaved with non-bereaved persons,
the overall adjusted hazard ratio at one year post-loss was 2.53
(95% CI: 2.39-2.67) for persons who lost a child, 2.14 (95% CI:
2.08-2.19) for persons who lost the spouse, 1.27 (95% CI: 1.23- DISCUSSION
1.30) for persons who lost a parent, and 1.85 (95% CI: 1.74-
1.97) for persons who lost a sibling (Figures 2 and 3). In this comprehensive nationwide cohort study, loss of a
The hazard ratio of developing a serious mental health con- close relative was associated with higher risk of suicide, delib-
dition was generally highest for 18-39 year-olds after loss of erate self-harm or psychiatric illness for up to ten years after
spouse (5.78; 95% CI: 4.70-7.10) and for 40-49 years-olds after the loss, but particularly within the first year. Risk profiles var-
loss of child (6.13; 95% CI: 5.21-7.20). The overall risk was simi- ied according to the bereaved persons relation to the deceased,
lar for males and females, except after loss of child, where age, gender, history of mental illness, and cause of death. We
females were at higher risk (hazard ratio: 2.68; 95% CI: 2.51- generally found higher risks for persons who lost a child or the
2.87) than males (hazard ratio: 2.29; 95% CI: 2.06-2.49). spouse, with a risk difference of 18.9 in 1,000 persons after loss

Figure 3 Adjusted hazard ratios (HR) and cumulative incidence proportion (CIP, %) within one year of the loss of the spouse or a sibling
according to demographic variables and health status at the time of the loss
of a child and 16.0 in 1,000 persons after loss of the spouse. Haz- highest risk of negative health consequences27,28, whereas others
ard ratios were generally highest in younger persons and after have found persons aged >60 years to be at highest risk, espe-
sudden and unnatural loss. One in three persons with a history cially of prolonged or complicated grief and suicide3,13,26,32. The
of psychiatric disorders experienced at least one of the three proportion of sudden and unnatural losses was higher in youn-
investigated outcomes within the first year of bereavement. ger age groups, whereas losses in older age were more often due
Our finding of increased risk of suicide and psychiatric illness to disease and expected deaths, which may contribute to explain
after the loss of a close relative is consistent with earlier studies, the more severe acute grief responses of the former. Age-specific
which have shown that risk is particularly high within the first vulnerabilities could also offer an explanation: younger persons
year after the loss3-6,26-29. To the best of our knowledge, this is might lack experience with loss adjustment and emotional suf-
the first large-scale study to explore the importance of the fering, which may result in susceptibility to mental illness.
bereaved persons relation to the deceased while accounting for Risk of serious mental health conditions was similar for males
gender and age. Death of the spouse has, for many years, ranked and females, yet females were at higher risk after loss of a child.
as the life event demanding the most intense readjustment Increased risk of mortality after loss of the spouse has been
when measured by the Social Readjustment Rating Scale30, but established in males26,27,33-35, while increased psychiatric mor-
recent studies with data on younger populations have suggested bidity following loss has especially been found in females2,15,36.
that loss of a child is also associated with intense and persistent Different risk profiles have been explained on the basis of differ-
grief2,31, mental illness, and suicide6. In our study, the largest ences in attachment patterns, social interaction, and coping
risk difference for developing a serious mental health condition strategies2,6,33,36: males tend to be less prone to seek help and
was actually seen in persons who lost a child. more likely to suffer from undertreated substance abuse and act
The absolute and relative risk of a serious mental health con- on impulse, which increases their risk of deliberate self-harm
dition increased with young age at the time of bereavement, and suicide33. Females tend to be more prone to rumination
except for persons who lost of child, for whom the risk peaked and react with emotional coping strategies, making them more
at the age of 40-49 years. Earlier findings have been inconsis- susceptible to anxiety, depression, and post-traumatic stress,
tent. Some studies have reported that younger spouses are at which could complicate their grief response.

Figure 4 Hazard ratios of serious mental health outcomes for persons who lost a relative due to a disease (triangles) or to an unnatural cause
of death (diamonds) according to time since bereavement (0-10 years)
Our study also showed that a history of mental illness is explain the results. However, information on reasons for contacts
associated with substantial increase in risk, as is sudden loss with psychiatric outpatient clinics or psychiatric emergency care
from suicide, accidents or homicide. Previous studies have units was not included. As severity of mental health issues may
established comorbidity between mental illness, substance vary in these contacts, the adversity could have been overesti-
abuse, and prolonged or complicated grief13,37-39, between sui- mated. Yet, only contacts with a psychiatric unit were recorded,
cide and a family history of suicidal behavior40,41, and between whereas information on persons who were treated for mental
violent deaths and increased risk of prolonged or complicated disorders in primary care was not included.
grief, mental illness, or suicide during bereavement13,42,43. Although we adjusted for several potential confounding fac-
Nevertheless, in our study, one in three bereaved with a history tors, residual confounding by unmeasured factors cannot be
of mental illness experienced a serious mental health condi- ruled out. Unfortunately, data on socio-economic factors, edu-
tion after loss; this has never previously been established while cational level and lifestyle factors were not available. However,
also adjusting for age and gender. Our finding points to the loss-induced changes in lifestyle, such as alcohol intake, diet or
role of personal vulnerability in adjustment to loss. sleeping pattern, are considered as intermediate steps on the
The sample size of this study is unparalleled by other stud- causal pathway and should not be adjusted for. Furthermore,
ies on risk of health consequences after loss and provides esti- our register-based study had no information on potentially
mates with high statistical precision, while controlling for modifying factors, such as genetic variables, family attachment
several confounders, such as history of mental or physical pattern, social network, and distress.
health, that might have been shared with the deceased family The generalizability of our findings may be limited to simi-
member and affected the health of the bereaved person. lar Western societies, with comparable health behaviors and
In the Danish registration system, the overall validity and risk factors. Yet, the estimates in this study provide significant
completeness of the records of death is close to 100%, which information on the far-reaching health consequences of famil-
ensured accurate classification of people exposed to bereave- ial loss.
ment. We followed the entire Danish population for up to 19 Serious mental health conditions and suicide after loss of a
years without loss to follow-up; thus, selection bias cannot close relative are potentially preventable13,44. Early mitigation

of risk may have wide-ranging beneficial effects, especially 13. Shear MK. Complicated grief. N Engl J Med 2015;372:153-60.
14. Maciejewski PK, Maercker A, Boelen PA et al. Prolonged grief disorder
for distinct high-risk groups. Suicide and psychiatric illness
and persistent complex bereavement disorder, but not complicated
after bereavement may be prevented by early identification of grief, are one and the same diagnostic entity: an analysis of data from the
symptom severity and adjustment problems. Future public Yale Bereavement Study. World Psychiatry 2016;15:266-75.
15. Kersting A, Brahler E, Glaesmer H et al. Prevalence of complicated grief in a
health strategies should consider policy implications of dissem-
representative population-based sample. J Affect Disord 2011;131:339-43.
inating knowledge about high-risk groups as well as strength- 16. Bonanno GA. Loss, trauma, and human resilience: have we underesti-
ening the professional competencies in assessing symptom mated the human capacity to thrive after extremely aversive events? Am
severity. Hence, more studies are needed about assessment Psychol 2004;59:20-8.
17. Stroebe W, Schut H, Stroebe MS. Grief work, disclosure and counseling: do
methods and early identification of adjustment problems. they help the bereaved? Clin Psychol Rev 2005;25:395-414.
In conclusion, this nationwide study provides the first com- 18. Erlangsen A, Fedyszyn I. Danish nationwide registers for public health and
prehensive assessment of the incidence of serious mental health-related research. Scand J Publ Health 2015;43:333-9.
19. Pedersen CB. The Danish Civil Registration System. Scand J Public Health
health conditions after the loss of a close relative. A signifi- 2011;39:22-5.
cantly elevated risk of suicide, deliberate self-harm and psy- 20. Helweg-Larsen K. The Danish Register of causes of death. Scand J Publ
chiatric illness is shown, particularly in the first year after the Health 2011;39:26-9.
21. Munk-Jorgensen P, Mortensen PB. The Danish Psychiatric Central Regis-
loss. Loss of child or spouse resulted in higher risk, and young ter. Dan Med Bull 1997;44:82-4.
age, a history of mental illness and sudden losses were found 22. Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register.
to be specific risk factors. This study points to early identifica- Scand J Public Health 2011;39:30-3.
23. Nordentoft M, Mortensen PB, Pedersen CB. Absolute risk of suicide after first
tion of high-risk persons displaying adjustment problems in hospital contact in mental disorder. Arch Gen Psychiatry 2011;68:1058-64.
order to mitigate distress and reduce the risk of serious mental 24. Statistics Denmark. Degree of urbanization. Copenhagen: Statistics Den-
health conditions after loss of a close family member. mark, 2016.
25. Kildemoes HW, Sorensen HT, Hallas J. The Danish National Prescription
Registry. Scand J Publ Health 2011;39:38-41.
26. Erlangsen A, Jeune B, Bille-Brahe U et al. Loss of partner and suicide risks
ACKNOWLEDGEMENTS among oldest old: a population-based register study. Age Ageing 2004;33:
378-83.
This study was supported by an unrestricted grant from the Lundbeck Founda-
27. Manor O, Eisenbach Z. Mortality after spousal loss: are there socio-demo-
tion (no. R155-2012-11280), TrygFonden, and the Danish Cancer Society. J. Li
graphic differences? Soc Sci Med 2003;56:405-13.
was supported by the European Research Council (ERC-2010-StG 2010-
260242-ProgEuro), Danish Council for Independent Research (project no. 28. Schaefer C, Quesenberry CP Jr, Wi S. Mortality following conjugal bereavement
6110-00019A) and the Nordic Cancer Union (2015-176673). The funders had and the effects of a shared environment. Am J Epidemiol 1995;141:1142-52.
no role in study design, data collection and analysis, decision to publish, or 29. Kaltman S, Bonanno GA. Trauma and bereavement: examining the impact
preparation of the manuscript. of sudden and violent deaths. J Anxiety Disord 2003;17:131-47.
30. Holmes TH, Rahe RH. The social readjustment rating scale. J Psychosom
Res 1967;11:213-8.
31. Zetumer S, Young I, Shear MK et al. The impact of losing a child on the
clinical presentation of complicated grief. J Affect Disord 2015;170:15-21.
REFERENCES 32. Elwert F, Christakis NA. Widowhood and race. Am Sociol Rev 2006;71:16-41.
1. Wilcox HC, Mittendorfer-Rutz E, Kjeldgard L et al. Functional impairment 33. Agerbo E. Midlife suicide risk, partners psychiatric illness, spouse and
due to bereavement after the death of adolescent or young adult offspring child bereavement by suicide or other modes of death: a gender specific
in a national population study of 1,051,515 parents. Soc Psychiatry Psy- study. J Epidemiol Community Health 2005;59:407-12.
chiatr Epidemiol 2015;50:1249-56. 34. Espinosa J, Evans WN. Heightened mortality after the death of a spouse:
2. Stroebe M, Schut H, Stroebe W. Health outcomes of bereavement. Lancet marriage protection or marriage selection? J Health Econ 2008;27:1326-42.
2007;370:1960-73. 35. Li G. The interaction effect of bereavement and sex on the risk of suicide
3. Christakis NA, Allison PD. Mortality after the hospitalization of a spouse. in the elderly: an historical cohort study. Soc Sci Med 1995;40:825-8.
N Engl J Med 2006;354:719-30. 36. Stroebe M. Gender differences in adjustment to bereavement: an empiri-
4. Elwert F, Christakis NA. The effect of widowhood on mortality by the cal and theoretical review. Rev Gen Psychol 2001;5:62-83.
causes of death of both spouses. Am J Publ Health 2008;98:2092-8. 37. Lobb EA, Kristjanson LJ, Aoun SM et al. Predictors of complicated grief: a
5. Li J, Precht DH, Mortensen PB et al. Mortality in parents after death of a systematic review of empirical studies. Death Stud 2010;34:673-98.
child in Denmark: a nationwide follow-up study. Lancet 2003;361:363-7. 38. Melhem NM, Rosales C, Karageorge J et al. Comorbidity of axis I disorders
6. Li J, Laursen TM, Precht DH et al. Hospitalization for mental illness among in patients with traumatic grief. J Clin Psychiatry 2001;62:884-7.
parents after the death of a child. N Engl J Med 2005;352:1190-6. 39. Simon NM, Pollack MH, Fischmann D et al. Complicated grief and its cor-
7. Qin P, Mortensen PB. The impact of parental status on the risk of complet- relates in patients with bipolar disorder. J Clin Psychiatry 2005;66:1105-10.
ed suicide. Arch Gen Psychiatry 2003;60:797-802. 40. Guldin MB, Li J, Pedersen HS et al. Incidence of suicide among persons
8. Wilcox HC, Kuramoto SJ, Brent D et al. The interaction of parental history of who had a parent who died during their childhood: a population-based
suicidal behavior and exposure to adoptive parents psychiatric disorders on cohort study. JAMA Psychiatry 2015;72:1227-34.
adoptee suicide attempt hospitalizations. Am J Psychiatry 2012;169:309-15. 41. Qin P, Agerbo E, Mortensen PB. Suicide risk in relation to family history of
9. Bolton JM, Au W, Chateau D et al. Bereavement after sibling death: a completed suicide and psychiatric disorders: a nested case-control study
population-based longitudinal case-control study. World Psychiatry 2016; based on longitudinal registers. Lancet 2002;360:1126-30.
15:59-66. 42. Tal Young I, Iglewicz A, Glorioso D et al. Suicide bereavement and compli-
10. Bonanno GA, Neria Y, Mancini A et al. Is there more to complicated grief cated grief. Dialogues Clin Neurosci 2012;14:177-86.
than depression and posttraumatic stress disorder? A test of incremental 43. van Denderen M, de Keijser J, Kleen M et al. Psychopathology among
validity. J Abnorm Psychol 2007;116:342-51. homicidally bereaved individuals: a systematic review. Trauma Violence
11. Maercker A, Brewin CR, Bryant RA et al. Diagnosis and classification of Abuse 2015;16:70-80.
disorders specifically associated with stress: proposals for ICD-11. World 44. Nordentoft M, Madsen T, Fedyszyn I. Suicidal behavior and mortality in
Psychiatry 2013;12:198-206. first-episode psychosis. J Nerv Ment Dis 2015;203:387-92.
12. Prigerson HG, Horowitz MJ, Jacobs SC et al. Prolonged grief disorder: psy-
chometric validation of criteria proposed for DSM-V and ICD-11. PLoS
Med 2009;6:e1000121. DOI:10.1002/wps.20422

REAPPRAISAL
A critique of the ultra-high risk and transition paradigm

Jim van Os1,2, Sinan Guloksuz1,3
1
Department of Psychiatry and Psychology, Maastricht University Medical Centre, Maastricht, the Netherlands; 2Kings College London, Kings Health Partners, Department of
Psychosis Studies, Institute of Psychiatry, London, UK; 3Department of Psychiatry, Yale University School of Medicine, New Haven, CT, USA
The transdiagnostic expression of psychotic experiences in common mental disorder (anxiety/depression/substance use disorder) is associated
with a poorer prognosis, and a small minority of people may indeed develop a clinical picture that meets criteria for schizophrenia. However,
it appears neither useful nor valid to observe early states of multidimensional psychopathology in young people through the schizo-prism,
and apply misleadingly simple, unnecessary and inefficient binary concepts of risk and transition. A review of the ultra-high risk (UHR)
or clinical high risk (CHR) literature indicates that UHR/CHR samples are highly heterogeneous and represent individuals diagnosed with
common mental disorder (anxiety/depression/substance use disorder) and a degree of psychotic experiences. Epidemiological research has
shown that psychotic experiences are a (possibly non-causal) marker of the severity of multidimensional psychopathology, driving poor out-
come, yet notions of risk and transition in UHR/CHR research are restrictively defined on the basis of positive psychotic phenomena alone,
ignoring how baseline differences in multidimensional psychopathology may differentially impact course and outcome. The concepts of risk
and transition in UHR/CHR research are measured on the same dimensional scale, yet are used to produce artificial diagnostic shifts. In
fact, transition in UHR/CHR research occurs mainly as a function of variable sample enrichment strategies rather than the UHR/CHR
criteria themselves. Furthermore, transition rates in UHR/CHR research are inflated as they do not exclude false positives associated with the
natural fluctuation of dimensional expression of psychosis. Biological associations with transition thus likely represent false positive find-
ings, as was the initial claim of strong effects of omega-3 polyunsatured fatty acids in UHR samples. A large body of UHR/CHR intervention
research has focused on the questionable outcome of transition, which shows lack of correlation with functional outcome. It may be more
productive to consider the full range of person-specific psychopathology in all young individuals who seek help for mental health problems,
instead of policing youngsters for the transdiagnostic dimension of psychosis. Instead of the relatively inefficient medical high-risk approach,
a public health perspective, focusing on improved access to a low-stigma, high-hope, small scale and youth-specific environment with accept-
able language and interventions may represent a more useful and efficient strategy.
Key words: Ultra-high risk, transition, psychotic experiences, common mental disorder, transdiagnostic expression of psychosis, public
health perspective
Over the last two decades, more than 1,500 studies have and outcome that lie on the same unidimensional scale, and
been published revolving around the concept of ultra-high obscure the temporality and dynamics of multidimensional
risk (UHR) or clinical high risk (CHR) for transition to a psychopathological states in young people.
psychotic disorder. The basic assumptions behind these studies We do not wish to dispute that it is better to intervene early
are as follows: in a group of young people seeking help for men- rather than late. Rather, we wish to argue that it is conceptual-
tal problems, one can apply criteria for a binary risk diagnosis ly flawed to frame the treatment of early psychopathology in
predicting schizophrenia spectrum disorder, and true positives diagnosed help-seeking individuals as prevention of psychotic
are people that meet criteria for transition at follow-up. disorder, just because there is some degree of transdiagnostic
Reviews of UHR/CHR studies tend to be upbeat, taking the expression of psychotic experiences.
shape of evidence-based recommendations or guidance, stat-
ing that the young field of preventive research in psychosis has
already resulted in sufficient evidence to formulate recommenda- CLINICAL HIGH RISK SAMPLING IS SELECTIVE AND
tions for an early detection of psychosis in the clinical practice1, NON-EPIDEMIOLOGICAL
and that psychological, in particular cognitive-behavioural, as
well as pharmacological interventions are able to prevent or at In practice, studies that want to apply the UHR/CHR paradigm
least postpone a first psychotic episode in adult CHR patients2. have to search for young individuals who are slightly-but-not-
However, the question arises of the degree to which this opti- quite psychotic and have also expressed a wish to receive help.
mism is based on logical reasoning and scientific evidence. Sampling strategies differ widely from study to study and are
There is a growing literature on the complexities underlying based on a mix of advertising, service filters and active searches,
UHR/CHR research, that are not resolved, clouding the inter- thus per definition resulting in selected, non-representative sam-
pretation of data3-11. In this paper, we critically review the as- ples that cannot readily be compared across studies.
sumptions underlying UHR/CHR research. In particular, we For example, in the North-American multicentre prediction
focus on outstanding issues to do with sampling variability and study12, it was stated that each site recruited potential subjects
basic epidemiological parameters, the fixation on psychosis at through clinical referrals as stimulated by talks to school coun-
the expense of other psychopathology, and the lack of transpar- selors and mental health professionals in community settings.
ency arising from the use of two binary concepts for diagnosis In the European Prediction of Psychosis Study (EPOS)13, UHR/

CHR sampling was described as follows: knowledge about ear- admixture is important, as it provides a crucial link to the epi-
ly warning signs (e.g., concentration and attention disturban- demiological literature with findings derived from representa-
ces, unexplained functional decline) and inclusion criteria was tive population-based samples. Attenuated psychotic symp-
disseminated (through local workshops, articles in professional toms at the population level are closely associated with non-
journals and newsletters, informational flyers, and web sites) to psychotic diagnoses and/or sub-diagnostic non-psychotic psy-
mental health professionals as well as institutions and persons chopathology including anxiety, depression, attention-deficit/
who might be contacted by at-risk persons seeking help. hyperactivity disorder, post-traumatic stress disorder, sub-
Of the two largest CHR psychotherapy trials to date, one did stance use disorder, eating disorder and many other forms of
not provide details about the sampling procedure except that psychopathology21. Psychosis can thus be regarded as a trans-
it took screening of 5,705 subjects to include 201 patients diagnostic dimension of psychopathology22.
(3.5%) in the trial14 and the other described sampling as fol- Epidemiological studies show that the presence of attenuat-
lows: our ascertainment strategy was to make services famil- ed psychotic symptoms in non-psychotic disorders is associat-
iar with our entry criteria and to liaise on a regular basis; no ed with greater severity and poorer response to treatment23-26.
In fact, research has shown that more exposure to genetic and
systematic screening of service populations was carried out15.
environmental risk factors is associated with more severe non-
What becomes clear is that CHR studies have to invest a great
psychotic psychopathology which in turn is associated with a
deal of resources in detecting and sampling subjects who meet
greater probability of the person also having some degree of
the inclusion criteria. The cost of finding rare UHR/CHR sub-
expression of psychosis24,27,28.
jects is considerable, but not included in cost-effectiveness analy-
ses of UHR/CHR research. Given the apparent rarity of UHR/CHR
states, it becomes a priori unlikely that early intervention along
the UHR/CHR paradigm will have public health impact. A recent PSYCHOTIC EXPERIENCES IN NON-PSYCHOTIC
study, investigating an early intervention service in an inner city DISORDER: MARKER OR CAUSE OF POOR PROGNOSIS?
area, found that only a tiny proportion (4.1%) of patients with a
first-episode psychotic disorder attending mental health services Psychotic experiences can thus be considered a marker for the
had been in previous contact with the local prodromal service, severity of non-psychotic states. However, it may not be valid to
indicating that the impact of prodromal services in public health see them as causal for a poor prognosis, as the evidence shows
terms may be negligible in relation to their costs16. Such a lack of that psychosis may also be considered as something that follows
impact associated with the high-risk approach is a well-known passively as a function of the general severity of multidimensional
phenomenon, referred to as the prevention paradox17. psychopathology22. This is essential with regard to the UHR/CHR
Given the absence of a consistent sampling frame, it is un- framework, where the clinical focus is solely on the binary risk
likely that CHR samples are readily comparable from study to concept of psychosis (risk and transition, measured on the
study. For example, samples differ widely in exclusion criteria same dimensional scale), while the multidimensional severity of
regarding previous use of antipsychotics and mood stabilizers, the psychopathological context is ignored. In the UHR/CHR
framework, the binary presence of psychotic experiences, under
previous episodes of mania, and previous drug-induced psy-
the implicit assumption of impending, mostly schizo outcome29,
chotic states. Therefore, referring to CHR patients as if they
trumps all other dimensional expressions of psychopathology.
were a class is not warranted. Although many meta-analyses
A whole generation of UHR/CHR studies has been analyzed
of UHR/CHR samples have been conducted, the question
from the perspective that outcome of common mental disor-
arises whether these studies are sufficiently similar. der with a degree of psychosis admixture is best predicted on
Nevertheless, two issues appear to be consistent across the basis of a binary psychosis risk criterion. An alternative
UHR/CHR samples. The first is that these samples in essence hypothesis, however, is that outcome in these states is in fact a
consist of individuals with a current diagnosis of mainly anxi- consequence of baseline severity of multidimensional psychopa-
ety, depression or substance use18,19. The second is that, of the thology rather than a binary psychosis risk criterion (Figure 1).
various CHR criteria, the attenuated symptom defines the Studies that have looked beyond UHR/CHR criteria confirm this
great majority of individuals20, the others having minimal rele- prediction13,30-34. In other words, what is presented as risk
vance. In other words, CHR samples are individuals with com- may be better summarized as baseline differences in the severity
mon mental disorder or a substance use disorder who also of multidimensional psychopathology.
present with low-grade psychotic symptoms.
DOES THE CONCEPT OF TRANSITION REPRESENT

CLINICAL HIGH RISK 5 COMMON MENTAL A QUALITATIVE SHIFT?
DISORDER WITH SUBTLE PSYCHOSIS ADMIXTURE
In UHR/CHR research, high risk and transition are typical-
The fact that UHR/CHR samples in fact consist of individu- ly measured on the same dimensional scale rating frequency/
als with anxiety/depression/substance use with subtle psychosis duration of attenuated positive psychotic symptoms, usually the

Figure 1 Relative blindness of the ultra-high risk (UHR)/transition paradigm. On the left, the natural development of multidimensional psy-
chopathology over time. Black circles indicate (attenuated) positive psychotic symptoms. Other gray-scale circles indicate other psychopathol-
ogy. As the UHR paradigm ignores multidimensional psychopathology, it remains blind and only sees psychotic phenomena as precursors
of schizo-transition (i.e., more severe psychosis; below on the right), while these phenomena are in fact a marker of relative poor outcome of
multidimensional psychopathology (below on the left). The restricted focus on positive symptoms in the UHR paradigm means that consider-
able potential for prevention in phases 1-4 is missed.
Comprehensive Assessment of At-Risk Mental States (CAARMS)35 ingly important qualitative diagnostic change: as the attenuat-
or the Scale of Prodromal Symptoms (SOPS)36. These frequency/ ed psychotic symptoms in the UHR/CHR state cannot be
duration ratings appear either impossibly precise (e.g., at least counted as a full psychotic symptom in the DSM/ICD diag-
once a month to twice a week more than one hour per occa- nostic system, the diagnosis in the UHR/CHR risk state re-
sion, or at least 3 to 6 times a week less than one hour per mains per definition non-psychotic. However, with the
occasion) or rather broad (e.g., present for at least 1 week and dimensional shift in the CAARMS/SIPS towards transition,
no longer than 5 years). The scales for positive symptoms range the attenuated psychotic symptom can now be used as a true
from 0 to 6, where 3-5, for example, represents risk for psy- psychotic symptom, automatically resulting in a diagnosis of
chosis and 6 represents psychosis. Other symptom domains psychotic disorder in DSM/ICD. Thus, dimensional shifts are
are ignored, regardless of their severity. Transition can be pre- used to evoke the notion that a diagnosis is born, creating
sent with a 1-point shift on the dimensional scale, thus repre- the suggestion of a qualitative distinction.
senting a quantitative, not a qualitative shift from risk to
transition status.
While UHR/CHR criteria are generally clearly described in IS TRANSITION CONFOUNDED BY NATURAL
the literature, accounts of transition are usually kept vague. FLUCTUATION OF DIMENSIONAL EXPRESSION OF
For example, in one recent large UHR/CHR trial15, transition PSYCHOSIS?
was described as operationally defined on the CAARMS using
the recommended criteria of a global rating scale score of 6 on Given the fact that transition in fact represents a dimen-
either unusual thought content, non-bizarre ideas, or disor- sional shift, false positive ratings of transition are likely to
ganised speech, or 5-6 on perceptual abnormalities, with an occur given the natural fluctuation in severity of the transdiag-
associated frequency score of 4-6, and with these experiences nostic psychosis dimension within and between individuals22.
lasting longer than one week15. In another trial14, it was sim- The only study to date that attempted to reduce false posi-
ply stated that the primary outcome of this study was the tive ratings of transition by serial examination of individuals,
transition to psychosis; the transition is defined by the excluding individuals rated as UHR that in fact were in a natu-
CAARMS criteria. Considering the importance of valid out- ral low of a clinical psychotic syndrome, reported a 2-year
comes in randomized controlled trials, these descriptions are transition rate of 8%15, well below the meta-analytical estimate
opaque and appear to rely on small dimensional shifts. These of 19% in studies that did not attempt to exclude such false
shifts are nevertheless subsequently transformed into a seem- positive ratings2.

IS THE CONCEPT OF TRANSITION RELEVANT? biological research on the diagnosis of schizophrenia. These
studies have reported a range of biological associations with
There is a lack of research on the clinical relevance of the transition, published in high-impact academic journals. For
transition outcome37. However, evidence from long-term fol- example, studies have reported that transition to psychosis
low-up studies suggests that the binary transition concept is was associated with thalamic dysconnectivity44, progressive
not particularly relevant in terms of predicting clinical and func- reduction of cortical thickness45, and increased glutamate lev-
tional outcome, and that other symptom domains (affective, els in the associative striatum46.
cognitive, negative but also how mixed and how severe psy- Given the uncertain status of the transition concept, these
chopathology is) are more impactful in this respect13,32,38,39. findings cannot be readily interpreted and appear to be false
This observation is supported by the fact that meta-analyses positives unless true, rather than approximate, replication is
of UHR/CHR intervention studies, focussing on the prevention attempted47. Analogously, one trial reported an apparently very
of transition, fail to show effect on functional outcome1. strong effect of fish oil in reducing transition rates48, which
became an informative null finding in the replication study49.
THE TRUE TRANSITION RATE OF ATTENUATED

PSYCHOTIC SYMPTOMS IS <1%: THE ROLE OF DOES UHR/CHR REPRESENT A VALID AND USEFUL
SAMPLING ENRICHMENT SURROGATE FOR EARLY INTERVENTION?
A common and persisting misunderstanding is that the To lay the groundwork for the current UHR/CHR construct,
risk function in UHR/CHR research is caused by the UHR/ the architects of the construct started with reviewing the previous
CHR criteria themselves. However, already more than a decade literature of the prodromal phase: narratives, early depictions,
ago, it was pointed out that high risk for transition does not so frequency and pattern of formation of signs and symptoms. This
much depend on UHR/CHR criteria themselves, but rather on comprehensive review of the prodromal period clearly showed
the way the sampling procedures ensure progressive enrich- that non-psychotic symptoms concentration difficulties, moti-
ment in risk4,40. Thus, the true yearly transition rate of attenu- vational impairment, depressed mood, sleep disturbance, and
ated psychotic symptoms in the general population, estab- anxiety frequently emerge prior to onset of psychotic symp-
lished in a meta-analysis of representative, population-based toms50. However, these symptoms were considered not specific
samples, is less than 1%41. The fact that the transition rate is enough to target with a therapeutic intervention, because the
much higher in UHR/CHR samples, similarly defined by the main driving force was to reproduce successful medical models
presence of attenuated psychotic symptoms20, has to do with of indicated prevention for schizophrenia.
the sampling strategies in UHR/CHR research. A recent meta- This was a hazardous pursuit for several reasons. First, early
analysis showed that the CHR sampling risk enrichment strat- detection and intervention in psychiatry cannot be easily fit
egy occasioned a 3-year transition rate of 15%42, thus account- into the framework of preventive medicine, because: a) natural
ing for half of the most recent meta-analytical 3-year tran- history and underlying biological mechanisms of mental disor-
sition rate of 29% attributed to CHR criteria2. Other reasons for ders have yet to be understood; b) there are no objective
the inflated transition rates in UHR/CHR research (e.g., natural screening tools; c) there is no specific treatment. Second,
fluctuation) were discussed earlier. UHR/CHR is conceptualized after schizophrenia, which is a
Direct evidence that the transition rate is caused by sam- classic case of the no true Scotsman fallacy, as formulated by
pling enrichment and not CHR criteria came from a study in an Robins and Guze51: good prognosis schizophrenia is not
early psychosis service for young people, showing that young mild schizophrenia, but a different illness. From this perspec-
people presenting to the service meeting UHR criteria had tive, setting the goal of preventing transition to schizophre-
essentially the same 10-year transition rate (17.3%) as young nia by intervening at the level of UHR/CHR creates a paradox,
people presenting to the same service with non-psychotic dis- or even a self-fulfilling prophecy of failure. Third, there is a
orders (14.6%)43. degree of tautology in the claim that an intervention specific to
positive symptoms the initial research agenda of prodromal
research was antipsychotic trials in the UHR/CHR population
DOES BIOLOGICAL RESEARCH OF TRANSITION shall prevent transition to psychosis by reducing positive
MAKE SENSE? symptoms in UHR/CHR states that are primarily defined on
the basis of milder positive symptoms. This can be likened to
Given the attractive binary outcome of transition, a range of saying that increased cholesterol would be reduced by anti-
biological studies have attempted to find differences between cholesterol treatment to prevent high cholesterol.
those who do and those who do not make a transition, resem- Perhaps not surprisingly, findings of UHR/CHR studies
bling the classical case-control paradigm that has dominated have confirmed what could have been expected: the pragmatic

UHR/CHR construct overlooking early expression of nonspe- CONCLUSIONS
cific psychopathology (Figure 1) indeed backfires on early
detection and intervention. A retrospective investigation52 of Early intervention is a progressive movement and should be
the population of the psychiatric case register in The Hague, supported. However, the CHR-cum-transition concept is overly
the Netherlands, revealed that over half of the patients who simplified and uncritically presented as evidence. The tools
developed psychosis had received treatment for non-psychotic solely rely on positive symptoms and a family history of psy-
conditions (mood, anxiety and substance use disorders) dur- chotic disorders. The implicit paradigm is to treat any sub-
ing the prodromal phase, revealing a lot more prevention threshold positive symptom as a pathway to schizophrenia.
potential than the negligible percentage of the prodromal ser- Currently, less emphasis is put on antipsychotic treatment,
vice, that is limited by the prevention paradox16,17. Similarly, which is a good point. However, the transition concept is not
the vast majority of the North American UHR/CHR cohort just fuzzy but overreaching, and should not be used as an
had received psychosocial or pharmacological treatment long outcome in research or clinical practice.
before the onset of subthreshold symptoms53,54. These find- It may be asked why, if this is the state of the evidence, the
ings bring into question the utility of UHR/CHR concept: how CHR-cum-transition concept continues to be pushed in re-
early is early intervention? search and clinical practice. In two separate articles, Schmidt
et al1 and Schultze-Lutter et al2 appear to provide guidance on
CHR research and clinical practice on behalf of the European
Psychiatric Association. In these days of heightened awareness
SHOULD TREATMENT FOCUS ON PREVENTION OF of the role of not just commercial, but also academic funding, as
TRANSITION? well as other interests in research57, and the vagaries of research
in small and selected samples, the meta-analysis of which does
There is no doubt that it is useful to offer early treatment to
not resolve the issue of multiple sources of bias58,59, one would
young individuals with anxiety/depression/substance use and a expect guidance by professional bodies to be critical and objec-
degree of psychosis admixture as a marker of relatively poor tive. It may be more useful to reserve journal space for academic
prognosis. It may be expected that non-specific psychotherapeu- debate, rather than uncritically perpetuating fashionable re-
tic interventions will be beneficial, similar to the non-specific search notions and the academic interests that come with it.
effects of a range of psychotherapies in anxiety/depression55. For Instead of the medical, relatively inefficient high-risk ap-
example, there is evidence that simple interventions such as non- proach, a public health perspective, focusing on improved
directive listening yield better results than cognitive-behavioural access to a low-stigma, high-hope, small-scale and youth-
therapy in UHR/CHR individuals56. specific environment with acceptable language and interven-
There is a body of intervention research, consisting of most- tions, as embedded in the recent Headspace initiative60, may
ly small, highly heterogeneous and variably controlled studies, represent a more useful and more efficient strategy61.
focusing on the outcome of transition in UHR/CHR individ-
uals1. However, given the questionable validity and clinical rel-
evance of the transition concept, coupled with the fact that REFERENCES
these interventions do no impact functioning1, there seems to
1. Schmidt SJ, Schultze-Lutter F, Schimmelmann BG et al. EPA guidance on
be an urgent need to reconceptualize and reorient treatment the early intervention in clinical high risk states of psychoses. Eur Psychia-
strategies in individuals with anxiety/depression/substance try 2015;30:388-404.
use and a degree of psychosis admixture as a marker of rela- 2. Schultze-Lutter F, Michel C, Schmidt SJ et al. EPA guidance on the early
detection of clinical high risk states of psychoses. Eur Psychiatry 2015;30:
tively poor prognosis. 405-16.
The available evidence suggests that the tradition to observe 3. Kablinger AS, Freeman AM 3rd. Prodromal schizophrenia and atypical
antipsychotic treatment. J Nerv Ment Dis 2000;188:642-52.
these states through the schizo-prism may be not useful and
4. Van Os J, Delespaul P. Toward a world consensus on prevention of schizo-
ethically questionable. Instead, it may be more productive to phrenia. Dialogues Clin Neurosci 2005;7:53-67.
consider the full range of person-specific psychopathology 5. Lacluyse K, van Bouwel L, Demunter H et al. Clinical assessment of the
ultra high risk of developing a psychotic disorder; review and critical
in all young individuals with mental health problems and to reflection. Tijdschr Psychiatr 2011;53:153-62.
not become disproportionally fixated on the transdiagnostic 6. Amos A. Assessing the cost of early intervention in psychosis: a systematic
manifestation of psychosis. Although psychotic experiences review. Aust N Z J Psychiatry 2012;46:719-34.
7. Marshall C, Addington J, Epstein I et al. Treating young individuals at clini-
in common mental disorder may be associated with a poorer cal high risk for psychosis. Early Interv Psychiatry 2012;6:60-8.
prognosis, and a small minority of people may indeed develop a 8. Fusar-Poli P, Van Os J. Lost in transition: setting the psychosis threshold in
clinical picture that meets criteria for schizophrenia, it appears prodromal research. Acta Psychiatr Scand 2013;127:248-52.
9. Amos AJ. Evidence that treatment prevents transition to psychosis in
neither useful nor scientifically valid to reduce the transdiagnos- ultra-high-risk patients remains questionable. Schizophr Res 2014;153:
tic expression of psychosis in early states of multidimensional 240.
10. Simon AE, Umbricht D, Lang UE et al. Declining transition rates to
psychopathology to the misleadingly simple binary concepts of
psychosis: the role of diagnostic spectra and symptom overlaps in
risk and transition, with the implicit suggestion that all or individuals with attenuated psychosis syndrome. Schizophr Res 2014;159:
most psychosis leads to schizophrenia. 292-8.

11. Mittal VA, Dean DJ, Mittal J et al. Ethical, legal, and clinical considerations 35. Yung AR, Yuen HP, McGorry PD et al. Mapping the onset of psychosis: the
when disclosing a high-risk syndrome for psychosis. Bioethics 2015;29: comprehensive assessment of at-risk mental states. Aust N Z J Psychiatry
543-56. 2005;39:964-71.
12. Cannon TD, Cadenhead K, Cornblatt B et al. Prediction of psychosis in 36. Miller TJ, McGlashan TH, Rosen JL et al. Prospective diagnosis of the ini-
youth at high clinical risk: a multisite longitudinal study in North America. tial prodrome for schizophrenia based on the Structured Interview for Pro-
Arch Gen Psychiatry 2008;65:28-37. dromal Syndromes: preliminary evidence of interrater reliability and
13. Salokangas RK, Heinimaa M, From T et al. Short-term functional outcome predictive validity. Am J Psychiatry 2002;159:863-5.
and premorbid adjustment in clinical high-risk patients. Results of the 37. Simon AE, Velthorst E, Nieman DH et al. Ultra high-risk state for psychosis
EPOS project. Eur Psychiatry 2014;29:371-80. and non-transition: a systematic review. Schizophr Res 2011;132:8-17.
14. van der Gaag M, Nieman DH, Rietdijk J et al. Cognitive behavioral therapy 38. Lin A, Wood SJ, Nelson B et al. Neurocognitive predictors of functional
for subjects at ultrahigh risk for developing psychosis: a randomized con- outcome two to 13 years after identification as ultra-high risk for psycho-
trolled clinical trial. Schizophr Bull 2012;38:1180-8. sis. Schizophr Res 2011;132:1-7.
15. Morrison AP, French P, Stewart SL et al. Early detection and intervention 39. Brandizzi M, Valmaggia L, Byrne M et al. Predictors of functional outcome
evaluation for people at risk of psychosis: multisite randomised controlled in individuals at high clinical risk for psychosis at six years follow-up.
trial. BMJ 2012;344:e2233. J Psychiatr Res 2015;65:115-23.
16. Ajnakina O. First episode psychosis: looking backwards and forwards. 40. Simon AE, Roth B, Zmilacher S et al. Developing services for the early
https://kclpure.kcl.ac.uk. detection of psychosis: a critical consideration of the current state of the
17. Rose G. Strategy of prevention: lessons from cardiovascular disease. BMJ art. Eur Child Adolesc Psychiatry 2007;16:96-103.
(Clin Res Ed) 1981;282:1847-51. 41. Kaymaz N, Drukker M, Lieb R et al. Do subthreshold psychotic experien-
18. Addington J, Case N, Saleem MM et al. Substance use in clinical high risk for ces predict clinical outcomes in unselected non-help-seeking population-
psychosis: a review of the literature. Early Interv Psychiatry 2014;8:104-12. based samples? A systematic review and meta-analysis, enriched with new
19. Fusar-Poli P, Nelson B, Valmaggia L et al. Comorbid depressive and anxi- results. Psychol Med 2012;42:2239-53.
ety disorders in 509 individuals with an at-risk mental state: impact on 42. Fusar-Poli P, Schultze-Lutter F, Cappucciati M et al. The dark side of the
psychopathology and transition to psychosis. Schizophr Bull 2014;40: moon: meta-analytical impact of recruitment strategies on risk enrichment
120-31. in the clinical high risk state for psychosis. Schizophr Bull 2016;42:732-43.
20. Fusar-Poli P, Cappucciati M, Borgwardt S et al. Heterogeneity of psychosis 43. Conrad AM, Lewin TJ, Sly KA et al. Utility of risk-status for predicting psy-
risk within individuals at clinical high risk: a meta-analytical stratification. chosis and related outcomes: evaluation of a 10-year cohort of presenters
JAMA Psychiatry 2016;73:113-20. to a specialised early psychosis community mental health service. Psychi-
21. Linscott RJ, van Os J. An updated and conservative systematic review and atry Res 2017;247:336-44.
meta-analysis of epidemiological evidence on psychotic experiences in 44. Anticevic A, Haut K, Murray JD et al. Association of thalamic dysconnec-
children and adults: on the pathway from proneness to persistence to tivity and conversion to psychosis in youth and young adults at elevated
dimensional expression across mental disorders. Psychol Med 2013;43: clinical risk. JAMA Psychiatry 2015;72:882-91.
1133-49. 45. Cannon TD, Chung Y, He G et al. Progressive reduction in cortical thick-
22. van Os J, Reininghaus U. Psychosis as a transdiagnostic and extended phe- ness as psychosis develops: a multisite longitudinal neuroimaging study
notype in the general population. World Psychiatry 2016;15:118-24. of youth at elevated clinical risk. Biol Psychiatry 2015;77:147-57.
23. Perlis RH, Uher R, Ostacher M et al. Association between bipolar spectrum 46. de la Fuente-Sandoval C, Leon-Ortiz P, Azcarraga M et al. Striatal gluta-
features and treatment outcomes in outpatients with major depressive mate and the conversion to psychosis: a prospective 1H-MRS imaging
disorder. Arch Gen Psychiatry 2011;68:351-60. study. Int J Neuropsychopharmacol 2013;16:471-5.
24. Kelleher I, Keeley H, Corcoran P et al. Clinicopathological significance of 47. Maxwell SE. The persistence of underpowered studies in psychological
psychotic experiences in non-psychotic young people: evidence from four research: causes, consequences, and remedies. Psychol Methods 2004;9:
population-based studies. Br J Psychiatry 2012;201:26-32. 147-63.
25. Wigman JT, van Nierop M, Vollebergh WA et al. Evidence that psychotic 48. Amminger GP, Schafer MR, Papageorgiou K et al. Long-chain omega-3 fat-
symptoms are prevalent in disorders of anxiety and depression, impacting ty acids for indicated prevention of psychotic disorders: a randomized,
on illness onset, risk, and severity implications for diagnosis and ultra- placebo-controlled trial. Arch Gen Psychiatry 2010;67:146-54.
high risk research. Schizophr Bull 2012;38:247-57. 49. McGorry PD, Nelson B, Markulev C et al. Effect of omega-3 polyunsaturat-
26. Wigman JT, van Os J, Abidi L et al. Subclinical psychotic experiences and ed fatty acids in young people at ultrahigh risk for psychotic disorders: the
bipolar spectrum features in depression: association with outcome of psy- NEURAPRO randomized clinical trial. JAMA Psychiatry 2017;74:19-27.
chotherapy. Psychol Med 2014;44:325-36. 50. Yung AR, McGorry PD. The prodromal phase of first-episode psychosis:
27. Guloksuz S, van Nierop M, Lieb R et al. Evidence that the presence of psy- past and current conceptualizations. Schizophr Bull 1996;22:353-70.
chosis in non-psychotic disorder is environment-dependent and mediat- 51. Robins E, Guze SB. Establishment of diagnostic validity in psychiatric ill-
ed by severity of non-psychotic psychopathology. Psychol Med 2015;45: ness: its application to schizophrenia. Am J Psychiatry 1970;126:983-7.
2389-401. 52. Rietdijk J, Hogerzeil SJ, van Hemert AM et al. Pathways to psychosis: help-
28. van Nierop M, Viechtbauer W, Gunther N et al. Childhood trauma is asso- seeking behavior in the prodromal phase. Schizophr Res 2011;132:213-9.
ciated with a specific admixture of affective, anxiety, and psychosis symp- 53. Woods SW, Addington J, Bearden CE et al. Psychotropic medication use in
toms cutting across traditional diagnostic boundaries. Psychol Med 2015; youth at high risk for psychosis: comparison of baseline data from two
45:1277-88. research cohorts 1998-2005 and 2008-2011. Schizophr Res 2013;148:99-104.
29. Woods SW, Addington J, Cadenhead KS et al. Validity of the prodromal risk 54. Woodberry KA, Seidman LJ, Bryant C et al. Treatment precedes positive
syndrome for first psychosis: findings from the North American Prodrome symptoms in North American adolescent and young adult clinical high
Longitudinal Study. Schizophr Bull 2009;35:894-908. risk cohort. J Clin Child Adolesc Psychol 2016;5:1-10.
30. Cannon TD, Cadenhead K, Cornblatt B et al. Prediction of psychosis in 55. Cuijpers P, Donker T, van Straten A et al. Is guided self-help as effective as
youth at high clinical risk: a multisite longitudinal study in North America. face-to-face psychotherapy for depression and anxiety disorders? A sys-
Arch Gen Psychiatry 2008;65:28-37. tematic review and meta-analysis of comparative outcome studies. Psy-
31. Velthorst E, Nieman DH, Becker HE et al. Baseline differences in clinical
chol Med 2010;40:1943-57.
symptomatology between ultra high risk subjects with and without a tran-
56. Stain HJ, Bucci S, Baker AL et al. A randomised controlled trial of cognitive
sition to psychosis. Schizophr Res 2009;109:60-5.
behaviour therapy versus non-directive reflective listening for young peo-
32. Carrion RE, McLaughlin D, Goldberg TE et al. Prediction of functional out-
ple at ultra high risk of developing psychosis: the detection and evaluation
come in individuals at clinical high risk for psychosis. JAMA Psychiatry
of psychological therapy (DEPTh) trial. Schizophr Res 2016;176:212-9.
2013;70:1133-42.
57. Smith R, Feachem R, Feachem NS et al. The fallacy of impartiality:
33. Nelson B, Yuen HP, Wood SJ et al. Long-term follow-up of a group at ultra
competing interest bias in academic publications. J R Soc Med 2009;
high risk (prodromal) for psychosis: the PACE 400 study. JAMA Psychia-
try 2013;70:793-802. 102:44-5.
34. Falkenberg I, Valmaggia L, Byrnes M et al. Why are help-seeking subjects 58. Open Science Collaboration. An open, large-scale, collaborative effort to
at ultra-high risk for psychosis help-seeking? Psychiatry Res 2015;228:808- estimate the reproducibility of psychological science. Perspect Psychol Sci
15. 2012;7:657-60.

59. van Assen MA, van Aert RC, Nuijten MB et al. Why publishing everything is 61. Fusar-Poli P, Yung AR, McGorry P et al. Lessons learned from the psychosis
more effective than selective publishing of statistically significant results. high-risk state: towards a general staging model of prodromal interven-
PLoS One 2014;9:e84896. tion. Psychol Med 2014;44:17-24.
60. McGorry PD, Tanti C, Stokes R et al. Headspace: Australias National Youth
Mental Health Foundation where young minds come first. Med J Aust
2007;187:S68-70. DOI:10.1002/wps.20423

INSIGHTS
Treatment of people at ultra-high risk for psychosis

The ultra-high risk (UHR) criteria were defined to identify measured caregiver warmth, family communication and so-
young people at high and imminent risk of developing a first cial functioning as outcomes9. All studies showed feasibility
episode of psychosis1. The criteria have now been in use and either significant results or trends to significance, indi-
worldwide for over 20 years and have shown predictive validity cating future avenues of research.
for psychotic disorders across different countries and service The above approaches are moving towards developing in-
settings. UHR individuals have a risk of developing a full psy- terventions that are more tailored to underlying pathophysiol-
chotic disorder of 15-30% within 12 months, and over 36% ogy. Given the heterogeneity of the UHR group and our
after 3 years2. These transition rates are several hundred- knowledge that poor outcomes include development or persist-
fold above that of the general population. Most individuals ence of non-psychotic disorders and chronic social disability,
who develop a psychotic disorder have a diagnosis of schizo- this is a movement that should be welcomed. One problem is
phrenia or a schizophrenia-spectrum disorder. Identification that we lack understanding of the factors that predict these
of UHR individuals, therefore, presents the opportunity for different outcomes, including underlying biological mecha-
prevention of onset of full psychotic disorder, or at least nisms. This means that we are unable to individualize treat-
reduction in disability and delay of onset of first-episode ments. Thus, some UHR individuals are having unnecessary
psychosis. treatment, and others are having ineffective interventions that
Treatment of UHR individuals has two aims: to manage cur- potentially delay initiation of effective treatment. There is,
rent symptoms and problems, and to reduce the risk of devel- therefore, a need for investigation into factors that predict
oping a psychotic disorder1. Intervention trials tend to have different trajectories and outcomes. The aim is to stratify the
transition to psychosis as the primary outcome, with symp- UHR group according to their underlying pathological processes
toms, level of functioning and distress sometimes included as and target treatment accordingly.
secondary outcome measures. A recent meta-analysis studied Clearly, we will also need to better understand the mecha-
10 randomized trials that reported effects on transition rates nisms of action of the interventions. Examples include deter-
of low-dose antipsychotic medication, cognitive behavioural mining if a subtype of the UHR group has high levels of
therapy (CBT), omega 3 fatty acid and integrated treatment oxidative stress and using an agent that has reduction in oxida-
including family therapy, cognitive remediation, social skills tive stress as its mechanism of action. For example, N-acetyl
training and CBT3. This study found that receipt of any spe- cysteine (NAC) is an antioxidant and may be indicated in such
cific intervention significantly reduced the risk of developing a individuals. Studies will need to measure both reduction in oxi-
first episode of psychosis both at 12 months and over the dative stress and its correlation with improvement in symptoms
longer term (2-4 years), albeit with diminished effects over and functioning as outcomes. We will need to investigate if the
time. The reduced effect at long-term follow-up suggests that mechanism of action of NAC in the UHR group is through
at least some UHR individuals remain at risk, and that inter- reduction in oxidative stress or through some other process
ventions might delay, rather than prevent, onset of psychosis. (such as reduction in inflammation or an effect on neurotrans-
Even so, such a delay could be of benefit, enabling people to, mitters). Similarly, some UHR individuals may have high levels
for example, finish education and develop supportive net- of dysfunctional metacognitive beliefs that lead to misinterpret-
works outside the family of origin. Additionally, individuals ing events and difficulty in dealing with stressful situations.
who develop a first episode of psychosis after having been These individuals could benefit from metacognitive therapy.
treated in the prodromal phase have improved outcomes Reduction in dysfunctional metacognitive beliefs should be
compared to their counterparts who did not receive such very measured as an outcome as well as symptoms and distress10.
early intervention4. Transition to psychosis will also still be a relevant outcome in
Recently some novel treatments have also been piloted in both scenarios.
the UHR group. These have had more targeted outcomes, Another issue in treatment of UHR individuals is whether
based on hypothesized mechanisms of action of the interven- specialized services are indicated and if so, where they should
tion rather than global aims of reducing transition risk. For be located. A major reform of early intervention in psychosis
example, a small study of lithium postulated that it may have services has recently been implemented in England. All these
a neuroprotective effect and examined hippocampal T2 relax- services are now required to assess for presence of the UHR
ation time and proton magnetic resonance spectroscopy as state (there called the at risk mental state) and provide man-
outcomes testing this hypothesis5. Glycine has been tested in agement of UHR individuals. Patients detected through this
two small pilot trials with outcomes of symptoms and neuro- pathway are likely to have high levels of symptoms as they will
cognitive functioning6. A study of biofeedback measured anxi- have originally been referred as possible first-episode psy-
ety and distress as outcomes7, and a trial of processing speed chosis. They will likely resemble the original cohort of UHR
training examined improvement in processing speed and its cor- patients identified mainly through this route over two decades
relation with social functioning8. A trial of a family intervention ago1. It may be therefore that the transition rate in this group

is also higher than cohorts detected through more generalist Alison R. Yung
pathways such as adolescent health services. Thus, it might Division of Psychology and Mental Health, Faculty of Biology, Medicine and Health,
University of Manchester and Greater Manchester Mental Health NHS Foundation
be that the integration of UHR and early intervention in psy- Trust, Manchester Academic Health Science Centre, Manchester, UK
chosis services is indicated, facilitating timely treatment of
psychosis should that occur. 1. Yung A, McGorry PD, McFarlane CA et al. Schizophr Bull 1996;22:283-303.
2. Fusar-Poli P, Bonoldi I, Yung AR et al. Arch Gen Psychiatry 2012;69:220-9.
On the other hand, we now know that young people with 3. van der Gaag M, Smit F, Bechdolf A et al. Schizophr Res 2013;149:56-62.
depressive and anxiety disorders frequently experience psychotic- 4. Valmaggia L, Byrne M, Day F et al. Br J Psychiatry 2015;207:130-4.
like symptoms and may meet criteria for the UHR state. For these 5. Berger GE, Wood SJ, Ross M et al. Curr Pharm Des 2012;18:570-5.
6. Woods SW, Walsh B, Hawkins K et al. Eur Neuropsychopharmacol 2013;23:
individuals, who will most likely present to primary care or
931-40.
adolescent services, it may be that management is optimal in an 7. McAusland L, Addington J. Early Interv Psychiatry (in press).
enhanced primary care youth service, such as Headspace in Aus- 8. Choi J, Corcoran CM, Fiszdon JM et al. Psychiatr Rehabil J 2017;40:33-42.
tralia. Ideally we need to know more about the different subtypes 9. OBrien MP, Gordon JL, Bearden CE et al. Schizophr Res 2006;81:269-75.
10. Cotter J, Yung AR, Carney R et al. Behav Res Ther 2017;90:25-31.
of UHR individuals and move towards stratified pathways of
care depending on need, risk profile and likely underlying DOI:10.1002/wps.20424
pathophysiology.
Persistent persecutory delusions: the spirit, style and content

of targeted treatment
We believe that treatments for persecutory delusions can be that they are now safe. Learning of safety counteracts the para-
substantially better. Current standard psychological and phar- noia. The fundamental learning is that the difficulty is one of tol-
macological treatments have small to moderate effects1,2. The erating high anxiety, rather than that there is an external threat.
severity of the problems associated with paranoia is typically The spirit, style and content of the 20-session Feeling Safe
considerable, but the treatments are less effective than those Programme has emerged from theoretical understanding, pa-
for problems such as anxiety disorders. The isolation, feelings of tient feedback, and our own clinical experience7. To start, the
hopelessness, and missed opportunities for patients with perse- three overarching goals of treatment, shared with patients, are
cutory delusions demand a step change in treatment outcomes. simple: to feel safer, happier, and to get people back doing
This is a clinical area that is beginning to receive a degree of more of what they want to be doing. These positively framed
attention. There are innovations in understanding and treat- goals are popular with patients, enhance engagement, and
ment emerging3-5. Central to our own strategy for improving embed the mechanism of change developing feelings of
treatment have been three inter-connected elements: a sus- safety from the outset. The goals also orient the intervention
tained, specific focus upon persecutory delusions; the devel- to the future. We are explicit that no significant time is spent
opment of a precise theoretical model with causal elements going over the past, unless that is requested by a patient.
amenable to intervention; and a style and content of interven- Secondly, our perspective that there are multiple causal fac-
tors, and the consequent development of multiple treatment
tion that follows from our understanding of delusions. Our
modules, allows both individual tailoring of the intervention
objective has been to achieve a much higher recovery rate for
and patient preference. A brief assessment, combining clinical
persecutory delusions.
interview and questionnaires, identifies with patients the factors
The strategy behind building a new treatment has been to
target in separate interventions each key causal factor identified contributing to their difficulties, and leads to the presentation
from our theoretical model, demonstrate that each reduces the of a treatment menu. Patients choose which interventions they
delusion, and then bring the evaluated individual components would like and in which order. This gives patients real control
together into one coherent framework called the Feeling Safe from the outset.
Programme that can be personalized for patients. Thirdly, targeting each maintenance factor, focusing on one
Persecutory delusions are conceptualized as threat beliefs, at a time, provides a method to address the undoubted complex-
developed in the context of genetic and environmental risk, ity (and often associated feelings of hopelessness) of presenting
that are maintained by several psychological processes, includ- problems. We acknowledge the complexity with patients, but
ing excessive worry, low self-confidence, intolerance of anxious explain that a way to deal with it is to tackle one problem, then
affect (and other internal anomalous experiences), reasoning move on to the next, starting with the most manageable. This
biases, and the use of defence strategies6. Therefore, the clinical reduces the influence of maintenance factors but also raises
strategy is first to limit the maintenance factors one by one, then patients capacity and confidence to face the demands of directly
enable patients to enter their feared situations in order to learn learning safety in vivo.

Fourthly, throughout the programme, we monitor the caus- recovery in half of patients with persistent delusions. If this is
al mechanism targeted in a module, as well as the three over- achieved, there will then be a problem of accessibility. We have
arching goals of the intervention. This enables us to track and developed the programme in a highly manualized form to aid
demonstrate change with patients. Scores are also used in the later dissemination, but technological solutions may also prove
regular, frequent supervision, particularly to rapidly identify important. For example, we have found that immersive virtual
cases requiring greater discussion. reality can help patients learn safety9. Mobile apps and web-
Fifthly, the style that has evolved from this systematic step- based programs also offer alternative delivery methods10.
by-step approach is akin to interval training: bursts of activity New treatments for persecutory delusions obviously require
and intensity followed by periods of reflection and integration. empirical testing in rigorous trials. Different forms of treatment
Of course, within this approach, the absolute pace of the inter- should not be regarded as a single class, given the varied mecha-
vention remains tailored to the individuals needs and prefer- nistic targets, delivery methods, and outcomes pursued. We
ence. Time is predominately dedicated to the implementation believe that the concept of specificity, inherent in our approach,
of strategies in day-to-day life. Substantial additional contact should be retained when evaluating treatment developments. In
(e.g., telephone calls) between weekly sessions is expected. this way, promising routes to improved outcomes for patients
This is not low intensity working. with persistent delusions will not be obscured.
Finally, the clarity of the model, and strong evidence-base
for each element, enables the therapeutic style to be encourag- Daniel Freeman, Felicity Waite
Department of Psychiatry, University of Oxford, Warneford Hospital, Oxford, UK
ing and optimistic, often holding hope when the patient strug-
gles (e.g., many patients with persistent delusions, right at the D. Freeman is supported by a National Institute for Health Research Professor-
start, are not expecting improvement). Transparency, offering ship.
direct answers to questions, and providing expert opinion (that

is accurate), in tandem with the monitoring of progress and col- 1. Leucht S, Cipriani A, Spineli L et al. Lancet 2013;382:951-62.
2. Van der Gaag M, Valmaggia LR, Smit F. Schizophr Res 2014;156:30-7.
laborative style, helps substantiate that optimism for patients. 3. Moritz S, Pfuhl G, L
udtke T et al. J Behav Ther Exp Psychiatry (in press).
All written materials are shared between therapist and patient. 4. Lincoln TM, Hartmann M, K other U et al. Psychiatry Res 2015;228:216-22.
5. Wickham S, Taylor P, Shevlin M et al. PLoS One 2014;9:e105140.
There is no separate therapist manual. The therapy booklets 6. Freeman D. Lancet Psychiatry 2016;3:685-92.
provide the framework and key messages of the intervention, 7. Freeman D, Bradley J, Waite F et al. Behav Cogn Psychother 2016;44:539-
but are not prescriptive. Creativity by both the therapist and 52.
8. Freeman D, Waite F, Emsley R et al. Trials 2016;17:134.
patient is often fostered, ensuring personal meaning and suc- 9. Freeman D, Bradley J, Antley A et al. Br J Psychiatry 2016;209:62-7.
cessful embedding of strategies for change. 10. Hardy A, Garety P, Freeman D et al. Front Public Health 2016;4.
We are currently testing the full Feeling Safe Programme in a
randomized controlled trial8. There are, of course, caveats. The
approach does not benefit all patients: our target at this stage is DOI:10.1002/wps.20425
Does neuroimaging have a role in predicting outcomes in psychosis?

A key difficulty in the management of psychotic disorders is outcomes2. This inconsistency may reflect the use of patient
that clinical outcomes are difficult to predict on the basis of the samples that were small, and heterogeneous for age, stage of ill-
patients clinical features. As a result, patients with psychosis are ness, and pharmacological treatment, all of which can affect neu-
generally treated in a similar way, even though there may be roimaging findings. Moreover, clinical outcomes have often been
marked differences in their course of illness or response to medi- determined retrospectively, on the basis of clinical records.
cation. However, recent research using neuroimaging suggests Recent neurochemical imaging studies have suggested that
that, within a sample of patients with psychosis, the pattern of the response to antipsychotic medication in patients with psy-
abnormalities may vary in relation to different clinical outcomes. chosis is related to both subcortical dopamine function, as meas-
This raises the possibility that neuroimaging could be used to ured using positron emission tomography, and regional brain
stratify patients according to clinical outcome; subgroups of glutamate levels, as assessed using magnetic resonance spec-
patients could then be offered different forms of treatment. troscopy. A good therapeutic response has been associated with
Data from a number of structural magnetic resonance imaging elevated dopamine function and relatively normal glutamate
(MRI) studies suggest that patients with relatively poor outcomes levels, whereas a poor response has been linked to normal
have, compared to those with good outcomes, more marked re- dopamine function and elevated glutamate levels3. Indepen-
ductions in total and regional grey matter volume, and greater dent work has also linked the response to antipsychotic medi-
ventricular enlargement1. However, other studies have not found cation to differences in cortical gyrification4, and to diffusion
a relationship between alterations in brain structure and clinical tensor imaging measures of white matter integrity5. However,

again, these studies involved relatively small samples, and the Future studies may also benefit from using more than one
patients were scanned after they had been treated with antipsy- modality of neuroimaging; there is some evidence that this
chotic medication: it is thus unclear whether the neuroimaging may improve prediction of outcomes9, although other data do
findings predated treatment or were secondary to it. not support this10. Similarly, integrating neuroimaging data
Most studies to date have related clinical outcomes to a sin- with non-imaging measures that have independently been
gle cross-sectional neuroimaging measure. Serial neuroimag- linked with altered outcomes in psychosis, such as polygenic
ing measurements provide data on how the brain changes over risk score, substance use, inflammatory markers and central
time within the same patient, and recent studies involving lon- nervous system autoantibodies, may enhance predictive pow-
gitudinal scanning of patients suggest that measuring the pro- er. However, although this may be a reasonable expectation, it
gression of findings facilitates the prediction of outcome6. For has yet to be tested.
example, longitudinal data from patients with first episode Even if a neuroimaging measure is established as a robust sta-
psychosis and from those with childhood-onset schizophrenia tistical predictor of clinical outcomes, this does not necessarily
suggest that reductions in hippocampal volume over the first mean that it can be translated into mainstream clinical practice.
few years of illness are associated with poorer functioning at Financial and practical considerations will apply, such as the
follow-up7. cost of scanning and the availability of the scanner. The develop-
All of the studies mentioned above reported differences ment of tools that can be used in a clinical setting is likely to
between groups of patients. However, in order for neuroimag- require neuroimaging measures that can be acquired without
ing to be useful in a clinical setting, it must be able to facilitate the need for highly specialized training or equipment. Some
outcome prediction using data from an individual patient. ongoing studies are explicitly focused on the development of
Multivariate statistical approaches such as machine learning such tools for psychosis (see, for instance, www.psyscan.eu).
provide a means of addressing this issue. For example, appli- Given that psychotic disorders are pathophysiologically het-
cation of machine learning analyses to MRI data from patients erogeneous, it is reasonable to expect that neuroimaging tech-
with first episode psychosis showed that baseline neuroimag- niques which can identify pathophysiological differences within
ing data could predict a non-remitting course of illness over patient samples may be useful in predicting clinical outcomes.
the subsequent six years with an accuracy of 72%8. However, at present, it is unclear which particular neuroimag-
Ongoing studies in this field are seeking to address the meth- ing measures will be the most useful, and whether combining
odological issues that may have limited earlier work. Sample sizes these with non-imaging biomarkers will enhance their ability
can be increased through the involvement of multiple research to facilitate prediction of outcomes in psychosis.
sites. Although multi-centre studies are logistically challenging,
and there are significant confounding factors associated with Philip McGuire, Paola Dazzan
Department of Psychosis Studies, Institute of Psychiatry, Psychology and Neurosci-
acquiring data on a variety of different scanners, these disad- ence, Kings College London; National Institute for Health Research (NIHR) Mental
vantages are probably outweighed by the increased statistical Health Biomedical Research Centre at South London and Maudsley NHS Foundation
power that results from having much larger samples. Similarly, Trust and Kings College London, London, UK
serial neuroimaging studies are more difficult to carry out than

1. Dazzan P, Arango C, Fleischhacker W et al. Schizophr Bull 2015;41:574-83.
those involving a single scan, but may provide more predictive 2. Sharma T, Kerwin R. Br J Psychiatry 1996;169(Suppl. 31):5-9.
power. Ongoing studies have also sought to enroll samples that 3. Demjaha A, Egerton A, Murray RM et al. Biol Psychiatry 2014;75:e11-3.
4. Palaniyappan L, Marques TR, Taylor H et al. JAMA Psychiatry 2013;23:
are homogeneous with respect to stage of illness and previous
1031-40.
treatment, and that are treated in a standardized way subse- 5. Reis Marques T, Taylor H, Chaddock C et al. Brain 2014;137:172-82.
quent to scanning. A good example of this is OPTiMiSE (Opti- 6. Cahn W, van Haren NE, Hulshoff Pol HE et al. Br J Psychiatry 2006;189:
381-2.
mization of Treatment and Management of Schizophrenia in
7. Lappin JM, Morgan C, Chalavi S et al. Psychol Med 2014;44:1279-91.
Europe), a large multicenter study funded by the European 8. Mourao-Miranda J, Reinders AA, Rocha-Rego V et al. Psychol Med 2012;42:
Commission1. This involves a neuroimaging assessment of a large 1037-47.
multi-centre sample of medication-nave or minimally treated 9. Suckling J, Barnes A, Job D et al. Hum Brain Mapp 2010;31:1183-95.
10. Reig S, Sanchez-Gonzalez J, Arango C et al. Hum Brain Mapp 2009;30:
first episode patients, all of whom are then treated with ami- 355-68.
sulpride following a standardized protocol. Their clinical out-
DOI:10.1002/wps.20426
comes are evaluated prospectively.
The role of expectations in mental disorders and their treatment

Expectations are defined as cognitions which are future- discussed as a central mechanism of change2,3. This focus on
directed and focused on the incidence or non-incidence of a expectations does not disregard any past experiences, but con-
specific event or experience1. In the treatment of mental dis- siders them only of relevance if they determine predictions
orders, examining and modifying patients expectations is about future events.

The relevance of expectations for clinical conditions and viduals suffering from depression show more depression-
their treatment can be illustrated by the following example: specific negative expectations7. Even in general medical condi-
temporary ear noises are no problem for most people, as long tions, expectations and expectation-associated concepts (e.g.,
as the affected persons expect them to vanish promptly. How- fear avoidance in chronic pain) have been shown to predict per-
ever, the same experience is difficult to bear if affected people sistence and survival8.
expect them to last forever. Analogously, it may be not the neg- Expectations about treatment success are the most prominent
ative mood, the unpleasant stimulation, the adverse life event predictor of outcome, both in psychopharmacological and psy-
per se that determines whether exposed people develop a mental chological interventions, and they are considered to be a major
or psychosomatic disorder, but the expectation about the time- determinant of placebo effects9. In most psychopharmacological
line of the aversive condition, expected future threats, expected trials, placebo responses represent a substantial proportion of
curability, and expected competence to cope with the unpleas- the overall treatment effect. Optimizing treatment expectations
ant experiences. can result in improved outcome and prevention of treatment
Neurobiology and psychological sub-disciplines such as de- side effects, while the induction of negative expectations can
velopmental psychology and social psychology have focused abolish the effects of highly effective medications10.
on expectations for decades. They provide us with detailed If expectations are one of the most powerful predictors of
knowledge on how expectations are formed, under what cir- outcome, interventions must maximally modify illness-specific
cumstances they are modified, or when they persist despite expectations, and positive outcome expectations should be suffi-
contradictory experiences. ciently established before treatment starts. One of the traditional
Expectations lead to brain activities that sensitize for the psychological interventions that may be considered a powerful
expected experience4, and they are closely linked to affective tool to change expectations is exposure therapy. However, tradi-
reactions5. Prediction error paradigms and their association tional exposure therapy needs to be reformulated to better focus
with dopaminergic activation, amygdala activation during aver- on the change of expectations (e.g., explicit comparison between
sive coding, and the role of contextual information in the gener- pre-exposure expectations and post-exposure experiences)3.
ation of expectancies are just a few neurophysiological examples Expectation-focused psychological interventions (EFPI)7 place
of how the topic has been investigated. a strong focus on analyzing and summarizing disorder-specific
Associative learning, influences via group norms and media, expectations of the patient, developing situational tests to check
and the phenomenon of sticking to expectations despite expec- the credibility of these expectations, and re-evaluating expecta-
tation violations (cognitive immunization) are psychologically tions by comparing pre-existing expectations with the experience
relevant concepts to better understand why specific expecta- during exposure.
tions are present. In addition to disorder-specific expectations, the baseline
We currently face the challenge of investigating the role of expectations about positive and negative effects of interventions
expectations from a clinical perspective and transferring this should play an important role in treatment planning. If patients
knowledge into psychotherapeutic and psychopharmacological have negative attitudes about drug therapy, these attitudes should
practice. This approach may allow for a better understanding of be addressed before starting medication. In psychological thera-
the dynamics of mental and psychosomatic disorders, guiding pies, positive outcome expectations should be established before
more challenging interventions are suggested.
the development of tailored interventions based on highly effec-
Considering the large effects that must be attributed to placebo
tive mechanisms. Moreover, focusing on expectations and their
mechanisms in psychiatry, expectations and their modification
persistence helps to explain why some treatments fail.
can be considered the most powerful mechanism for successful
Some mental disorders are expectation disorders by defini-
treatment. Therefore, there is an urgent need to utilize knowledge
tion. This is particularly so in the case of anxiety disorders, such
about expectations to improve treatment outcomes.
as phobias, panic disorder and generalized anxiety disorder. In
these cases, patients expect adverse consequences when being Winfried Rief, Julia Anna Glombiewski
exposed to specific stimuli, situations, or experiences (e.g., the Department of Clinical Psychology and Psychotherapy, Philipps-University of Marburg,
Marburg, Germany
phobic stimulus, the experience of palpitations). In obsessive-
compulsive disorder, the patient expects dreadful consequences 1. Rief W, Glombiewski JA, Gollwitzer M et al. Curr Opin Psychiatry 2015;28:
if compulsive behaviors are prohibited. 378-85.
The role of expectations in post-traumatic stress disorder 2. Craske MG, Treanor M, Conway CC et al. Behav Res Ther 2014;58:10-23.
3. Craske MG. Verhaltenstherapie 2015;25:134-44.
(PTSD) seems to be more complex. While most people feel 4. Koyama T, McHaffie JG, Laurienti PJ et al. PNAS 2005;102:12950-5.
secure and do not expect horrible events, this basic confidence 5. Schwarz KA, Pfister R, Buchel C. Trends Cogn Sci 2016;20:469-80.
in everyday life situations is violated if people suffer from 6. Janoff-Bulman R. Soc Cogn 1989;7:113-36.
7. Rief W, Glombiewski JA. Verhaltenstherapie 2016;26:47-54.
trauma6. Some patients with PTSD do not want to talk about
8. Barefoot JC, Brummett BH, Williams RB et al. Arch Intern Med 2011;171:
the trauma because they do not expect to be able to bear the 929-35.
emotions that will arise. 9. Schedlowski M, Enck P, Rief W et al. Pharmacol Rev 2015;67:697-730.
10. Enck P, Bingel U, Schedlowski M et al. Nat Rev Drug Discov 2013;12:191-204.
In other mental disorders, expectations are not part of the
diagnostic criteria, but are also of relevance. For example, indi- DOI:10.1002/wps.20427

LETTERS TO THE EDITOR
Why ultra high risk criteria for psychosis prediction do not work well
outside clinical samples and what to do about it
The use of ultra high risk (UHR) criteria in selected help- offered to selected samples of subjects already distressed by
seeking samples is the only clinical possibility to alter the mental problems and seeking help for them.
course of psychosis by preventing its onset. The UHR paradigm At the same time, because of the potential benefits yielded by
can additionally reduce the duration of untreated psychosis1 the UHR paradigm, it seems important to continue exploring the
and provide extended benefits to patients who are experiencing usefulness of an extended application of UHR assessment in sev-
a first episode of psychosis2. eral different samples. A first pragmatic approach to estimating
Because of these potentials, there is a great interest in the the prognostic accuracy of the UHR assessment in several scenar-
use of UHR outside clinical samples, such as in the general ios would be to use the meta-analytical Fagans nomogram that
population. The first epidemiological study investigating the we presented in a previous paper in this journal9. This nomogram
significance of UHR criteria in the non-help-seeking general is based on the intrinsic properties of the UHR assessment (such
population aged 8-40 was published in this journal3. It indi- as the positive and negative likelihood ratios7) and can be applied
cated that only 1.3% of the general population met the UHR to different populations with a given pre-test risk of psychosis
criteria of the Structured Interview for Psychosis-Risk Syn- onset to estimate their post-test risk of psychosis at 38 months.
dromes (SIPS)3. The longitudinal fate of these individuals has Importantly, our nomogram has now been externally vali-
just been released4: 143 UHR and 131 controls were followed dated. In fact, with that nomogram, we had estimated a small
up for an average of 2.5 years, with three transitions to psycho- post-test psychosis risk (less than 5% at 38 months) in the gen-
sis in the UHR group (psychosis risk 5 2.09%) and no transi- eral population, a value that is similar to the real value observed
tion in the control group. in the epidemiological study discussed above4. Similarly, with
These results are of great interest, as they may support the our nomogram, we had estimated a post-test psychosis risk of
epidemiological validity of the UHR paradigm, although they 26% for patients affected with the 22q11.2 deletion syndrome9,
are likely to be underpowered (assuming a 0.001% risk in the which exactly matches to the real value recently reported in this
control group as continuity correction and an alpha 5 0.05, the journal10.
resulting power would be of 38% only). Beyond these limita- The use of our nomogram can thus provide reliable estimates
tions, the key finding of 2.09% psychosis risk (at 2.5 years) in (along with 95% CIs) for post-test risk of psychosis in individuals
people meeting UHR from the general population is of crucial meeting UHR criteria, given a determined pre-test risk. Using the
clinical relevance. It is strikingly lower than the annualized 2- nomogram, researchers can simulate the expected prognostic
year 20% (95% CI: 17%-25%)5 transition risk in help-seeking accuracy, and estimate the required sample size needed to test
UHR samples, that are characterized by frequent comorbid their hypotheses.
affective disorders and functional impairments6. Since the use of the UHR assessment outside clinical samples
These findings clearly confirm that the prognostic accuracy is likely to be associated with low predictive power, it is funda-
of the UHR criteria strictly depends on the sample to which mental to perform accurate power calculations. In this scenario
they are being applied. Indeed, clinical help-seeking samples of and considering the probability of infrequent events, a second
individuals undergoing UHR assessment are characterized by a approach could involve using sequential testing methods11, for
substantial pre-test risk enrichment (pre-test risk for psycho- example by using the SIPS in samples already enriched for psy-
sis)7 of up to 15% at 38 months8. As demonstrated in a previous chosis risk, as shown in this journal12. Sequential testing is tradi-
paper in this journal9, the use of UHR assessment is associated tionally adopted in medicine to enrich the risk of samples that
with a small positive likelihood ratio of 1.82 at 38 months and a are selected to undergo different diagnostic or prognostic tests.
modest ability to rule in psychosis9. Therefore, to reach some A third practical approach could be to better investigate the
prognostic accuracy of clinical utility in individuals meeting factors that modulate pre-test risk enrichment in samples under-
UHR criteria, it is necessary to apply them to samples that are going a UHR assessment. We have recently shown that it may be
already enriched in psychosis risk, i.e., with a significant pre- possible to stratify help-seeking individuals undergoing UHR
test risk. For example, a recent study published in this journal10 assessment through the use of simple socio-demographic and
has shown that meeting the UHR criteria given an underlying clinical variables13. The predictive model has been externally
22q11.2 deletion syndrome, a condition that is characterized by validated and can be used to inform future research in the field,
a substantial pre-test risk for psychosis, is associated with a with the scope to improve prognostic accuracy of psychosis pre-
27.3% risk of psychosis at 32 months. diction.
These considerations clearly limit the practical utility of the
UHR outside of clinical samples, as recently recognized by the Paolo Fusar-Poli
Kings College London, Institute of Psychiatry, Psychology and Neuroscience; OASIS
recommendation no. 4 of the European Psychiatric Association, Service, South London and Maudsley NHS Foundation Trust, London, UK
which suggests that the UHR assessment should be primarily

1. Valmaggia LR, Byrne M, Day F et al. Br J Psychiatry 2015;207:130-4. 9. Fusar-Poli P, Cappucciati M, Rutigliano G et al. World Psychiatry 2015;14:
2. Fusar-Poli P, Diaz-Caneja CM, Patel R et al. Acta Psychiatr Scand 2016;133: 322-32.
76-85. 10. Schneider M, Armando M, Pontillo M et al. World Psychiatry 2016;15:
3. Schimmelmann BG, Michel C, Martz-Irngartinger A et al. World Psychiatry 259-65.
2015;14:189-97. 11. Schmidt A, Cappucciati M, Radua J et al. Schizophr Bull (in press).
4. Michel C, Schimmelmann BG, Schultze-Lutter F. Early Interv Psychiatry 12. Calkins ME, Moore TM, Satterthwaite TD et al. World Psychiatry 2017;16:
2016;10(Suppl. 1):129. 62-76.
5. Fusar-Poli P, Cappucciati M, Borgwardt S et al. JAMA Psychiatry 2016;73: 13. Fusar-Poli P, Rutigliano G, Stahl D et al. JAMA Psychiatry 2016;73:1260-7.
113-20.
6. Fusar-Poli P, Rocchetti M, Sardella A et al. Br J Psychiatry 2015;207:198-206. DOI:10.1002/wps.20405
7. Fusar-Poli P, Schultze-Lutter F. Evid Based Ment Health 2016;19:10-5.
8. Fusar-Poli P, Schultze-Lutter F, Cappucciati M et al. Schizophr Bull 2016;
42:732-43.
Drug use disorders: impact of a public health rather than a criminal

justice approach
The Outcome Document of the 2016 United Nations Gen- Eliminate stigma and discrimination toward individuals with
eral Assembly Special Session on drugs (UNGASS 2016), unan- substance use disorders. Increasing public awareness of addic-
imously approved by the 193 Member States, has recognized tion/dependence as a chronic but treatable disorder is need-
drug addiction as a complex multifactorial health disorder ed to overcome stigma and promote a shift from exclusion
characterized by chronic and relapsing nature that is pre- and blame toward support and compassion. This should
ventable and treatable and not the result of moral failure or a include national policies that address substance use disor-
criminal behavior. Historically, most nations strategies for ders as neurobiological disorders having complex social and
addressing substance use disorders have centered on punish- developmental underpinnings.
ment, and thus recognition of the need to shift from a criminal Address substance use disorders as public health problems
instead of criminal justice issues. A comprehensive public
justice to a public health approach represents a major shift in
health approach should offer accessible evidence-based pre-
mentality by United Nations Member States.
vention, treatment, and recovery options to drug users, and
This achievement was the result of a continuous dialogue
engage those who commit criminal offences in evidence-
between policy makers and the scientific community during
based treatment during and following, or in lieu of, incarcer-
recent sessions of the United Nations Commission on Narcotic
ation, to prevent relapse and recidivism. It also includes nal-
Drugs. In 2015, the United Nations Office of Drugs and Crime
oxone distribution for overdose prevention3, and integration
and the World Health Organization created an Informal Inter-
of treatment of substance use disorders with prevention and
national Scientific Network, consisting of experts in addiction
treatment of infectious diseases (HIV and hepatitis C)4 and
sciences, to advise the Commission. Network members were
of co-occurring psychiatric conditions5.
appointed by Member States and represented widely diverse
Implement evidence-based prevention programs. Substance
geographical regions, political systems, and cultures.
use disorders are fully preventable. The use of evidence-
The Networks input for the Commissions preparation of
based prevention programs, both universal and targeted to
UNGASS 2016 provided the scientific support for the concept
high-risk individuals, has shown positive outcomes in reduc-
that substance use disorders are brain disorders1; that they
ing drug initiation and escalation of use. Since prevention
can be treated; that people with even the most severe forms
programs address risk and protective factors that are com-
can recover with access to evidence-based treatment and so-
mon to a range of behavioral problems, they produce positive
cial supports2; and that criminal sanctions are ineffective at
outcomes not just in drug taking but also in reducing aggres-
preventing or addressing these disorders. It also highlighted
sion, early pregnancies, and drugged driving, and improve
evidence-based approaches to drug policy based on public mental health and educational outcomes. Highest priority
health principles, emphasizing social protection and health should be given to interventions targeting children and youth,
care instead of conviction and punishment. since the earlier the use of drugs the greater the risk for sub-
The Network issued eight recommendations, which were stance use disorders and the higher their severity6.
adopted unanimously by all the United Nations Member Implement evidence-based treatments for substance use dis-
States at UNGASS 2016 and summarized in the Outcome Doc- orders. Abundant research shows that these disorders are
ument of that meeting. These recommendations are a testa- treatable and that people do recover when given evidence-
ment to a momentous shift in mentality, to which science and based care, including behavioral therapies for all these dis-
the Network have contributed. orders and medication-assisted treatments for alcohol and
The recommendations are as follows: opioid use disorders and for smoking cessation7,8. However,

because changes in the brain function in these disorders ing controlled drugs for legitimate medical needs, such as
can be long-lasting, an individual may be at increased risk analgesic drugs in the more than 150 countries where pain is
for relapse even after years of abstinence. Effective treat- undertreated9.
ment thus requires a chronic care model as used for other
chronic conditions such as cardiovascular disease or diabe- The public health goal of reducing the worlds drug prob-
tes, which along with routine screening should be integrat- lems cannot be achieved without addressing substance use
ed into the general health care system and be affordable disorders with the same scientific rigor, compassion, and com-
and accessible. mitment that other physical and mental health problems are
Collect and utilize scientific data and engage scientific experts addressed. Substance use disorders are common psychiatric
in policy making. Reliable epidemiological data on the eco- disorders, and access to affordable, quality health care for
nomic and social factors that contribute to drug use and sub- such disorders has been declared an inherent right for all
stance use disorders should be gathered and analyzed to United Nations Member State citizens.
drive planning and evaluation of drug policy interventions The strong consensus reached by the Network scientists
and decision making. The scientific community should pro- representing very different countries that have widely varying
vide knowledge of effective prevention and treatment inter- policies, political views, and stages of development is an
ventions as well as training in their implementation and unprecedented and positive step toward a world where science
ongoing evaluation. Member States should establish national guides nations approach to drug misuse and its associated
early warning systems to monitor changing drug trends and health and safety consequences. Adopting these recommenda-
identify emerging public safety and health threats. tions will be crucial to fulfilling Member States joint commit-
Engage diverse stakeholders in coordinated policy making. Be- ment to effectively address and counter the world drug problem.
cause of the complexity of the health and safety issues related
Nora D. Volkow1, Vladimir Poznyak2, Shekhar Saxena2, Gilberto
to substance use disorders, policy makers should involve di- Gerra3, and the UNODC-WHO Informal International Scientific
verse stakeholders, including public health, education, law Network
1
enforcement, science, and health care systems, as well as solicit National Institute on Drug Abuse, National Institutes of Health, Bethesda, MD, USA;
2
Department of Mental Health and Substance Abuse, World Health Organization,
input from countries with different cultures, resources, and
Geneva, Switzerland; 3Drug Prevention and Health Branch, United Nations Office on
experiences. Stakeholders should cooperate in the planning, Drugs and Crime, Vienna, Austria
implementation, and evaluation of science-informed inter-
ventions and policies that address the demand as well as the The authors alone are responsible for the views expressed in this article and
they do not necessarily represent the views, decisions or policies of the institu-
supply of drugs. This would include diverting offenders into tions with which they are affiliated. The authors thank E.M. Wargo for his valu-
treatment, combating drug production and trafficking, cre- able editorial work. The members of the Network are C.F. Damin, G. Fischer, T.
Mota Ronzani, S.T. Zhenkova, M. Zhao, Z. Liu, O. Scoppetta Diaz Granados, M.
ating alternative opportunities for communities dependent Mahfouz, P. Arwidson, T. Pfeiffer-Gerschel, E. Adjei-Acquah, R. Lal, S. Ben Ezra,
on the drug trade, and ensuring the safety and protection of P. Rosca, C. Leonardi, I. Maremmani, M.T. Matar, J.A. Villatoro Velazquez, J.
Toufiq, I. Obot, J.G. Bramness, K.M. Ostaszewski, M.J. Rodrigues Dias, S.N.
the most vulnerable as it relates to drug taking but also engage- Al-Emadi, E.A. Bryun, G. Korchagina, E. Krupitskiy, O.A. Alibrahim, T. Hernandez
Fernandez, N. Stenstrom, M.P. Schaub, M. Boyle, I. Grant, S. Gust, A.T. McLellan,
ment in drug trading. N.D. Volkow, S. Weiss, G. Campello, E. Mattfeld, E. Saenz, A. Busse and N. Clark.
Support drug-related research. Ongoing research must ad-
dress the effects of drugs (especially emerging new syn-
1. Volkow ND, Koob GF, McLellan AT. N Engl J Med 2016;374:363-71.
thetic drugs) on the brain and behavior; the social and 2. Chandler RK, Fletcher BW, Volkow ND. JAMA 2009;301:183-90.
public health impact of different drug policies; the best 3. World Health Organization. Community management of opioid overdose.
ways to tailor prevention and treatment modalities to differ- Geneva: World Health Organization, 2014.
4. World Health Organization. WHO, UNODC, UNAIDS technical guide for
ent cultural contexts; and the therapeutic potential of con- countries to set targets for universal access to HIV prevention, treatment
trolled substances (e.g., cannabinoids). Regulatory impedi- and care for injecting drug users. Geneva: World Health Organization, 2012.
ments to conducting research on scheduled drugs should 5. World Health Organization. mhGAP intervention guide for mental, neuro-
logical and substance use disorders in non-specialized health settings.
be minimized and policies that facilitate research across Geneva: World Health Organization, 2010.
these areas implemented. 6. Robins LN, Przybeck TR. Age of onset of drug use as a factor in drug and
Ensure access to scheduled medications for therapeutic use. other disorders. Bethesda: National Institute on Drug Abuse, 1985.
7. National Institute on Drug Abuse. Principles of drug addiction treatment: a
Some controlled and dependence-producing psychoactive research-based guide (3rd ed). Bethesda: National Institute on Drug Abuse,
drugs are necessary medicines for treating serious health con- 2012.
ditions. The international drug conventions are designed to 8. World Health Organization. Guidelines for the psychosocially assisted
pharmacological treatment of opioid dependence. Geneva: World Health
ensure legitimate medical access to such medicines, under Organization, 2009.
appropriate supervision, through a distribution chain that 9. Berterame S, Erthal J, Thomas J et al. Lancet 2016;387:1644-56.
deters and combats illicit manufacture, sale, and diversion.
Necessary steps should be taken to remove barriers to access- DOI:10.1002/wps.20428

Prevention and early intervention for borderline personality
disorder: a novel public health priority
There is now a broad evidence-based consensus that border- in May 2014. The Alliance calls for action through a set of sci-
line personality disorder (BPD) is a reliable, valid, common and entifically based clinical, research and social policy strategies
treatable mental disorder1. The adverse personal, social and eco- and recommendations.
nomic consequences of BPD are severe. They include persistent Clinical priorities include: a) early intervention (i.e., diagnosis
functional disability2, high family and carer burden3, incomplete and treatment of BPD when an individual first meets DSM-5 cri-
education with fewer qualifications and disproportionately high teria for the disorder, regardless of his/her age) should be a rou-
unemployment4, physical ill health5, greater burden of mental tine part of child and youth mental health practice; b) training of
disorders, recurrent self-harm, and a suicide rate of around 8%1. mental health professionals in evidence-based early interven-
The high economic costs of BPD (estimated to be e16,852 per tions should be prioritized; c) indicated prevention (preventing
patient per annum in the Netherlands) are attributable to high the onset of new cases by targeting individuals showing sub-
direct treatment costs and high indirect costs, chiefly work- threshold features of BPD) currently represents the best starting
related disability1. BPD is a stronger predictor of being on dis- point toward developing a comprehensive prevention strategy
ability support than either depressive or anxiety disorders6. for BPD; d) early identification should be encouraged through
Although BPD usually has its onset in the period between workforce development strategies (knowledge about BPD as a
puberty and emerging adulthood (young people)7, delay in the severe mental disorder affecting young people should be dissem-
diagnosis and treatment is the norm, and discrimination a- inated among trainees and clinicians in the child and youth
gainst people with BPD is widespread. Specific treatment is mental health professions; programs should address clinician-
usually only offered late in the course of the disorder, to rela- centred discomfort with the label, mistaken beliefs, and prejudi-
tively few individuals, and often in the form of inaccessible, cial and discriminatory attitudes and behaviour); e) the diagnosis
highly specialized and expensive services4. Accumulating evi- of BPD should not be delayed (non-diagnosis of BPD is discrimi-
dence indicates that such late intervention often reinforces natory because it denies individuals the opportunity to make
functional impairment, disability and therapeutic nihilism. informed and evidence-based treatment decisions, and excludes
The proliferation of knowledge about BPD in adolescents BPD from health care planning, policy and service implementa-
and emerging adults (youth) over the past two decades8,9 tion, ultimately harming the young peoples prospects); f) mis-
has provided a firm basis for establishing early diagnosis and leading terms, or the intentional use of substitute diagnoses,
treatment (early intervention) for BPD and for subthreshold should be discouraged (when sub-threshold BPD is present,
borderline personality pathology7. Several salient issues arise terms such as BPD features or borderline pathology are pre-
from this literature. First, personality disorder begins in child- ferred); g) family and friends should be actively involved as col-
hood and adolescence, and can be diagnosed in young people. laborators in prevention and early intervention (typically, family
Second, DSM-5 BPD is as valid and reliable a diagnosis in and friends are the front line for young people with BPD, and
adolescence as it is in adulthood, based on similarity in preva- their central role should be recognized and supported).
lence, phenomenology, stability and risk factors, marked sep- Research priorities are as follows: a) prevention and early
aration of course and outcome from other disorders, and efficacy intervention for BPD must be integrated with similar efforts
of disorder-specific treatment. Third, BPD is common among for other severe mental disorders, such as mood and psychotic
young people: the estimated prevalence is 1-3% in the com- disorders, acknowledging the equifinal and multifinal path-
munity, rising to 11-22% in outpatients, and 33-49% in inpa- ways for the development of psychopathology; b) building a
tients7,8. Fourth, when BPD is compared with other mental knowledge base for a health care system response to prevention
disorders, it is among the leading causes of disability-adjusted and early intervention for BPD can take two approaches (for
life years (DALYs) in young people9. BPD is also a substantial indicated prevention and early intervention, a critical task is to
financial burden for the families of young people, with esti- identify risk factors for the persistence or worsening of problems,
mated average costs per annum in the US of $14,606 out-of- rather than the onset or incidence of disorder per se; or treat-
pocket, plus $45,573 billed to insurance10. Fifth, the first ment development can be based upon causal mechanisms that
wave of evidence-based treatments has demonstrated that underlie risk, such as environmental adversities); c) novel, low-
structured treatments for BPD in young people are effective4. cost preventive interventions that can be widely disseminated
Finally, the weight of empirical evidence has led the DSM-5 should be developed and evaluated (such interventions will need
and the UK and Australian national treatment guidelines to to be developmentally appropriate, and stage/phase specific,
legitimize the diagnosis of BPD prior to age 18. incorporating stepped care service models); d) education and
The Global Alliance for Prevention and Early Intervention skill development programs for families with a young person
for BPD had its origins at a meeting convened under the aus- with BPD are a key priority for treatment research; e) research
pices of the National Education Alliance for BPD in New York needs to fully quantify the educational, vocational and social

outcomes for young people with BPD; f) further development A. Chanen and C. Sharp are joint first authors of this letter. The Global Alliance
for Prevention and Early Intervention for Borderline Personality Disorder
and validation of brief and user-friendly assessment tools is includes: B. Aguirre, G. Andersen, R. Barkauskiene, A. Bateman, E. Bleiberg, M.
needed to promote the systematic use of standardized evaluation Bohus, R. Brunner, A. Chanen, L. Courey, S. Crowell, F. de Fruyt, M.-P. De
Valdivia, M. Debban e, B. De Clercq, K. Ensink, D. Flynn, P. Fonagy, A. Fossati,
in research and clinical settings; g) detailed health economic A. Fruzetti, L. Gervinskaite-Paulaitiene, M. Goodman, K. Goth, K. Gratz, J. Gun-
data are needed to support prevention and early intervention derson, K. Hall, S.B. Hansen, S. Herpertz, H. Herrman, C. Hessels, P. Hoffman,
J. Hutsebaut, M. Jacobsen, M. Kaess, C. Kaplan, C. Kempinsky, R. Kissell, M.
programs for BPD and should be included in all clinical trials; h) Kongerslev, B. Krueger, P. Luyten, K. Lyons-Ruth, J. Mazza, L. McCutcheon, P.
research identifying methods to improve access to evidence- McGorry, L. Mehlum, A. Miller, C. Mirapeix, A. New, J. Oldham, J. Paris, J.
Rathus, M.E. Ridolfi, T. Rossouw, S. Schlu ter-Mu
ller, C. Schmahl, K. Schmeck,
based treatments and reduce treatment dropout is a priority (this C. Sharp, R. Shiner, E. Simonsen, M. Speranza, B. Stanley, S. Stepp, J. Tackett,
should include novel locations and formats for delivery of treat- . Urnes, R. Verheul, M. Wells, C. Winsper, S. Yen, M. Zanarini; the Interna-
tional Society for the Study of Personality Disorders, the European Society for
ments, such as in schools, out-of-home care, or youth forensic the Study of Personality Disorders, the North American Society for the Study of
settings). Personality Disorders, the National Education Alliance for Borderline Personal-
ity Disorder USA, the National Education Alliance for Borderline Personality
Social and policy priorities include the following: a) BPD Disorder Australia, the National Education Alliance for Borderline Personality
needs to be recognized as a severe mental disorder at all levels Disorder Israel, the National Education Alliance for Borderline Personality Dis-
order Italy, and the Sashbear Foundation.
of the health system; b) evidence-based policy is needed to
address BPD from primary through to specialist care, with the 1. Leichsenring F, Leibing E, Kruse J et al. Lancet 2011;377:74-84.
aim of building a health care system response to prevention 2. Gunderson JG, Stout RL, McGlashan TH et al. Arch Gen Psychiatry 2011;
68:827-37.
and early intervention with young people and those who care
3. Bailey RC, Grenyer BF. Harv Rev Psychiatry 2013;21:248-58.
for them as its focus, and including young people and families 4. Chanen AM. J Clin Psychol 2015;71:778-91.
as partners in the design of such systems; c) discriminatory 5. El-Gabalawy R, Katz LY, Sareen J. Psychosom Med 2010;72:641-7.
6. Ostby KA, Czajkowski N, Knudsen GP et al. Soc Psychiatry Psychiatr Epide-
practices in health care systems must be eliminated, especially
miol 2014;49:2003-11.
regarding BPD as a diagnosis of exclusion from services and 7. Chanen AM, McCutcheon LK. Br J Psychiatry 2013;202:s24-9.
refusing health insurance coverage for people with BPD. 8. Sharp C, Fonagy P. J Child Psychol Psychiatry 2015;56:1266-88.
9. The Public Health Group. The Victorian burden of disease study. Melbourne:
Andrew M. Chanen1, Carla Sharp2, Perry Hoffman3 and the Global Victorian Government Department of Human Services, 2005.
10. Goodman M, Patil U, Triebwasser J et al. J Person Disord 2011;25:59-74.
Alliance for Prevention and Early Intervention for Borderline
Personality Disorder
1
Orygen, National Centre of Excellence in Youth Mental Health & Centre for Youth DOI:10.1002/wps.20429
Mental Health, University of Melbourne, Melbourne, Australia; 2University of Houston,
Houston, TX, USA; 3National Education Alliance for Borderline Personality Disorder, USA
Integrated care for mental, neurological and substance use disorders

in non-specialized health settings: rising to the challenge
Worldwide, mental, neurological and substance use (MNS) that people with these disorders receive care that is effective
disorders are major contributors to the global burden of dis- and affordable, and respects their rights and dignity3,4. In line
ease as estimated by disability adjusted life years, and this is with the World Health Organization (WHO)s leadership in the
rising especially in low- and middle-income countries (LMIC)1. field of global public health, the Mental Health Gap Action
MNS disorders commonly co-occur with other chronic health Programme (mhGAP)5 was initiated, with the objectives to
conditions, both communicable (e.g., HIV/AIDS) as well as non- scale up services and enhance coverage. Through its objectives,
communicable (e.g., diabetes, cardiovascular disease) and, if the mhGAP is contributing towards achieving the targets of the
untreated, worsen the outcome of these conditions. People with Comprehensive Mental Health Action Plan 2013-2020, particu-
MNS disorders and their families are doubly challenged by larly in providing comprehensive, integrated and responsive
stigma that further worsens their quality of life, affects social ac- mental health and social care services in community-based set-
ceptability, employability and interferes with help seeking. tings. The underlying principle of mhGAP is to strengthen non-
Financial resources for developing and maintaining mental specialist primary health care systems and providers to deliver
health services in LMIC are very low. The level of public expen- MNS services, thus facilitating the vital link to integrate mental
diture on mental health is less than US$2 per capita. Further- and physical health6.
more, the number of mental health workers is below 1 per To support countries to strengthen MNS care by non-spe-
100,000 in LMIC compared to over 50 in high-income coun- cialist health care providers, the mhGAP Intervention Guide
tries2. The scarcity and unequal distribution of services means (mhGAP-IG) was developed in 2010 using evidence-based guid-
that 76-85% of people with MNS disorders in LMIC do not ance and extensive stakeholder consultation. The mhGAP-IG
receive the care they need. was translated in over 20 languages and has had widespread
Recognizing the urgent priority to scale up services for MNS application by a range of stakeholders in over 90 countries for
disorders, global initiatives have pressed for reforms to ensure integrated management of priority MNS disorders. It has been

used as a key tool in the phased approach to scale up mental World Bank Group and WHO co-hosted the Out of the Shadows:
health services and reduce the treatment gap on a regional- Making Mental Health a Global Priority meeting in April 2016,
national-subnational level, as a capacity building tool for a wide that emphasized the cross-cutting nature of mental health
range of health professionals and para-professionals, and for issues and the need to integrate mental health services into
developing and updating undergraduate and postgraduate cur- general health systems10.
ricula for health professionals. It has also been used to scale up To realize the goal of universal health coverage, it is essen-
mental health response in emergency settings7,8. tial for health care providers and planners to maximize efforts
The WHO has incorporated feedback and recommendations to scale up care for people with MNS disorders, and the
from international experts as well as latest evidence in the field mhGAP-IG Version 2.0 will be a valuable tool to facilitate this
to update the mhGAP-IG and has now released the mhGAP-IG process.
Version 2.09. The key developments include: content update in
various sections based on new evidence; design changes for Tarun Dua1, Allie Sharma2, Archana Patel3, Fahmy Hanna1, Neerja
Chowdhary1, Shekhar Saxena1
enhanced usability; a streamlined and simplified clinical assess- 1
World Health Organization, Geneva, Switzerland; 2Mental Health Service Corps,
ment that includes an algorithm for follow-up; inclusion of two New York, NY, USA; 3Boston Childrens Hospital, Harvard Medical School, Boston,
new modules (Essential Care and Practice, and Implementa- MA, USA
tion), and revised modules for Psychoses, Child and Adolescent
T. Dua, F. Hanna, N. Chowdhary and S. Saxena are staff members of the WHO.
Mental and Behavioural Disorders, and Disorders due to Sub- The authors alone are responsible for the views expressed in this letter and
stance Use. An interactive electronic version of the mhGAP-IG is they do not necessarily represent the decisions, policy or views of the WHO.
currently under development and will have benefits in terms of

1. Whiteford HA, Degenhardt L, Rehm J et al. Lancet 2013;382:1575-86.
increased ease of use, added functionality and cost savings. 2. World Health Organization. Mental health atlas 2014. Geneva: World
The inclusion of mental health and substance abuse in the Health Organization, 2015.
3. Collins PY, Patel V, Joestl SS et al. Nature 2011;475:27-30.
Sustainable Development Goals (SDGs) at the 70th Session of
4. Patel V, Collins PY, Copeland J et al. Br J Psychiatry 2011;198:88-90.
the United Nations General Assembly in September 2015 has 5. World Health Organization. mhGAP: Mental Health Gap Action Pro-
paved the way for mental health to be integrated into the gramme. Geneva: World Health Organization, 2008.
6. Saxena S, Funk M, Chisholm D. Lancet 2013;381:1970-1.
broader development plans of countries over the next 15 years.
7. World Health Organization. mhGAP: supporting Ebola survivors in Guinea.
There is now fresh impetus for countries to provide sufficient Geneva: World Health Organization, 2016.
financial and human resources for mental health care; im- 8. Budosan B, OHanlon KP, Mahoney J et al. Int J Med Med Sci 2016;8:112-9.
9. World Health Organization. mhGAP intervention guide for mental, neuro-
prove access to care for people with mental illness and their
logical and substance use disorders in non-specialized health settings
families; and integrate mental health care across different sec- Version 2.0. Geneva: World Health Organization, 2016.
tors such as social, education and employment, and implement 10. Kleinman A, Estrin GL, Usmani S et al. Lancet 2016;387:2274-5.
community programmes. In order to initiate a collaborative,
multisectoral commitment to put the mental health agenda at DOI:10.1002/wps.20430
the centre of global health and development priorities, the
Causes and predictors of premature death in first-episode

schizophrenia spectrum disorders
As highlighted by the Forum in the February 2017 issue of specific mortalities was used to compute cause-specific ex-
this journal1, patients with schizophrenia spectrum disorders pected numbers of deaths. Crude standardized mortality ratios
have significantly higher risk of premature death due to sui- (SMRs) were calculated as observed deaths/expected deaths.
cide and physical illness; their expected reduction in life Thirty-one participants (11%) were dead at follow-up (SMR
expectancy is 10-20 years2-4. Since the disorders affect 2-3% of 11.56; 95% CI: 7.86-16.42). Sixteen (6%) died by suicide (SMR
the population, with peak onsets in early adulthood, their 46.50, 95% CI: 26.58-75.51); seven (2.5%) by accidental overdo-
impact on public health is considerable5. ses or other accidents, and eight (2.8%) from physical illnesses,
We report findings from a 10-year prospective study of 281 including three (1%) from cardiovascular illness. Time to death
patients with DSM-IV schizophrenia spectrum disorders was significantly shorter in those who committed suicide com-
recruited consecutively at first treatment in four Nordic catch- pared to the two other groups (mean 1,274 6 1,032 days vs.
ment areas over four years. They were assessed during their 2,706 6 1,046 days for accidents and 3,000 6 792 days for natu-
first week of treatment, with follow-ups after one, two, five and ral deaths, p<0.001). Six (37.5%) of those who died by suicide
ten years6,7. Data were linked to the central registries of per- did so within the first two years (two-year SMR estimate 81.91,
sons and causes of death at Statistics Norway and Statistics 95% CI: 30.05-178.28). Only one accident and no natural deaths
Denmark. Information about two- and ten-year average age- occurred in this period.

All-cause mortality was higher for men than for women. and thus very early compared to studies recruiting at discharge
Univariate analyses showed that those alive at the ten-year fol- from first inpatient treatment or later. The findings can thus
low-up were significantly older at baseline compared to those be seen as an illustration of the particularly high risk for sui-
who died by suicide, and significantly younger than those who cide at this early stage, and underline that mortality estimates
died from other reasons. Those alive had significantly shorter based on multi-episode patient samples significantly underes-
duration of untreated psychosis (DUP), lower baseline rates of timate the suicide risk in schizophrenia spectrum patients.
drug and alcohol misuse, longer education and higher employ- The number of deaths from cardiovascular disorders was low.
ment than those with all-cause deaths. The participants were, however, still in their late thirties and
There were no significant associations with baseline clinical not yet into the main cardiovascular risk period.
symptoms or lifetime/current measures of depression/suicidal In conclusion, we found a high mortality rate during the
first ten years of treatment, with the risk of dying by suicide
behaviors and no significant between-group differences in
being particularly high during the first two years. Long DUP
time to first remission or time being psychotic or in treatment
and substance misuse at baseline were significant predictors
during the first two years (including length/dosage of antipsy-
of all-causes mortality. This is of clinical importance, since
chotic medication and number/length of hospital admissions).
help-seeking behaviors and substance use can be responsive
Measures of depression and suicidal behavior at last follow-up
to interventions.
were, however, significantly higher in those who died by sui-
cide. Ingrid Melle1, Jan Olav Johannesen2,3, Ulrik H. Haahr4,5,
A multinomial logistic regression analysis indicated signifi- Wenche ten Velden Hegelstad2, Inge Joa2,3, Johannes Langeveld2,
cant influences of lower age, longer DUP and baseline alcohol Tor K. Larsen2,6, Stein Ilner Opjordsmoen7,8, Ping Qin9,
Jan Ivar Rssberg7,8, Bjrn Rishovd Rund10,11, Erik Simonsen5,12,
misuse on increasing risk of death by suicide; and of higher Per J.W. Vaglum13, Thomas H. McGlashan14, Svein Friis7,8
age, longer DUP and baseline drug misuse on increasing risk 1
NORMENT K.G. Jebsen Centre for Psychosis Research, Institute of Clinical Medicine,
of death from other reasons. Kaplan-Meyer survival analyses University of Oslo and Oslo University Hospital, Oslo, Norway; 2Division of Psychia-
try, Stavanger University Hospital, Stavanger, Norway; 3Faculty of Social Sciences,
showed that long DUP and baseline substance misuse (alco- University of Stavanger, Stavanger, Norway; 4Early Psychosis Intervention Center,
hol 1 drugs) were significantly increasing risk of all-cause Roskilde, Denmark; 5Institute of Clinical Medicine, University of Copenhagen,
Copenhagen, Denmark; 6Department of Clinical Medicine, University of Bergen,
mortality (Mantel-Cox v2 (3)536.98, p<0.001), with a signifi-
Bergen, Norway; 7Division of Mental Health and Addiction, Oslo University Hospital,
cant contribution of substance misuse also after removing Oslo, Norway; 8Division of Mental Health and Addiction, Institute of Clinical Medi-
overdose deaths. cine, University of Oslo, Oslo, Norway; 9National Center for Suicide Research and
Prevention, Institute of Clinical Medicine, University of Oslo, Oslo, Norway; 10Depart-
Our results confirm previous findings of high mortality rates ment of Psychology, University of Oslo, Oslo, Norway; 11Vestre Viken Hospital Trust,
in patients with schizophrenia spectrum disorders. We clearly Drammen, Norway; 12Institute of Clinical Medicine, University of Copenhagen,
demonstrate for the first time that long DUP is a significant Roskilde, Denmark; 13Department of Behavioural Sciences in Medicine, University of
Oslo, Oslo, Norway; 14Department of Social and Behavioral Health, Yale School of
risk factor for all-cause mortality, including suicides, accidents Medicine, New Haven, CT, USA
and physical illnesses. Long DUP can in this context best be
seen as a marker of problematic help-seeking behaviors, in The study has been supported by grants from the Research Council of Norway;
the Norwegian Department of Health and Social Affairs; the National Council
line with recent register studies reporting that patients with for Mental Health/Health and Rehabilitation; Rogaland County; Oslo County;
schizophrenia dying from physical illnesses enter treatment the Theodore and Vada Stanley Foundation; the Regional Health Research
Foundation for Eastern Region, Denmark; Roskilde County; Lundbeck Pharma,
late8. Eli Lilly, and Janssen-Cilag Pharmaceuticals, Denmark; the National Alliance
That substance use diagnoses increase risk of premature for Research on Schizophrenia and Depression; the US National Institute of
Mental Health; the Regional Health Authority in Western Norway; and the
death in patients with severe mental disorders has been dem- Regional Health Authority in South-Eastern Norway.
onstrated previously9. We here show that also substance use
below the diagnostic threshold for use disorders is a risk factor. 1. Liu NH, Daumit GL, Dua T et al. World Psychiatry 2017;16:30-40.
The strong association between baseline substance misuse 2. Chesney E, Goodwin GM, Fazel S. World Psychiatry 2014;13:153-60.
3. Nordentoft M, Mortensen PB, Pedersen CB. Arch Gen Psychiatry 2011;68:
and all-cause mortality is striking. This can be based in shared
1058-64.
underlying risk factors for suicide, including impulsivity, emo- 4. Tiihonen J, Lonnqvist J, Wahlbeck K et al. Lancet 2009;374:620-7.
tion regulation difficulties and interpersonal problems. The 5. Whiteford HA, Degenhardt L, Rehm J et al. Lancet 2013;382:1575-86.
6. Hegelstad WT, Larsen TK, Auestad B et al. Am J Psychiatry 2012;169:374-80.
effects of substances during intoxication can also increase 7. Melle I, Larsen TK, Haahr U et al. Arch Gen Psychiatry 2004;61:143-50.
impulsive behavior and lack of self-care, adding to risks for 8. Crump C, Winkleby MA, Sundquist K et al. Am J Psychiatry 2013;170:324-33.
accidents or physical illnesses. 9. Hjorthoj C, Ostergaard ML, Benros ME et al. Lancet Psychiatry 2015;2:801-8.
10. Fedyszyn IE, Robinson J, Harris MG et al. Early Interv Psychiatry 2014;8:
The two-year SMR estimate for suicide was >80. Previous 387-95.
studies have shown a particularly high suicide risk before or
during the first months of treatment10. Our participants were DOI:10.1002/wps.20431
recruited through an early treatment and intervention study

A reassessment of the relationship between depression and all-cause
mortality in 3,604,005 participants from 293 studies
As reported in the February issue of this journal1, over three side effects of medications used to treat existing pathologies. Pre-
decades of research suggest that depression is associated with selecting participants on the basis of medical conditions could
an increased risk of all-cause mortality, although some large therefore result in confounding by reverse causality among those
recent studies have found negative or null associations2,3. To who are physically unwell at baseline. Given that somatic symp-
better inform clinical decision making and evidence-based toms that are not confounded by physical conditions are integral
service provision, it is crucial to resolve this discrepancy. to a diagnosis of major depression, studies based on medical
Here we summarize the principal findings of the largest samples that use rating scales (instead of diagnostic interviews
ever investigation of the relationship between depression and that query the source of these somatic symptoms) may be partic-
all-cause mortality, comprising 3,604,005 participants and over ularly likely to misclassify individuals who are of relatively poorer
417,901 deaths, based on a reassessment of 293 studies derived health as depressed. Furthermore, we found that over forty dif-
from 15 systematic reviews. We observed that several factors ferent instruments were used to measure depressive symptoms,
moderate the relationship between depression and mortality, which is problematic due to the considerable content heteroge-
and found no evidence of an association when controlling for neity among commonly used instruments7. Even studies that
comorbid mental disorders and health behaviors (see https:// used the same questionnaire frequently adopted different cutoff
osf.io/svywu/ for the complete report and the extracted data). scores for a probable diagnosis of major depressive disorder. The
The purpose of this reassessment was to better understand interaction of three of the aforementioned points the use of
the features of studies that have sought to address the depres- scales encompassing physical symptoms that may indicate
sion-mortality relationship, to delineate some methodological comorbid medical conditions; the use of samples pre-selected
reasons for heterogeneity between studies (sample size and based on medical conditions; the lack of adjustment for comor-
characteristics, number of deaths and follow-up periods, and bidities when estimating the effect of major depressive disorder
adjustment for mental disorders and health behaviors), and to on mortality points to significant weaknesses in the literature.
explore whether estimates of the relationship between depres- We therefore estimated the association of depression and
sion and mortality on the basis of the methodologically most mortality among studies that used DSM-based structured inter-
rigorous studies differed from those of previous meta-analyses. views requiring the presence of core depressive symptoms (sad
The three main results of the study are as follows. mood or anhedonia) prior to assessing for more general physi-
First, there was a pronounced publication bias4, as indicated cal, somatic and cognitive symptoms, in community-based
by the positive intercept (1.02; 95% CI: 0.72-1.31) of effect esti- samples and based on survival analysis methodology. Only four
mates on their standard errors favoring imprecise studies with estimates (1% of all studies) met these criteria, among which
large positive associations. The largest estimates consistently the pooled hazard ratio was 1.17 (95% CI: 0.75-1.60).
came from studies with small samples, low number of deaths, Given the overall poor quality of the available evidence, we
and brief follow-up periods. are unable to draw strong conclusions about the relationship
Second, only 16 (5%) of the included studies adjusted for between depression and mortality. Studies with large samples,
at least one comorbid mental condition. This is surprising, extensive follow-up periods, adjustment for mental disorders
given that more than half of individuals diagnosed with major and health behaviors, and time-to-event outcomes assessed
depressive disorder suffer from at least one additional comor- using survival analysis methodology are especially needed.
bid mental disorder in their lifetime5. The pooled relative risk More work of a higher quality is also required to examine
(RR) of these 16 estimates (1.08; 95% CI: 0.98-1.18) was smaller which variables related to depression and mortality may modify
than the RR of the 266 estimates that were unadjusted for this relationship. For example, the subsequent onset of health
comorbid mental disorders (1.33; 95% CI: 1.29-1.37). Addition- behaviors such as smoking, drinking and physical inactivity ap-
ally, there was no evidence of an association between depres- pear to play an important role in mediating the risk of adverse
sion and all-cause mortality among the fraction of eight of cardiovascular outcomes among depressed individuals8. This
these estimates that also adjusted for health behaviors (smok- could account for a variety of adverse health outcomes that are
ing, drinking or physical inactivity) (1.04; 95% CI: 0.87-1.21). not limited to cardiovascular disorders. Moreover, the risk of
Third, apart from sample size, follow-up duration, and lack depression and mortality are both influenced by a subset of
of adjustment for important variables, other substantial sources common variables. For example, smoking at baseline is associ-
of heterogeneity between studies emerged. Over two-thirds of ated with increased risk of depression onset at follow-up9, and
the estimates comprised respondents who were pre-selected on smoking is associated with many causes of death10.
the basis of medical conditions. This is problematic, because More rigorous research is needed to better understand wheth-
many symptoms of major depression (e.g., insomnia, fatigue) er depression does, in fact, pose an increased risk of all-cause
are shared with various physical conditions6, or may arise as mortality. We hope that our work will encourage such efforts.

Beyon Miloyan1,2, Eiko Fried3 5. Hasin DS, Goodwin RD, Stinson FS et al. Arch Gen Psychiatry 2005;62:
1 1097-106.
Department of Mental Health, Bloomberg School of Public Health, Johns Hopkins
University, Baltimore, MD, USA; 2Faculty of Health, Department of Psychology, 6. Zimmerman M, Chelminski I, McGlinchey JB et al. J Nerv Ment Dis 2006;
Federation University, Mount Helen, VIC, Australia; 3Department of Psychology, 194:893-7.
University of Amsterdam, Amsterdam, The Netherlands 7. Fried EI. J Affect Disord 2017;208:191-7.
8. Whooley MA, Wong JM. Annu Rev Clin Psychol 2013;9:327-54.
1. Liu NH, Daumit GL, Dua T et al. World Psychiatry 2017;16:30-40. 9. Mojtabai R, Crum RM. Am J Publ Health 2013;103:1656-65.
2. Chwastiak LA, Rosenheck RA, Desai R et al. Psychosom Med 2010;72:817- 10. Carter BD, Abnet CC, Feskanich D et al. N Engl J Med 2015;372:631-40.
22.
3. Eaton WW, Roth KB, Bruce M et al. Am J Epidemiol 2013;178:1366-77.
4. Egger M, Smith GD, Schneider M et al. BMJ 1997;315:629-34. DOI:10.1002/wps.20439
Correction
It has been brought to our attention that the Acknowledgements section of the paper Disorders related to sexuality and gender iden-
tity in the ICD-11: revising the ICD-10 classification based on current scientific evidence, best clinical practices, and human rights
considerations, by Reed et al, published in the October 2016 issue of World Psychiatry, should contain the following additional state-
ment: The authors are grateful to the other members of the 2011-2013 ICD-11 Working Group on Sexual Disorders and Sexual
Health, including R. Coates, J. Cottingham, S. Krishnamurti, A. Marais, E. Meloni Vieira, S. Winter and A. Giami, for their contributions
to the proposals discussed in this article.

WPA NEWS
Report on WPA activities in the triennium 2014-2017

The WPA has delivered on almost all Bill of Rights. Over 60 organizations Childrens Day in 2016, WPA launched a
major aspects of the Action Plan 2014- around the globe have signed up their Bill of Rights for Children and Young
2017 as had been agreed and approved support for this. This Bill of Rights was People. On International Womens Day
by the General Assembly in Madrid in launched in the House of Lords in Lon- in 2017, WPA launched a position state-
September 2014. don in October 2016. The logo of Blue ment on perinatal mental health.
Butterfly is the campaign symbol for Translations. As part of the Action Plan,
Public mental health. A curriculum and Social Justice for People with Mental it was agreed that papers from other
policy statements on the following top- Illness3. A special issue of the Interna- languages should be translated into
ics have been produced: a) gender- tional Review of Psychiatry4 has pub- English. Papers from Spanish and Por-
based interpersonal violence and men- lished these findings in details along tuguese have been translated in special
tal health (chaired by P. Chandra and D. with examples of minimum standards of issues of the International Review of
Stewart); b) childrens physical, emo- service as well as good clinical practice5. Psychiatry that have been published or
tional and sexual abuse (chaired by G. Round table meetings. We had a very are in publication9,10. Papers from Ital-
Milovic and B. Leventhal); c) lesbian, successful round table meeting on vio-
ian, Mandarin and Russian are being
gay, bisexual and transgender (LGBT) lent radicalization hosted by the WPA
translated.
individuals (chaired by P. Levounis and Collaborating Centre in London in Oc-
WPA-Lancet commission on psychiatry.
K. Eckstrand); d) migrant mental health tober 2016 and another one on early
This commission has completed its
(chaired by M. Schouler-Ocak and M. interventions in psychiatry hosted by
data analysis and six writing groups
Kastrup)1; e) individuals with intellec- the Hong Kong College of Psychiatrists
were set up to explore and discuss these
tual disability and their mental health in December 2016. These have been in
findings. The first draft of the report
(chaired by S. Bhaumik)2; f) prisoners addition to round table meetings in Co-
will be ready by September 2017 and
mental health (chaired by A. Forrester lombia, Dominican Republic, Costa Rica,
the full report will be published in time
and M. Piper); g) mental health pro- Mexico and Guatemala6. There was a
for the World Congress of Psychiatry in
motion (several short films about follow-up round table on migrant men-
October 2017.
mental illness have been produced by tal health in Oslo in March 2017.
Diplomas in psychological medicine and
A. Sharma). All these documents and Position statements. The WPA has launch-
in mental health. This initiative, in part-
films are available on the WPA web- ed several position statements which
nership with the University of Mel-
site. are available on the WPA website. These
bourne, aimed at psychiatrists and other
World Mind Matters Day. September 5 include those on migrant mental health
mental health professionals, continues
of each year has been identified as in Europe (with Careif), on migrant
apace and we hope to have it ready later
World Mind Matters Day. In 2015, the mental health in Latin America (with
this year.
WPA launched statements on migrant the Latin American Psychiatric Associa-
WPA collaborating centres. These cen-
mental health. This had major implication, APAL), on high quality training in
tres have been established in London
tions as globally there are several million psychiatry, on five reasons to be a psy-
(K.S. Bhui), Naples (M. Maj), Nairobi (D.
migrants, refugees and asylum seekers. chiatrist, on gender identity and same-
Ndetei), Cape Town (D. Stein), Cairo (T.
The Second World Mind Matters Day in sex orientation, attraction and behav-
Okasha), Hong Kong (L. Lam) and
2016 saw the launch of our findings from iours7, on environmental sustainability,
Bangalore (S. Chaturvedi). Each centre is
a global survey covering 193 countries on recruitment in psychiatry, on suicide
focusing on training, research or policy
(which are member states of the Unit- prevention, on preventive psychiatry,
ed Nations) on discrimination against and on spirituality and religion in psy- and will also act as repository for infor-
persons with mental illness. The WPA chiatry8. Joint declarations on migrant mation for the region11.
looked at discrimination against peo- mental health were launched following WPA goodwill ambassadors. WPA good-
ple with mental illness in four areas: round table meetings in Costa Rica, Gua- will ambassadors have been identified
political, social, economic and personal temala, Mexico and Dominican Repub- in India and three major figures have
in the laws of each country. The levels lic. Other statements included the Buch- agreed to take on this role which focuses
of discrimination remain a cause of arest Declaration on Primary Care Men- on disseminating messages on mental
worry across the globe. Nearly one third tal Health (2015), the Manila Declara- health promotion. M. Agashe (psychia-
of the countries do not allow people tion on Mental Health Promotion trist and film actor/director); M. Shivani
with mental illness to get married. (2016), the Kochi Declaration on Col- (a religious leader) and S. Oberoi (a Bol-
Fewer countries provide supported laborative Care (2015), and the Tbilisi lywood actor) have consented to take on
employment or right to vote. As a result Declaration on Integrated and Collab- this role. We are looking for goodwill
of these findings, the WPA produced a orative Care (2016). On International ambassadors in other countries.

Dinesh Bhugra 3. Bhugra D. World Psychiatry 2015;14:254. 9. Villasenor-Bayardo S, Rojas-Malpica C, Romero
President, World Psychiatric Association 4. Bhugra D. Int Rev Psychiatry 2016;28:335-419. A. Int Rev Psychiatry 2016;28:129-230.
5. Bhugra D. Int Rev Psychiatry 2016;28:342-74. 10. Moreira-Almeida A, Oliveira e Oliveira F. Int
1. Schouler-Ocak M, Wintrob R, Moussaoui D 6. Javed A. World Psychiatry 2016;15:191-2. Rev Psychiatry (in press).
et al. Int J Cult Ment Health 2016;9:216- 7. Bhugra D, Eckstrand K, Levounis P et al. World 11. Bhui KS, Fiorillo A, Stein D et al. World Psychi-
32. Psychiatry 2016;15:299-300. atry 2016;15:300.
2. Bhaumik S, Kiani R, Dasari MM et al. Int J Cult 8. Moreira-Almeida A, Sharma A, Janse van Rens-
Ment Health 2016;9:417-29. burg B et al. World Psychiatry 2016;15:87-8. DOI:10.1002/wps.20435
News from WPA Scientific Sections

The current triennium has seen an in- innovative ways of engaging medical stu- their enthusiasm for future contributions
creasing visibility of Scientific Sections dents to increase their interest in the towards WPAs work.
as important and integral components specialty. Section officers and members are also
of the WPA, especially in supporting the There has always been an added value contributing extensively to the WPA offi-
Associations promotion and dissemina- of the contributions from the newly estab- cial journal World Psychiatry9-11. Their
tion of scientific knowledge around the lished Sections. Current interest in pro- interest and participation in the devel-
globe. We are now having 72 Sections moting intersectional work is equally opment of the chapter on mental disorders
and there is still an interest in having evident from activities of many of the new of the ICD-11 is another ongoing effort to
more of them to cover some other scien- Sections. Furthermore, several Sections improve classification and nosology in
tific sub-specialties. including those on Early Career Psychia- psychiatric practice12-15.
During the triennium there has been trists, Education in Psychiatry and Trans- It is hoped that the current enthusiasm
a noticeable increase in the number of cultural Psychiatry are actively involved of Sections leadership and their commit-
WPA co-sponsored meetings, joint inter- in organizing teaching and training work- ted work will keep on contributing to
sectional activities and other related in- shops aiming to develop leadership skills enhance the quality of scientific knowledge
tersectional accomplishments1,2. Sections in different fields of psychiatry. Whereas
of young psychiatrists.
participation in the recently held WPA this requires dedicated expertise, it is antic-
The WPA Action Plan for 2014-20178 has
International Meeting in Cape Town has ipated that the Sections will offer valuable
continued to be a focal action point for
been outstanding, with more than 26 ses- additions to the professions changing per-
many Sections activities. Sections have
sions in the programme. The Sections al- spectives. It is trusted that Scientific Sec-
adapted the theme of promotion of mental
so organized training workshops, round tions will also continue with their current
health as a priority in their work along with
table discussions and educational activi- contribution to the development of pio-
initiating various programmes in the areas
ties in that well attended meeting. neering approaches in psychiatric practice.
of preventive psychiatry by producing edu-
Scientific Sections have continued to
cational materials for the WPA website3. Afzal Javed
produce position statements and discus-
Whereas Sections enjoy a great degree WPA Secretary for Sections
sion documents on important topics that
of independence within the framework
correspond to their expertise. Adding 1. Javed A. World Psychiatry 2015;14:255-6.
of the statutes and by-laws of the WPA,
these documents to the WPA website did 2. Javed A. World Psychiatry 2016;15:191-2.
the relevant Operational Committee is 3. Kallivayalil RA. World Psychiatry 2015;14:374-5.
generate a lot of interest among the WPA
considering some revisions in the by- 4. Riba M. World Psychiatry 2016;15:88.
membership3,4. Several of these position 5. Moreira-Almeida A, Sharma A, Janse van Rens-
laws that would make the work of Sec- burg B et al. World Psychiatry 2016;15:87-8.
statements have been published in World
tions more in line with current needs 6. Bhugra D, Eckstrand K, Levounis P et al. World
Psychiatry, such as those on spirituality
and expectations. There have also been Psychiatry 2016;15:299-300.
and religion in psychiatry5; gender iden- 7. Shields G, Ng R, Ventriglio A et al. World Psy-
tity and same-sex orientation, attraction some initial discussions on clustering of chiatry 2017;16:113-4.
Sections on the basis of common inter- 8. Bhugra D. World Psychiatry 2014;13:328.
and behaviours6; and recruitment in psy-
9. Wahlbeck K. World Psychiatry 2015;14:36-42.
chiatry7. The Sections on Education and ests and activities. This will hopefully 10. Kasper S, Dold M. World Psychiatry 2015;14:
Early Career Psychiatrists have prepared promote further collaboration and links 107-8.
among various Sections. 11. Young A, Colasanti A. World Psychiatry 2016;
a key document for promoting psychiatry
15:239-41.
as an inspiring medical specialty and as a As we are approaching our next World 12. Luciano M. World Psychiatry 2015;14:375-6.
prospective career for medical students. Congress, Sections are also preparing their 13. Gureje O, Reed GM. World Psychiatry 2016;15:
triennium activity reports for the General 291-2.
This work was presented in Sections busi-
14. Jablensky A. World Psychiatry 2016;15:26-31.
ness meeting at Cape Town. It is hoped Assembly along with plans for their new 15. Maj M. World Psychiatry 2016;15:1-2.
that Sections will prepare further materi- elections. We are encouraging all the Sec-
als for promotion of psychiatry in under- tions to involve younger members in offi- DOI:10.1002/wps.20434
graduate medical education and explore cer and committee positions, to enhance

WPA International Competency-Based Curriculum for Mental
Health Providers on Intimate Partner Violence and Sexual
Violence Against Women
Intimate partner violence (IPV) and public health ecological model of risk fac- ment, psychological first aid, resources,
sexual violence (SV) are public health tors for IPV/SV includes those related to documentation, and psychiatric man-
and human rights problems worldwide the individual, partner, family and commu- agement of related mental health trau-
which have profound effects on the health nity (including social and cultural), which mas). The psychiatric management curric-
and wellbeing of individuals, families and may all contribute5. ulum includes the initiation and moni-
communities1. Both IPV and SV can result in numer- toring of first line methods indicated for
The World Health Organization (WHO) ous physical (including death) and men- the treatment of IPV/SV psychological
defines IPV as behaviour by an intimate tal health sequelae, thereby making it trauma, such as cognitive behavioural
partner that causes physical, sexual or imperative that health care professionals therapy with a focus on the trauma, ex-
psychological harm, including acts of know the risk factors, how to assist dis- posure therapy, eye movement desensi-
physical aggression, sexual coercion, psy- closure and safely respond6. Mental tization and reprocessing, pharmacol-
chological abuse or controlling behav- health sequelae may include depression, ogical interventions and comprehensive
iours2. It may be perpetrated by a current anxiety, post-traumatic stress disorder, care for complex post-traumatic stress
or past intimate partner, occur in het- psychosis, self-harm, sexual problems, disorder.
erosexual or same-sex relationships, and inability to trust others, and a host of Different types of educational tools are
include stalking. At its core it is a means psychosomatic conditions and risky be- employed in the curriculum to enhance
to control and dominate the abused haviours that may be referred. There- knowledge, attitudinal change and skills,
partner. fore, psychiatrists must be familiar with and thereby provide real life competen-
SV is defined by WHO as any sexual the best evidence-based short- and long- cies. The curriculum links to WHO clini-
act, attempt to obtain a sexual act, term management of IPV and SV mental cal and policy guidelines9, a WHO clinical
unwanted sexual comments or advan- health sequelae to best assist victims6. handbook2, key paper abstracts, a list
ces, or acts to traffic or otherwise direct- Evidence shows that 30% or more (de- of books, manuals and toolkits, and a
ed against a persons sexuality using pending on location and presenting teaching set of powerpoints on IPV and
coercion, by any person regardless of symptoms) of psychiatric patients have SV. It provides twelve international case
their relationship to the victim, in any been exposed to IPV or SV7. As IPV and vignettes with teaching points and two
setting, including but not limited to SV are often not disclosed, or enquired video-based learning vignettes in which
home and work2,3. It includes marital, about by psychiatrists, this may affect two senior psychiatrists each interview a
dating, stranger and acquaintance rela- diagnosis, treatment and outcome. Needs woman who has experienced IPV.
tionships, sexual slavery, forced marriage assessments have indicated that IPV and A trauma-informed model of care ad-
or cohabitation and wife inheritance and SV are key determinants of womens men- vocated by WHO uses the acronym
may occur during peace or war2,3. It may tal health, but 60% of mental health pro- LIVES, where L means listen (empath-
also take place when someone is not fessionals report that they lack adequate ic and non-judgmental), I means in-
able to give informed consent (e.g., a knowledge and want more education on quire (about needs and concerns), V
child or a person who is intoxicated, these topics8. means validate (believe and understand
drugged, asleep or physically or mentally The WPA Action Plan for 2014-2017 the victim), E means enhance safety
incapacitated). developed by President D. Bhugra listed (help protect against further harm), and
While IPV and SV can occur to both IPV and SV as priorities for a position S means support (help connect to serv-
women and men, the most serious forms paper and curriculum. Under the leader- ices and social support)2.
overwhelmingly happen to women at the ship of D.E. Stewart (Canada) and P.S. A WPA position paper on IPV/SV was
hands of men2,3. Chandra (India) and a Steering Group of also developed by the Steering Group of
Prevalence rates of IPV/SV varied from six experts from across the globe (three experts.
15 to 71% (lifetime) or 4 to 54% (last 12 from WHO), plus two educational con- The curriculum and the position paper
months) across a 10-country study by sultants, a competency-based curriculum were approved by the WPA Executive
WHO4. Prevalence is greatly underre- for medical students, psychiatric trainees Committee in July 2016 and posted on the
ported due to guilt, shame, social stigma, and practicing psychiatrists was devel- WPA website (www.wpanet.org). Initial
inadequate social support, financial or oped. This includes suggestions of how to response has been encouraging, with sev-
emotional dependence on the perpetrator, test nine core competencies and their eral universities, medical schools and
or fears for safety, child custody, immi- related subtopics (definitions, prevalence, non-governmental organizations across
gration status, or other repercussions3. A misconceptions, health sequelae, assess- five continents requesting permission to

use the curriculum in whole or in part. The 1. Stewart DE, Vigod S, Riazantseva E. Curr Psy- 6. Stewart DE, MacMillan H, Wathen N. Can J
chiatry Rep 2016;18:4. Psychiatry 2013;58:1-15.
curriculum has been presented at several 2. World Health Organization. Health care for 7. Oram S, Trevillion K, Feder G et al. Br J Psychi-
annual national psychiatric association women subjected to intimate partner violence atry 2013;202:94-9.
meetings (with more planned) and has or sexual violence. A clinical handbook. Geneva: 8. Nyame S, Howard LM, Feder G et al. J Ment
World Health Organization, 2014. Health 2013;22:536-43.
been featured in news articles. We wel- 3. World Health Organization/London School of 9. World Health Organization. Responding to inti-
come comments on the curriculum and Hygiene and Tropical Medicine. Preventing mate partner violence and sexual violence a-
feedback about its use at donna.stewart@ intimate partner and sexual violence against gainst women: WHO clinical and policy guide-
women: taking action and generating evidence. lines. Geneva: World Health Organization,
uhn.ca. Geneva: World Health Organization, 2010. 2013.
4. Ellsberg M, Jansen HAFM, Heise L et al. Lancet
Donna E. Stewart1, Prabha S. Chandra2 2008;371:1165-72.
1 DOI:10.1002/wps.20432
University Health Network Centre for Mental Health, 5. World Health Organization. World report on
University of Toronto, Toronto, Canada; 2Department violence and health. Geneva: World Health Or-
of Psychiatry, National Institute of Mental Health and ganization, 2002.
Neurosciences, Bangalore, India

WPS-C2-C3 4/19/17 11:15 AM Page 1
The World Psychiatric Association (WPA) World Psychiatry
The WPA is an association of national psychiatric societies World Psychiatry is the official journal of the World
aimed to increase knowledge and skills necessary for work in Psychiatric Association. It is published in three issues per year
the field of mental health and the care for the mentally ill. Its and is sent free of charge to psychiatrists whose names and
member societies are presently 138, spanning 118 different addresses are provided by WPA member societies and sections.
countries and representing more than 200,000 psychiatrists. Research Reports containing unpublished data are welcome
The WPA organizes the World Congress of Psychiatry every for submission to the journal. They should be subdivided into
three years. It also organizes international and regional con- four sections (Introduction, Methods, Results, Discussion).
gresses and meetings, and thematic conferences. It has 72 sci- References should be numbered consecutively in the text and
entific sections, aimed to disseminate information and pro- listed at the end according to the following style:
mote collaborative work in specific domains of psychiatry. It 1. Cuijpers P, Sijbrandij M, Koole SL et al. Adding psychother-
has produced several educational programmes and series of apy to antidepressant medication in depression and anxi-
books. It has developed ethical guidelines for psychiatric prac- ety disorders: a meta-analysis. World Psychiatry 2014;13:
tice, including the Madrid Declaration (1996). 56-67.
Further information on the WPA can be found on the web- 2. McRae TW. The impact of computers on accounting.
site www.wpanet.org. London: Wiley, 1964.
3. Fraeijs de Veubeke B. Displacement and equilibrium mod-
els in the finite element method. In: Zienkiewicz OC,
WPA Executive Committee
Hollister GS (eds). Stress analysis. London: Wiley, 1965:145-
President D. Bhugra (UK)
97.
President-Elect H. Herrman (Australia)
All submissions should be sent to the office of the Editor.
Secretary General R.A. Kallivayalil (India)
Secretary for Finances A. Soghoyan (Armenia)
Editor M. Maj (Italy).
Secretary for Meetings M. Takeda (Japan)
Editorial Board D. Bhugra (UK), H. Herrman (Australia), R.A.
Secretary for Education E. Belfort (Venezuela)
Kallivayalil (India), A. Soghoyan (Armenia), M. Takeda (Japan), E.
Secretary for Publications M. Riba (USA)
Belfort (Venezuela), M. Riba (USA), A. Javed (UK/Pakistan).
Secretary for Sections A. Javed (UK/Pakistan)
Advisory Board H.S. Akiskal (USA), R.D. Alarcn (USA), J.A.
Costa e Silva (Brazil), J. Cox (UK), M. Jorge (Brazil), H. Katschnig
WPA Secretariat (Austria), F. Lieh-Mak (Hong Kong-China), F. Lolas (Chile), J.E.
Geneva University Psychiatric Hospital, 2 Chemin du Petit Bel- Mezzich (USA), D. Moussaoui (Morocco), P. Munk-Jorgensen
Air, 1225 Chne-Bourg, Geneva, Switzerland. Phone: (Denmark), F. Njenga (Kenya), A. Okasha (Egypt), J. Parnas
+41223055737; Fax: +41223055735; E-mail: wpasecretariat@ (Denmark), V. Patel (India), P. Ruiz (USA), N. Sartorius
wpanet.org. (Switzerland), A. Tasman (USA), S. Tyano (Israel), J. Zohar
(Israel).
Office of the Editor Department of Psychiatry, University of

Naples SUN, Largo Madonna delle Grazie, 80138 Naples,
Italy. Phone: +390815666502; Fax: +390815666523; E-mail:
majmario@tin.it.
Acknowledgement
This publication has been partially supported by an unrestricted
educational grant from Janssen-Cilag SpA,
which is hereby gratefully acknowledged.
World Psychiatry is indexed in PubMed, Current Contents/Clinical Medicine, Current Contents/Social 2017 by WPA Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons
and Behavioral Sciences, Science Citation Index, and EMBASE. or property as a matter of products liability, negligence or otherwise, or from any use or oper-
ation of any methods, products, instructions or ideas contained in the material herein.
All back issues of World Psychiatry can be downloaded free of charge from the PubMed system Because of rapid advances in the medical sciences, in particular, independent verification of
(http://www.pubmedcentral.nih.gov/tocrender.fcgi?journal=297&action=archive). diagnoses and drug dosages should be made.
WPS-C1SP 4/20/17 4:28 PM Page 1
WPA
World Psychiatry
WORLD PSYCHIATRY
OFFICIAL JOURNAL OF THE WORLD PSYCHIATRIC ASSOCIATION (WPA)
Volume 16, Number 2 June 2017
EDITORIALS Shared decision making in mental health care 158

settings: perspective, purpose and practice
Shared decision making: everyone wants it, 117 S. TSE
so why isnt it happening?
Incorporating shared decision making in mental 160
A. COULTER
health care requires translating knowledge
Migration and psychosis: our smoking lung? 119 from implementation science
Vol. 16, No. 2, pp. 117-224

J.B. KIRKBRIDE I. SCHOLL, P.J. BARR
Mental health shared decision making in the US 161
SPECIAL ARTICLES R.E. DRAKE
Etiology in psychiatry: embracing the reality 121
of poly-gene-environmental causation of RESEARCH REPORTS
mental illness Prevalence, incidence and mortality from 163
R. UHER, A. ZWICKER cardiovascular disease in patients with pooled
The contemporary refugee crisis: an overview 130 and specific severe mental illness: a large-scale
of mental health challenges meta-analysis of 3,211,768 patients and 113,383,368
D SILOVE, P. VENTEVOGEL, S. REES controls
C.U. CORRELL, M. SOLMI, N. VERONESE ET AL
PERSPECTIVES Has the rising placebo response impacted 181
antidepressant clinical trial outcome? Data from
The clinical relevance of appraisals of psychotic 140 the US Food and Drug Administration 1987-2013
experiences A. KHAN, K.F. MAR, J. FAUCETT ET AL
P.A. GARETY, A. HARDY
Risk of suicide, deliberate self-harm and 193
Mating, sexual selection, and the evolution 141 psychiatric illness after the loss of a close relative:
of schizophrenia a nationwide cohort study
M.D. GIUDICE M.-B. GULDIN, M.I.S. KJAERSGAARD, M. FENGER-GRN ET AL
Validity and utility of the general factor 142
of psychopathology REAPPRAISAL
B.B. LAHEY, R.F. KRUEGER P. RATHOUZ ET AL A critique of the ultra-high risk and transition 200
Neuroticism is a fundamental domain of personality 144 paradigm
with enormous public health implications J. VAN OS, S. GULOKSUZ
FEBRUARY 2017
T.A. WIDIGER, J.R. OLTMANNS
INSIGHTS
FORUM SHARED DECISION MAKING
IN MENTAL HEALTH CARE Treatment of people at ultra-high risk 206
for psychosis
Implementing shared decision making in routine 146 A.R. YUNG
mental health care Persistent persecutory delusions: the spirit, 207
M. SLADE style and content of targeted treatment
D. FREEMAN, F. WAITE
Commentaries Does neuroimaging have a role in predicting 208
Shared decision making: a consideration of 154 outcomes in psychosis?
historical and political contexts P. MCGUIRE, P. DAZZAN
G. MEADOWS The role of expectations in mental disorders 209
Involvement in decision making: the devil is 155 and their treatment
in the detail W. RIEF, J.A. GLOMBIEWSKI
R. MCCABE
Psychiatric practice: caring for patients, 156 LETTERS TO THE EDITOR 211
collaborating with partners, or marketing to
consumers? WPA NEWS 220
D.J. STEIN
Common sense alone is not enough 157
S. PRIEBE
IMPACT FACTOR: 20.205 ISSN 1723-8617

World Psychiatry June 2017

Uploaded by

Copyright:

Available Formats

World Psychiatry June 2017

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

World Psychiatry June 2017

Uploaded by

Copyright:

Available Formats

WPS-C1SP 4/26/17 7:54 AM Page 1

Volume 16, Number 2 June 2017

EDITORIALS Shared decision making in mental health care 158

Vol. 16, No. 2, pp. 117-224

Neuroticism is a fundamental domain of personality 144

IMPACT FACTOR: 20.205 ISSN 1723-8617

The World Psychiatric Association (WPA) World Psychiatry

Office of the Editor Department of Psychiatry, University of

Shared decision making: everyone wants it, so why isnt it happening?

World Psychiatry 16:2 - June 2017 117

118 World Psychiatry 16:2 - June 2017

Migration and psychosis: our smoking lung?

World Psychiatry 16:2 - June 2017 119

120 World Psychiatry 16:2 - June 2017

Etiology in psychiatry: embracing the reality of

(World Psychiatry 2017;16:121129)

World Psychiatry 16:2 - June 2017 121

Autism Schizophrenia Bipolar disorder Depression

Number of genetic variants 4 128 18 17

Odds ratio of the most strongly 1.17 1.21 1.15 1.05

Proportion of variance explained 14% 23% 25% 5%

Figure 1 The heritability gap. Heritability (the proportion of causation

122 World Psychiatry 16:2 - June 2017

tionship between severe life events and psychopathology onset

World Psychiatry 16:2 - June 2017 123

124 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 125

126 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 127

128 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 129

The contemporary refugee crisis: an overview of mental health

(World Psychiatry 2017;16:130139)

130 World Psychiatry 16:2 - June 2017

OSCILLATIONS IN PUBLIC PERCEPTIONS AND

World Psychiatry 16:2 - June 2017 131

132 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 133

134 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 135

136 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 137

138 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 139

The clinical relevance of appraisals of psychotic experiences

140 World Psychiatry 16:2 - June 2017

Mating, sexual selection, and the evolution of schizophrenia

World Psychiatry 16:2 - June 2017 141

Validity and utility of the general factor of psychopathology

142 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 143

144 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 145

Implementing shared decision making in routine mental health care

(World Psychiatry 2017;16:146153)

146 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 147

148 World Psychiatry 16:2 - June 2017

World Psychiatry 16:2 - June 2017 149

150 World Psychiatry 16:2 - June 2017