DSM Versus RDoC

HHS Public Access
Author manuscript
Annu Rev Clin Psychol. Author manuscript; available in PMC 2017 February 03.
Author Manuscript
Published in final edited form as:

Annu Rev Clin Psychol. 2016 ; 12: 435–463. doi:10.1146/annurev-clinpsy-021815-093122.
Clashing Diagnostic Approaches: DSM-ICD versus RDoC

Scott O. Lilienfeld, Ph.D. and Michael T. Treadway, Ph.D.
Emory University
Abstract
Since at least the middle of the past century, one overarching model of psychiatric classification,
namely, that of the Diagnostic and Statistical Manual of Mental Disorders and International
Author Manuscript
Classification of Diseases (DSM-ICD), has reigned supreme. This DSM-ICD approach embraces
an Aristotelian view of mental disorders as largely discrete entities that are characterized by
distinctive signs, symptoms, and natural histories. Over the past several years, however, a
competing vision, namely, the Research Domain Criteria (RDoC) initiative launched by the
National Institute of Mental Health, has emerged in response to accumulating anomalies within the
DSM-ICD system. In contrast to DSM-ICD, RDoC embraces a Galilean view of psychopathology
as the product of dysfunctions in neural circuitry. RDoC appears to be a valuable endeavor that
holds out the long-term promise of an alternative system of mental illness classification. We
delineate three sets of pressing challenges – conceptual, methodological, and logistical/pragmatic
– that must be addressed for RDoC to realize its scientific potential, and conclude with a call for
further research, including investigation of a rapprochement between Aristotelian and Galilean
approaches to psychiatric classification.
Author Manuscript
“I can calculate the motion of heavenly bodies, but not the madness of people.”
-Sir Isaac Newton
The contentious field of psychiatric classification has long been marked by an analogous
clash of Aristolelian and Galiliean worldviews (Carson, 1996). The approach enshrined in
recent editions of the Diagnostic and Statistical Manual of Mental Disorders (DSM) and the
International Classification of Diseases (ICD), which we heretofore dub the DSM-ICD
model, adopts – at least implicitly – an Aristotelian model of categorization. In this
approach, which has been dominant in American psychiatry over at least the past half
century, psychiatric disorders are presumed to constitute largely discrete entities: They are
commonly assumed to differ qualitatively from normality and from each other. For example,
Author Manuscript
the influential framework for the validation of mental disorders advanced by Robins and
Guze (1970) delineated five criteria for ascertaining whether a diagnostic entity is genuine.
Among them was the “delimitation [of the disorder] from other disorders” (p. 108), the
Corresponding author: Scott O. Lilienfeld, Ph.D., Department of Psychology, Room 473, Emory University, 36 Eagle Row, Atlanta,
Georgia 30322. slilien@emory.edu.
Permission Source:
https://www.google.com/search?q=NIMH+RDoC
+matrix&biw=1920&bih=935&source=lnms&tbm=isch&sa=X&ved=0CAcQ_AUoAmoVChMIvNCH79XYxwIVCKKACh1ZnwUa
&dpr=1#imgrc=pFkp7ofm8D37bM%3A
Lilienfeld and Treadway Page 2
assumption being that a diagnosis that overlaps extensively with other diagnoses is of
Author Manuscript
doubtful validity.
DSM-ICD and Kraepelin: Hints of a Paradigm Shift

The reigning DSM-ICD approach is commonly referred as “neo-Kraepelinian” (Blashfield,
2002) in recognition of the contributions of the pioneering German systematist Emil
Kraepelin, who initially believed that superficially similar mental disorders, such as
dementia praecox (now termed schizophrenia) and manic-depression (now termed bipolar
disorder), could be differentiated by a detailed documentation of their (a) signs (observable
manifestations), (b) symptoms (subjective reports), and (c) natural history (trajectory over
time). Kraepelin, following in the footsteps of Linnean botanists (Compton & Guze, 1995),
regarded different mental disorders as akin to differing species or subspecies that could be
distinguished largely by their topography.
Author Manuscript
A key assumption of the neo-Kraepelian approach is that diagnostic signs and symptoms are
sufficient to differentiate mental disorders. Nevertheless, as the history of medicine reminds
us, this assumption may be unwarranted. For example, in the early 20th century, physicians
diagnosed dozens of different fevers – ranging alphabetically from blackwater fever to
yellow fever – and distinguished them on the basis of other co-occurring signs and
symptoms (Kihlstrom, 2002). Today, physicians recognize that fever is a nonspecific
indicator of a plethora of underlying pathologies rather than a disorder per se, and instead
attempt to identify the etiology of the fever prior to initiating treatment.
Despite the enormous impact of the DSM on everyday research and practice, there are
growing indications that its hegemony may at last be beginning to wane. Over the past
several years, rumblings of what some have described as a Kuhnian “paradigm shift” (Fu &
Author Manuscript
Costafreda, 2013) or at least a “reboot” (Kendler, 2015) in psychiatric classification have

become increasingly audible Among the first responders has been the NIMH’s Research
Domain Criteria initiative (RDoC; Insel et al., 2010), Though more of promissory note than
a formalized system at present, the RDoC aims to develop a system of psychiatric
classification based not on signs, symptoms, and course, ala Kraepelin, but instead on
markers of psychobiological systems linked to adaptive – and maladaptive – functioning.
In the pages to follow, we briefly examine the successes of the DSM-ICD approach, discuss
the anomalies that have led scholars to question its scientific status, and present the raison
d’etre of RDoC as a competing model. We focus on RDoC’s scientific potential, but also on
the often unappreciated challenges standing in the way of realizing this potential. We
conclude with recommendations for future directions in classification, including the
Author Manuscript
possibility of a rapprochement between these two overarching approaches.
DSM and ICD: Origins and Assumptions

The history of the DSM and its revisions has been recounted in numerous sources (e.g.,
Lieberman & Ogas, 2015; Lilienfeld, Smith, & Watts, 2014; Widiger & Clark, 2000; Wilson,
1993), so we reprise it only very briefly here. In response to the perceived shortcomings with
the first two DSMs, especially their often vague diagnostic descriptions and the low or best
modest inter-rater reliabilities of their diagnostic categories, the American Psychiatric

Author Manuscript
Association, with psychiatrist Robert Spitzer at the helm, released DSM-III in 1980
(American Psychiatric Association, 1980). DSM-III was influenced heavily by the seminal
work of the St. Louis psychiatric group at Washington University, who had introduced
preliminary diagnostic criteria and algorithms (decision rules) for 14 major mental disorders
in the early 1970s (Feighner et al., 1972). Another noteworthy and more immediate
precursor of DSM-III were the Research Diagnostic Criteria (RDC; Spitzer, Endicott, &
Robins, 1978), which delineated diagnostic criteria and algorithms for approximately 20
major diagnostic categories, along with subtypes within several of these categories.
Like the Feighner and RDC criteria, DSM-III was considerably more neo-Kraepelinian than
its DSM predecessors (Compton & Guze, 1995), as it emphasized the differentiation of
conditions on the basis on their signs, symptoms, and natural history. In accord with its neo-
Kraepelinian emphasis, DSM-III provided users with (a) standardized diagnostic criteria and
Author Manuscript
(b) algorithms, for each diagnosis. Rather than merely describing each diagnosis, as DSM-I
and DSM-II had done, DSM-III delineated the specific signs and symptoms comprising each
diagnosis and the method by which they needed to be combined to establish it. In this way, it
almost certainly boosted the inter-rater reliability of most diagnostic categories, although
some critics have maintained that these increases were modest at best (Kirk & Kutchins,
1992). Although DSM-III-R and DSM-IV introduced a number of important changes to
numerous diagnoses, they retained DSM-III’s basic structure and format.
In an effort to accommodate novel evidence that had emerged in the wake of the second
millennium, DSM-5, spearheaded by David Kupfer and Darrel Regier, was published in May
of 2013 amidst a host of searing criticisms, many of which centered on major alterations to a
number of diagnostic categories in the absence of adequate data (see Frances & Widiger,
Author Manuscript
2012). Although the early phases of planning for DSM-5 were rife with speculations
regarding drastic changes in its content and structure, such as a heightened focus on
neuroscientific markers or the inclusion of a potential dimensional system for personality
disorders (Skodol et al., 2011), DSM-5 by and large retained the principal format of DSM-
IV, as well as most of its major diagnoses. Although DSM-5 is the predominant system of
psychiatric classification around the world, Chapter V of the ICD, 10th revision (ICD-10) of
the World Health Organization (1993) is a neo-Krapelinian alternative to DSM-5 (which is
similar to the DSM in most important respects) that has been adopted in numerous countries
outside of the U.S (ICD-11 is under construction as of this writing).
The hue and cry regarding the most controversial changes in DSM-5, such as the deletion of
the bereavement criterion for major depressive disorder or the jettisoning of hypochondriasis
Author Manuscript
as a diagnosis (Frances, 2014), may have obscured a deeper point. With the potential
exception of (a) the inclusion of dimensional scales to capture the functioning and
impairment associated with major mental disorders and (b) the inclusion of a hybrid
prototype-dimensional model for personality disorders, both of which appeared in Section
III (“Emerging Measures and Models”), DSM-5 was every bit as neo-Kraepelinian as its
predecessors. Thus, despite initial hopes – or fears, depending on one’s perspective –
DSM-5 did not constitute a radical shift in classification or diagnosis, at least with respect to
its central presuppositions.
DSM-ICD Successes
Author Manuscript
Extreme skeptics of the DSM-ICD system (e.g., Greenberg, 2015) have at times argued that
this approach has yielded psychiatric categories with negligible validity. Similarly, in a
widely publicized blog posting announcing the shift of the National Institute of Mental
Health (NIMH) toward RDoC, NIMH Director Thomas Insel (2013) wrote that the DSM’s
“major weakness is its lack of validity.”
The assertion that DSM categories “lack validity” paints with too broad a brush. Many DSM
categories, such as major depression, panic disorder, bipolar disorder, obsessive-compulsive
disorder, and schizophrenia, clearly display at least some construct validity, as demonstrated
by consistent relations with laboratory indicators, biological correlates, natural history, and
family history. For example, although the DSM diagnosis of schizophrenia almost surely
encompasses a heterogeneous mélange of overlapping conditions, this diagnosis is
Author Manuscript
associated with certain external correlates, such as smooth pursuit eye movement
dysfunction, ventricular enlargement, a chronic and frequently relapsing course, and a
family history of schizophrenia among biological relatives (e.g., Tsuang, Stone, & Faraone,
2000).
Moreover, the development of lists of empirically supported therapies (ESTs), which are
psychological treatments that have been demonstrated to be efficacious for specific disorders
(e.g., cognitive-behavioral therapy for major depression, panic control treatment for panic
disorder; Chambless & Ollendick, 2001) is a testament to the treatment validity (Hayes,
Nelson, & Jarrett, 1996) of at least some DSM-ICD categories. .If the DSM and ICD were
entirely devoid of validity, one could not identify psychological interventions that are
differentially efficacious for different conditions (Garb, Lilienfeld, Nezworski, Wood, &
O’Donohue, 2009; but see “Inadequate treatment validity”). Moreover, by providing
Author Manuscript
researchers and clinicians with a lingua franca for facilitating diagnostic communication, the
DSM has greatly accelerated scientific progress regarding the correlates, etiology, and
treatment of mental disorders (see Kendell, 1975, for a broader discussion). These
impressive achievements are not to be dismissed lightly.
DSM-ICD Anomalies
At the same time, it has become evident that the DSM-ICD approach has been beset by a
growing number of anomalies (Lilienfeld, 2014b), many of which have not been adequately
resolved across DSM editions; we touch on the most salient of these conundrums here.
Scientifically arbitrary diagnostic cut-offs—The DSM is technically agnostic with

regard to whether the conditions it contains are categorical (taxonic) or dimensional in
Author Manuscript
nature. Still, the DSM adopts a categorical approach as a working model for measurement
purposes. This model is debatable for two major reasons. First, data from taxometric studies,
which allow investigators to ascertain whether an observed distribution can be decomposed
into two or more independent distributions (Meehl & Golden, 1982), indicate that most
DSM conditions, including the lion’s share of mood, anxiety, eating, personality, and
externalizing disorders, appear to be underpinned by one or more dimensions. The potential
exceptions are schizophrenia spectrum disorders, autism spectrum disorders, and some
substance use disorders (Haslam, Holland, & Kuppens, 2012). The absence of a point of
Author Manuscript
rarity (Sneath, 1957) demarcating most DSM conditions from normality raises questions
regarding the Aristotelian assumption of discrete conditions. Second, even if it turns out that
some DSM conditions are taxonic, it would not justify the imposition of a categorical
measurement model. A categorical measurement model decreases reliability and validity
relative to a dimensional model, because it forfeits valuable psychometric information
(Markon, Chmielewski, & Miller, 2011). Even taxonic variables, dimensional measures
almost always provide more sensitive indicators than do categorical measures, as they offer
information regarding individuals’ proximity to the natural cut-point.
Heterogeneity—The polythetic model of DSM-III-R and later DSM editions, sometimes

derided as the “Chinese menu” approach, provides criteria that are neither singly necessary
nor jointly sufficient for a diagnosis. This model has generated marked phenotypic
heterogeneity. In DSM-5, for example, 256 different sign/symptom combinations are
Author Manuscript
compatible with a diagnosis of borderline personality disorder. As an even more extreme

example, in DSM-5, there are 636,120 ways to meet diagnostic criteria for PTSD (Galatzer-
Levy & Bryant, 2013). It is even possible for two people to meet DSM-5 criteria for
obsessive-compulsive personality disorder and to share no diagnostic criteria. Although
phenotypic heterogeneity does not necessarily imply etiological heterogeneity, it seems
unlikely that two conditions that overlap minimally in their expression would stem from
similar, let alone identical, causal influences. In addition, such heterogeneity renders the
search for etiological agents more challenging (see Monroe & Anderson, 2015, for a
discussion of this issue with respect to major depressive disorder).
Comorbidity—An ideal taxonomy yields categories that are largely mutually exclusive,
with few or no intermediate cases (Frances, 1980). Yet across its multiple editions, the DSM
Author Manuscript
has consistently been bedeviled by a vexing problem termed comorbidity. This term was
coined by medical epidemiological Alvan Feinstein (1970) to denote the co-occurrence of a
“distinct clinical entity that has existed or that may occur during the clinical course of a
patient who has the index disease under study” (pp. 456-457). Although some scholars have
voiced concerns about the use of the “comorbidity” concept in most cases of
psychopathology research given that it is unclear how many DSM conditions are distinct
clinical entities (Lilienfeld, Waldman, & Israel, 1994), extremely high levels of covariation
among ostensibly separable conditions may raise questions concerning their etiological
independence.
The extent of the comorbidity problem is difficult to overstate. In the Australian National
Survey of Mental Health and Well-Being, 21% of participants with one DSM-IV disorder
Author Manuscript
met criteria for three or more DSM-IV disorders (Andrews et al., 2002); comparable data
derive from the United States National Comorbidity Study (Kessler et al., 1994). The
comorbidity problem is especially glaring in the personality disorders domain. In one study
of over 1000 patients (Stuart et al., 1998), patients with one DSM personality disorder met
criteria for an additional 1.7 DSM-III-R personality disorders on average; 9.4% met criteria
for 4 or more personality disorders. Notably, only 18% of patients with a personality
disorder did not meet criteria for at least one other personality disorder, indicating that
comorbidity is much more the rule than the exception. In one especially extreme case, a
Author Manuscript
participant in a research study simultaneously met criteria for all 10 DSM-IV (and DSM-5)
personality disorders (Lilienfeld, Smith, & Watts, 2013)!
In the eyes of many authors, the presence of rampant comorbidity is a red flag that the DSM
system is not drawing the correct diagnostic borders. Other authors (e.g., Maj, 2005) go
further, suggesting that such comorbidity reflects the propensity of the DSM to attach
different names to slightly different manifestations of a shared predisposition, a logical error
known as the jangle fallacy (Block, 1995). For example, many or most personality disorders
appear to be complex constellations or configurations of normal-range personality
dimensions (e.g., antagonism, low conscientiousness, introversion; see Widiger & Trull,
2007). If so, it is hardly surprising that many of these conditions, such as narcissistic and
borderline personality disorders, display substantial covariation, as it likely that these
disorders reflect highly overlapping densifications in multivariate space.
Author Manuscript
Large number of intermediate cases—As noted earlier, an optimal classification

system consists of categories that yield few intermediate cases (Frances, 1980). Yet for most
major classes of psychopathology, one of the most frequent diagnoses is NOS (Not
Otherwise Specified), meaning that most patients with mental disorders do not fit into any
extant category (Stein, Black, & Pienaar, 2000; Westen, 2012). Such patients must be
diagnosed by means of a “wastebasket category” that encompasses individuals who are
clearly disordered but whose signs and symptoms do not meet criteria for any extant
diagnosis. For eating disorders, NOS has typically been the modal diagnosis in routine
clinical settings (Fairburn & Bohn, 2005); in unstructured interviewed studies of personality
disorders, NOS is similarly the most frequent diagnosis (Verheul & Widiger, 2004).
Author Manuscript
The high prevalence of NOS diagnoses probably derives from what Hyman (2010) termed
the “problem of overspecification” (p. 166). For its diagnoses, the DSM-ICD prescribe strict
criteria and algorithms that almost certainly exclude many individuals who suffer from the
same dysfunction as do most individuals within the category, but who fall just short of the
diagnostic threshold. For example, in a research study an individual who met all of the
criteria for DSM-5 major depressive disorder (MDD) but whose episode lasted only 12 days
(rather than the required two weeks) would be precluded from receiving a formal diagnosis
of MDD. Nevertheless, it is unlikely that this person differs fundamentally from most people
with MDD.
Inadequate treatment validity.—As also noted earlier, the presence of lists of ESTs
suggests that at least some DSM categories possess treatment validity and clinical utility
Author Manuscript
(Garb et al., 2009). Nevertheless, it clear that the linkage between diagnosis and treatment is
not nearly as tight in psychiatry as it is in other domains of medicine. Increasing data
suggest although at least some treatment specificity clearly exists (Lilienfeld, 2014a),
common factors – those that cut across most or all efficacious treatments - account for hefty
chunks of variance in therapeutic outcomes (Laska, Gurman, & Wampold, 2014). More
broadly, the DSM era has not borne witness to large-scale reductions in the morbidity or
mortality associated with major mental disorders; nor to decreases in suicide rates. These
sobering facts stand in stark contrast to sharp declines in the death rates associated with
heart disease, stroke, cancer, and a number of major medical disorders (Insel, 2009). Of
Author Manuscript
course, it would be unfair to attribute all of the lack of progress in diminishing psychiatric
morbidity and mortality to our contemporary system of classification. Nevertheless, this
state of affairs raises the distinct possibility that fresh approaches to classification will be
needed to achieve breakthroughs in intervention.
Genetic and environmental nonspecificity.—If DSM-ICD conditions were largely

distinct, as would be expected from a neo-Kraepelinian framework, one might anticipate that
their genetic and environmental architecture would be largely distinct. Yet the more we learn
about the etiology of most DSM conditions, the more apparent it becomes that many of the
influences impinging on them are nonspecific. For example, one recent genome-wide
association study revealed that many single nucleotide polymorphisms (SNPs) of small
effect contribute to risk for five mental disorders traditionally deemed to be etiologically
disparate, namely schizophrenia, bipolar disorder, major depression, attention-deficit
Author Manuscript
hyperactivity disorder (ADHD), and autism spectrum disorder. For example, the genetic
correlation of SNPs in schizophrenia and bipolar disorder was.68 (Cross-Disorder Group of
the Psychiatric Genetics Consortium, 2013). The genetic overlap in this study is especially
striking given that several of these disorders (e.g., ADHD and schizophrenia) display little
overt phenotypic resemblance.
Similarly, there is compelling evidence for substantial environmental nonspecificity in the

etiology of mental disorders. For example, even after controlling statistically for childhood
adversity, childhood sexual abuse has emerged as modest nonspecific antecedent, and
perhaps a causal risk factor, for a host of different mental disorders, including multiple
mood, anxiety, and substance use disorders (Molnar, Buka, & Kessler, 2001). Although not
undermining the possibility of some disorder specificity, these well-replicated findings pose
Author Manuscript
a challenge to the assumption that DSM-ICD conditions are associated with distinct specific
etiologies (see Meehl, 1977, for a discussion of different forms of specific etiology).
DSM-ICD anomalies: Taking stock

In aggregate, the anomalies we have reviewed reveal that all is not well with the scientific
health of the DSM. With each new DSM edition, concerted efforts have been directed to
attempting to salvage the essence of the neo-Kraepelinian framework by a variety of ad hoc
maneuvers (Popper, 1959), such as removing, adding, or revising diagnostic criteria;
changing thresholds or age of onsets for diagnoses; adding new diagnoses or removing
others; adding new disorder subtypes; and adding or removing hierarchical exclusion rules
(which prohibit certain conditions from being diagnosed if they appear to be attributable to
another condition). Yet repeated attempts to nibble around the edges of the DSM anomalies
Author Manuscript
across editions of the manual have met with at best limited success. In the words of Kendler
(2015), the DSM may have found itself in a “box canyon”: a deep gorge from which it
cannot extricate itself. If so, small-scale revisions to the DSM-ICD, or, in the lingo of
Lakatos (1978), modifications to the auxiliary hypotheses of the DSM-ICD research
program, are unlikely to remedy its core shortcomings. In reacting to the lack of substantial
scientific progress in the DSM-ICD research program, a growing chorus of critics has
argued that this approach has failed to carve nature at its joints, to borrow Plato’s hallowed
Author Manuscript
phrase (see Gangestad & Snyder, 1985). .
The view that the DSM-ICD model is defective beyond repair is by no means new. In 1989,
eminent British psychiatrist R.E. Kendell (see also Widiger & Trull, 2007) wrote that:
Ninety years have now elapsed since Kraepelin first provided the framework of a
plausible classification of mental disorders. Why then, with so many potential
validators available, have we made so little progress since that time? …One
important possibility is that the discrete clusters of psychiatric symptoms we are
trying to delineate do not actually exist (Kendell, 1989, p. 313).
Ironically, a similar alarm call was sounded 13 years later by the two principal architects of
DSM-5: “Research exclusively focused on refining DSM-defined syndromes may never be
successful in uncovering their underlying etiologies. For that to happen, an as yet unknown
Author Manuscript
paradigm shift may need to occur” (Kupfer et al., 2002, xix). RDoC may be heralding the
leading edge of this very paradigm shift.
RDoC: Intellectual Antecedents

Many of the intellectual precursors of RDoC can be traced to early successes in the
biomedical sciences, which unearthed the etiology of what we now recognize as physical
disorders masquerading as mental disorders (Schildkrout, 2014). At the same time, many of
the ensuring failures of this approach to traditional mental disorders, such as schizophrenia,
imparted valuable sobering lessons that were to inform RDoC.
Laboratory methods in medicine

Author Manuscript
It is easy to forget that by the middle or even the end of 19th century, medical diagnosis,
much like modern-day psychiatric diagnosis, was based almost exclusively on a combination
of careful history taking from patients and a physical examination, an approach known as
bedside medicine (Ackerknect, 1967). Medical tests began to be introduced in the 18th and
19th centuries, with the advent of the stethoscope, opthalmoscope, and laryngoscope
becoming routine in medical examinations by the 1850s (Kihlstrom, 2002). Nevertheless, it
was not until the 20th century that laboratory tests became de rigueur for diagnosing many
medical conditions (Burke, 2002). The examination of cerebrospinal fluid to detect
meningitis emerged around the turn of the 20th century, and the Wasserman Test for the
detection of syphilis was developed in 1906 (Berger, 1999). Today, over 3000 standardized
laboratory tests are used in standard medical practice (Kapur, Phillips, & Insel, 2012).
Author Manuscript
Early triumphs in psychopathology research

The laboratory revolution in medicine led to two sensational triumphs in the
psychopathology domain. Specifically, medical science demonstrated that two conditions
long presumed to be largely or entirely psychogenic, namely general paresis and pellagra,
were of purely organic etiology. General paresis, also called general paralysis of the insane,
accounted for 5 to 10 percent of hospital admissions around the turn of the century (Tsay,
2013). Its primary psychological features included delusions, grandiosity, euphoria,
concentration difficulties, and poor impulse control. In 1913, Noguchi and Moore isolated
Author Manuscript
the syphilis spirochete (Treponema pallidum) in the brains of patients with general paresis,
demonstrating that condition was an outcome of tertiary syphilis, in which the bacterium had
invaded the central nervous system. Individuals suffering from pellagra, whose symptoms
included irritability, depression, insomnia, paranoia, confusion, and memory problems, were
also common fixtures in inpatient psychiatric hospitals around the turn of the 20th century.
Research later showed this condition to be a consequence of niacin deficiency (Sartorius,
1993).
The discovery of the medical causes of general paresis and pellagra were justifiably hailed
as triumphs of biomedicine, especially because they promptly led to effective treatments
(penicillin and niacin, respectively). Coming as they did against the backdrop of the
discovery of mendelian genetics, they encouraged several generations of psychopathology
researchers to believe that other major mental disorders, such as schizophrenia, bipolar
Author Manuscript
disorder, and obsessive-compulsive disorder, would similarly be traceable to a single

etiological agent (Kendler & Schaffner, 2011). In many respects, however, these successes
were misleading; in the case of schizophrenia, dozens if not hundreds of candidates for the
“schizochete” (Neimark, 2008) – a presumably unique biological cause of the disorder, such
as dopamine hyperactivity, a dominant gene, or hypofrontality – were proposed over the
years. Yet this “single bullet” approach ultimately failed, almost surely because the causes of
these disorders turned out to be exceedingly multifactorial (Kendler, 2006). This approach
also foundered because these causes are not unique to schizophrenia, but are shared across
numerous DSM-ICD disorders. This belated realization was almost certainly one of many
catalysts for the transdiagnostic approach embraced by RDoC.
Biological markers
Author Manuscript
The appreciation of multifactorial causality notwithstanding, the 1970s and 1980s saw keen
interest in the identification of biological markers for psychopathology (Iacono, 1985).
Sophisticated thinkers viewed these markers not as akin to the spirochete that causes general
paresis, but rather as fallible but nonetheless useful indicators that could assist in etiological
and perhaps diagnostic efforts. For example, the early 1980s bore witness to the
development of the Dexamethasone Suppression Test (DST), an indicator of cortisol
reactivity, to assist in the diagnosis of endogenous depression (Carroll, 1982), the discovery
of reduced rapid eye movement (REM) latency as an indicator of major depression (Kupfer
et al., 1978), and the development of measures of smooth pursuit eye movement dysfunction
(Iacono, Tuason, & Johnson, 1981) as an indicator of schizophrenia.
Nevertheless, the much vaunted search for biological markers, which continues apace today,
Author Manuscript
has met with at best qualified success. Although some of these putative markers offered
fruitful leads to etiology, no marker displayed sufficiently high levels of both sensitivity and
sensitivity (as well as positive and negative predictive power) to qualify as proxies for
research diagnoses, let alone as screening or laboratory tests for routine clinical use (Insel,
2014; Kapur et al., 2012). For example, while the DST displayed reasonably high sensitivity
for endogenous depression, its specificity was often poor, reflecting high levels of false
positive identifications among patients with schizophrenia, alcoholism, anxiety disorders,
and dementia (Coppen et al., 1983). As a consequence, a state of disenchantment regarding

Author Manuscript
the DST and other biological markers gradually set in. As of today, the lone laboratory-
based biological tests in the DSM are for a subset of sleep disorders, such as assay-based
measures of hypocretin levels for narcolepsy (American Psychiatric Association, 2013), and
even these tests are hardly infallible.
Nevertheless, it has long been unclear whether the failure to detect highly sensitive and
specific markers of mental disorders was a reflection of the failure of the biological strategy
per se rather than of the gross imprecision of diagnostic phenotypes themselves. As the 20th
century drew to a close, more and more researchers were placing their bets on the latter.
The slow progress of psychopharmacology

Another impetus for RDoC has been the lack of progress in the psychopharmacology
industry over the past few decades. The 1950s, which were marked by “the explosive
Author Manuscript
growth” of psychopharmacology, witnessed the discovery of the first tricyclic

antidepressant, imipramine (Tofranil), the first monoamine oxidase inhibitor, iproniazid
(Marsilid), the first benzodiazepine, chlordiazepoxide (Librium), and the first antipsychotic
medication, chlorpromazine (Thorazine; see Lopez-Munoz & Alamo, 2009). The next two to
three decades were awash in the development of scores of variants of these medications.
Yet for reasons that are still poorly understood, progress in psychopharmacology soon began
to slow to a crawl. Although the advent of selective serotonin uptake inhibitors in the 1980s
and atypical antipsychotic medications in the 1990s substantially ameliorated the
problematic side effect profiles of major classes of psychiatric drugs, there is little evidence
that these new classes of medications contributed to marked enhancements in overall
therapeutic efficacy (Anderson, 2000; Chakos et al., 2014). As another example, recent
Author Manuscript
enthusiasm concerning the potential of glutamate antagonists for schizophrenia gave way to
disillusionment in the wake of several failed clinical trials (Carroll, 2012). In 2011, two
premier drug companies, AstraZeneca and GlaxoSmithKline, announced that they were
terminating their search for new psychiatric medications in light of the bleak outlook for
major discoveries (Cowen, 2012), and other firms may soon be following suit. Although
some of lack of progress may stem from failures of innovation and risk-taking in the
psychopharmacology industry, much of it may also stem from the fact that the disorder
phenotypes afforded by DSM-ICD are too crude and heterogeneous to allow for effective
medication targets.
Precision medicine
A final intellectual cross-current that informed RDoC was the emergence of precision
Author Manuscript
medicine, sometimes also called personalized medicine, a new branch of medicine that
strives to harness the power of clinic-pathological laboratory measures, such as genetic tests,
to tailor treatments to individuals (Mirnemazi, Nicholson, & Darzi, 2012). For example,
researchers have developed a targeted drug treatment for a small subset (4%) of patients with
cystic fibrosis who possess a specific mutation (Insel, 2014). Similarly, the treatment of
breast cancer has been revolutionized by the development of oncotype testing, permitting
physicians to move from a “one size fits all” intervention approach to treatment geared to
specific genetic profiles (Afimos, Barthelemy, & Awada, 2014). The coming era of precision
Author Manuscript
medicine has stirred hopes for a similar revolution in psychiatry and clinical psychology.
The Birth of RDoC: Fundamental Presuppositions and Structure

In many ways, RDoC has DSM to thank for its birth. During his tenure as NIMH director,
Steven Hyman expressed his concern at seeing preclinical researchers attempting to develop
animal models that would mimic DSM-based diagnostic criteria. As Hyman put it “The
DSM system…created an unintended epistemic prison that was palpably impeding scientific
progress…Even animal studies that purported to develop disease models…were often
judged by how closely they approximated DSM disorders.” (Hyman, 2010, p. 157). Under
Hyman’s successor, Thomas Insel, the NIMH sought to develop a new framework for
organizing research that could liberate clinical and translational researchers from
the“epistemic prison” of DSM criteria. The primary concern was not that the DSM criteria
Author Manuscript
lacked validity or utility for clinical practice; rather, the issue centered on whether the de
facto whole-sale appropriation of DSM criteria by the research community was impairing
the elucidation of pathophysiology. In this sense, the birth of the RDoC initiative can be seen
as inextricably linked to the NIMH’s growing emphasis on translational neuroscience to
guide research priorities.
So-named as an homage to the RDC, the RDoC was formally launched in 2009 as a bold
initiative to transform the current framework of psychiatric classification into an explicitly
biological system (Cuthbert, 2014; Insel et al., 2010; Sanislow et al., 2010). RDoC’s aim
was ambitious: to inaugurate nothing short of a paradigm shift in descriptive psychiatry.
Rather than base psychiatric diagnosis on presenting signs and symptoms, as do DSM and
ICD, RDoC strives to anchor psychiatric classification and diagnosis in a scientifically
Author Manuscript
supported model of neural circuitry. RDoC conceptualizes mental disorders as dysfunctions

in brain systems that bear important adaptive implications, such as systems linked to reward
responsiveness and threat sensitivity (see also Harkness, Reynolds, & Lilienfeld, 2013).
An RDoC Primer
In a widely discussed and controversial blog posting issued several weeks prior to the
release of DSM-5, NIMH Director Thomas Insel (2013) staked out a bold claim for RDoC:
“…NIMH will be re-orienting its research away from DSM categories. Going forward, we
will be supporting research projects that look across current categories – or sub-divide
current categories – to begin to develop a better system.” Nevertheless, in subsequent
communications, Insel made clear that NIMH would continue to fund grants that included
data on DSM categories, but that it would prioritize those that emphasized transdiagnostic
Author Manuscript
etiological processes.
At least for the foreseeable future, RDoC is not envisioned as a system of psychiatric
classification in its own right. Instead, in the near term, RDoC and DSM-ICD are expected
to coexist. Nevertheless, the RDoC is intended to provide scaffolding for a large-scale
research program that will ultimately yield an alternative to DSM-ICD (MacDonald &
Krueger, 2013). .
RDoC adopts no a priori stance on what form a new model of classification will eventually
Author Manuscript
take (Insel, 2014). In the words of the current NIMH director and his colleagues, RDoC is “a
vision for the future” (Insel et al., 2010, p. 749) rather than a fleshed out proposal. As of this
writing, RDoC is largely fluid in its mission. Such flexibility is probably justifiable in the
early phases of scientific investigation, in which hypothesis generation (“the context of
discovery”) should often take precedence over hypothesis testing (“the context of
justification”; Kell & Oliver, 2003).
Still, the RDoC research program is not entirely open-ended. RDoC provides researchers
with an explicit rubric for guiding investigations. As can be seen in Figure 1, RDoC
proposes that research efforts be conducted within a two dimensional matrix (Morris &
Cuthbert, 2012; Weinberger & Goldberg, 2013). This matrix took shape following a series of
workshop meetings held over an 18 month period that involved panels of experts in different
domains of neural circuitry. On the horizontal axis are seven units of analysis organized
Author Manuscript
roughly from more to less “basic”: genes, molecules, cells, circuits, physiology, behavior,
and self-reports (the matrix also includes a column for “paradigms,” allowing investigators
to indicate which tasks are useful for the research question at hand). On the vertical axis are
five broad domains/constructs that correspond to brain-based circuits deemed relevant to
psychopathology: negative valence systems (e.g., threat, loss), positive valence systems (e.g.,
approach motivation, responsiveness to reward); cognitive systems (e.g., attention, working
memory), social processes (e.g., theory of mind, dominance), and arousal/regulatory
systems.
RDoC rests on several assumptions, four of which are especially crucial (Cuthbert & Insel,
2013). First, RDoC is explicitly transdiagnostic in seeking markers of dysfunctional
psychobiological circuitry that transcend multiple traditional disorder categories; it posits
Author Manuscript
that many superficially different conditions are underpinned by similar psychobiological

processes. Second, RDoC is translational in emphasis, encouraging researchers to apply the
basic science of brain systems and behavior to an understanding of mental disorders. Third,
RDoC adopts a dimensional framework in light of evidence that the activity of most brain
circuits, such as reward and threat systems, is continuously distributed, with few or no clear-
cut boundaries demarcating normality from abnormality. Fourth, RDoC strives to accord
roughly equal weight to different levels of analysis, including the biological and behavioral
(Cuthbert & Insel, 2013).
Consistent with the view of science as a self-correcting enterprise, RDoC is conceptualized

as provisional and open to revision in light of new scientific data. As a consequence, novel
brain-based constructs may be added to the matrix with the emergence of new neuroscience
Author Manuscript
evidence.
The Scientific Promise of RDoC

As of this writing, there is considerable excitement in many quarters regarding RDoC’s
scientific potential, and for good reason. RDoC has already begun to loosen the longstanding
grip of the DSM-ICD system over research and grant funding. This hegemony has stifled
investigators’ capacity to explore scientifically promising alternatives to the status quo
model of psychiatric classification (see also Berenbaum, 2013; Harkness & Lilienfeld, 2013,
and Markon, 2013). Moreover, RDoC’s adoption of a dimensional approach to

Author Manuscript
psychopathology as a starting assumption accords with the bulk of the evidence derived
from taxometric studies, which suggests that most conditions are underpinned by
dimensions rather than by taxa, that is, natural categories (Haslam et al., 2012). In this
respect, RDoC is more consistent with data on the latent structure of mental disorders than is
the DSM. At the same time, RDoC allows for the possibility of threshold effects (“tipping
points”) or other categorical effects (Cuthbert & Insel, 2013), whereby certain
psychopathological phenomena differ qualitatively rather than quantitatively from normality.
Furthermore, RDoC promises to capitalize on the burgeoning corpus of knowledge

concerning affective and cognitive neuroscience by applying it to the classification of mental
disorders. In this respect, it may ultimately allow for closer linkages between assessment and
diagnosis, on the one hand, and treatment and prevention, on the other. More ambitiously,
RDoC may eventually shift psychiatry and clinical psychology closer to the goal of
Author Manuscript
precision medicine (Insel, 2014), or at least the more modest goal of stratified medicine, in
which interventions are tailored to individuals within certain well-defined subgroups defined
by laboratory-based profiles (Kapur et al., 2012). With the enormous recent influx of
research attention and grant funding to RDoC, there is renewed hope of uncovering
laboratory-based assays with high levels of sensitivity and specificity for the proximal
pathological processes that contribute to psychopathological signs and symptoms.
In all of these respects, RDoC appears to be a valuable alternative to the current approach to
classification, especially given that the DSM-ICD’s scientific yield, which has been
substantial despite its noteworthy defects (Regier, Narrow, Kuhl, & Kupfer, 2009), appears
to be reaching an asymptote. We regard RDoC as holding out the promise of generating a
competing, and ideally more valid, system of classification relative to that of DSM-ICD. At
Author Manuscript
the same time, RDoC confronts a number of challenges, many of which have received short
shrift in ongoing discussions (but see Berenbaum, 2012; Lilienfeld, 2014b; Shankman &
Gorka, 2015).
In the following section, we examine three overarching sets of challenges to RDoC: (1)
conceptual, (2) methodological, and (3) practical/logistical (see also Lilienfeld, 2014b).
Although there is overlap among these three sets of challenges, we differentiate them for
expository purposes. Given that we intend to be constructive, we devote considerably more
space to RDoC’s potential weaknesses than to its potential strengths. Nevertheless, this
differential space allotment should not be construed as implying that we view RDoC’s
weaknesses as outweighing its strengths. To the contrary, we focus on these weaknesses in
the hope that researchers and theorists, including RDoC’s architects, will pay them greater
Author Manuscript
heed and thereby increase the likelihood that RDoC will bear scientific fruit. Because we do
not perceive any of these challenges as insurmountable, we hope that our delineation of them
plays a role in shaping the future direction of RDoC.
RDoC: Conceptual Challenges

Author Manuscript
Falsifiability
The philosopher of science Sir Karl Popper (1959) argued that for a theoretical model to be
scientific, it must be capable in principle of being proven wrong. Although few
contemporary philosophers of science accept Popper’s premise that falsifiability is the lone
demarcation criterion separating science from nonscience (Pigliucci & Boudry, 2013), many
concur that this feature is one key indicator of a model’s scientific status. Moreover, Popper
maintained that science ideally progresses by means of “severe” or “risky” tests (see also
Meehl, 1978), those that place theories at grave theoretical risk.
In the case of RDoC, it is not evident what findings would suggest that it is not mapping
well onto the state of nature, or at least falling considerably short of its avowed goals. Nor is
it entirely clear what would constitute “negative” findings – results that would indicate that
Author Manuscript
RDoC is a flawed endeavor that is not leading to a feasible alternative model of

classification. In our view, it is incumbent on RDoC’s developers to lay out a set of
provisional but explicit benchmarks by which its progress, or lack of progress, can be
gauged.
In a related vein, it will be critical to address the question of whether RDoC, especially once
it is in better-developed form, is superior to the DSM-ICD model in predicting theoretically
and pragmatically relevant external criteria, such as natural history and treatment response.
Based on our conversations with current and would-be RDoC grantees, it is our impression
that researchers who are submitting proposals within the RDoC framework are being
discouraged from pitting RDoC-relevant markers against DSM diagnoses in their capacity to
predict such criteria. Such advice would be ill-conceived, as it could hamper long-term
Author Manuscript
efforts to ascertain whether RDoC is performing as well as, or ideally better than, the extant
system.
A further issue that bears on the falsifiability of RDoC concerns the distinction between
convergent and discriminant validity (Campbell & Fiske, 1959; Cole, 1987). As of this
writing, the lion’s share of RDoC research focuses on convergent validity, viz.,
demonstrating that presumed markers are predictive of their hypothesized RDoC domains
(Shankman & Gorka, 2015). For example, it is useful to demonstrate that excessive levels of
fear-potentiated startle are related to intrusive memories and fear-related memories
associated with a traumatic event (Norholm et al., 2014). This finding suggests that fear-
potentiated startle is a valid indicator of the RDoC Negative Valence domain, especially its
acute threat (fear) subconstruct. Nevertheless, moving forward, it will be at least equally
Author Manuscript
crucial to demonstrate that putative markers are not associated, or at least less associated
with, non-hypothesized RDoC domains. For example, it will be important to show that fear-
potentiated startle is largely unassociated with markers of the RDoC Arousal/modulatory
domain, which should theoretically be largely independent of fear, although it may be
predictive of levels of baseline startle.
Potential overemphasis on biological “dysfunction” at the level of an individual

Author Manuscript
RDoC’s posits that mental disorders are “disorders of brain circuits” (Insel et al., 2013, p.
749). As a corollary, RDoC presumes that “the tools of clinical neuroscience…can be used
to identify dysfunction in neural circuits” (Morris & Cuthbert, 2012, p. 33). At some level,
this assumption is merely a truism given that all psychological disorders, and all
psychological phenomena for that matter, must be mediated by the brain and the remainder
of the central nervous system (Kendler, 2005). Still, researchers who adopt the RDoC
framework must be careful not to confuse biological mediation with biological etiology.
In the phraseology of Graham (2013), all mental disorders are in the brain, not all are
necessarily of the brain. That is, at least some conditions classically regarded as mental
disorders may stem from neural systems reacting more or less normally to harsh or extreme
environmental input. Such reactions may be acute–such as rapid, over-generalized fear
Author Manuscript
responses to trauma–or gradual, such as the slow accumulation of glucocorticoid resistance

and chronic inflammation following prolonged periods of life stress that may result in
structure damage to key brain regions involved in regulation of mood (Macqueen & Frodl,
2010; Treadway et al., 2015). Similar phenomena are observed in other areas of medicine;
hypothermia reflects the activity of the body reacting normally to extreme environmental
input, and some forms of cardiovascular disease may be mediated in part by normal
biological responses to a chronically unhealthy lifestyle. In other cases, biological mediation
may be more akin to a lesion of some kind, in which there is a genuine loss of function due
to an environmental insult, genetic predisposition, or their interaction. Many clinical
neuroscience studies neglect to specify these different classes of hypothesized mechanisms
clearly, sufficing to label observed differences between patient and healthy groups simply as
“deficits”, “dysfunction” or “alterations.” A limitation of the RDoC project as it is currently
implemented is that it makes few demands for better delineation of mechanisms at this level.
Author Manuscript
Although RDoC does not conflate biological mediation with etiology, it may inadvertently
foster this error by placing considerably less emphasis on psychosocial than on biological
variables (Hershenberg & Goldfried, 2015; Lilienfeld, 2014b). The RDoC matrix focuses
almost exclusively on intra-individual variables, with little or no explicit coverage of extra-
individual variables, such as the social or cultural context (Berenbaum, 2013; Shankman &
Gorka, 2015; Whooley & Horwitz, 2014). This omission is significant given that the
phenotypic expression of biological vulnerabilities may often be constrained by
sociocultural factors. For example, religious beliefs, as well as regional differences in the
pricing and availability of alcohol, are associated with – and probably causally linked to -
risk for alcohol use disorder (Kendler, 2012). Hence, even individuals with a potent genetic
propensity toward alcohol use disorder may display low rates of this condition if raised in a
Author Manuscript
socially traditional environment.
Furthermore, five of the seven RDoC units of analysis focus explicitly on biological
indicators, raising concerns that biological levels of analysis may receive undue attention by
investigators (Berenbuam, 2013). Although several RDoC publications (e.g., Morris &
Cuthbert, 2012) have acknowledged the importance of psychosocial variables and
developmental considerations in the RDoC program, these processes are not explicitly
represented in the matrix, an omission that RDoC may wish to remedy.
Fueling this concern is the fact that some researchers appear to have assumed erroneously
Author Manuscript
that indicators drawn from one level of analysis (e.g., physiological, behavioral) are
necessarily best suited for detecting abnormalities at that level. For example, in response to
Berenbaum’s (2013) concern that RDoC underemphasizes the role of beliefs, emotions, and
other potential “emergent” phenomena that are not readily reducible strictly to neural events
(Franklin et al., 2014, O’Connor, 1994; Paris, 2013), Cuthbert and Kosak (2013) wrote that
“Berenbaum is right in supposing that research that relies exclusively on self-report data
would fall outside of the RDoC approach” (p. 933).
The reasons for this a priori exclusion are unclear. Such decisions should be adjudicated by
data, not by fiat. There is no inherent reason why self-report measures, which can readily
capitalize on aggregation across indicators of behavior, cognition, and emotion across
diverse situations, cannot provide equally – or more – construct-valid measures of biological
systems relative to biological markers of these systems. For example, Patrick et al. (2013)
Author Manuscript
reported that scores on self-report measures of an externalizing propensity and aggression

were better markers of a latent disinhibition dimension than were P300 responses, an event-
related potential indicator of response to novelty. Hence, RDoC must remain open to the
possibility that self-report measures alone may in some cases be the optimal indicators of
relevant psychobiological systems. Put somewhat differently, RDoC should be guided
exclusively by data, and should adopt no a priori stance on which measures or sets of
measures will prove most valid for detecting individual differences in the behavioral
manifestations of neural circuitry.
Biological predispositions versus their behavioral manifestations

A potentially crucial distinction that has received little attention in the RDoC literature is
that between biological predispositions to psychopathology and their behavioral
Author Manuscript
manifestations (Lilienfeld, 2014b). In this respect, the distinction between basic tendencies
and characteristic adaptations in the personality literature provides a useful organizing
framework (McCrae & Costa.1995; see also Harkness & Lilienfeld, 1997). Basic tendencies
are personality traits, such as negative emotionality (NE), whereas characteristic adaptations
are the behavioral expressions of these traits, such as an anxiety disorder. Wakefield’s (1992)
influential harmful dysfunction framework, which posits that the definition of mental
disorder is a conjunction of (a) a failure in, or breakdown of, a naturally selected
psychological system (dysfunction) and (b) impairment (harm) is also broadly consistent
with this distinction (but see Lilienfeld & Marino, 1998, for a critique of aspects of
Wakefield’s framework). This model proposes that the presence of biological dysfunction
alone is not sufficient for psychopathology; this dysfunction must also be manifested in
social harm.
Author Manuscript
The distinction between basic tendencies and characteristic adaptations underscores the
point that certain adaptations to personality traits may be largely adaptive, whereas others
may be largely maladaptive. For instance, an individual with high levels of NE may manifest
this predisposition in an anxiety disorder; alternatively, she may manifest it in artistic
productivity, which is associated with high levels of NE (Sheldon, 1994). As another
example, the mean sensation seeking scores of prisoners are essentially indistinguishable
from those of firefighters (Zuckerman, 1994), suggesting that sensation seeking can be
Author Manuscript
manifested in either (a) socially and personally destructive outlets (e.g., crime, substance
abuse) or (b) socially and personally constructive outlets (e.g., firefighting, law enforcement)
depending on psychosocial factors and
Data on discordant monozygotic (MZ) twins among individuals with r mental disorders
afford another potential illustration of multifinality. For example, between 35% and 59% of
the co-twins (Carndo & Gottesman, 2000) of MZ twins with schizophrenia do not develop
the disorder. Some of this discordance may derive from epigenetic differences within MZ
pairs, in turn perhaps stemming partly from differential twin exposure to environmental
influences (Dempster et al., 2011; Kato et al., 2005; Petronis et al., 2003). More broadly, the
literature on MZ twin discordance in mental disorders raises the possibility that individuals
with similar biological liabilities, such as those reflecting a predisposition to schizophrenia,
display this risk in markedly different exophenotypes, only some of which are associated
Author Manuscript
with psychopathology.
The principles of (a) basic tendencies versus characteristic adaptations and (b) multifinality
highlight the point that individuals with similar biological predispositions toward
psychopathology may manifest these predispositions in different ways, in part as a
consequence of developmental and psychosocial factors. If so, RDoC may be insufficient as
a model for mental disorder, as it may often be unable to distinguish physiological risk
factors for psychopathology from psychopathology per se (see also Wakefield, 2014). If so,
RDoC, at least in its present form, may be better suited as a model of the predispositions
toward mental illness than of mental illness itself.
RDoC: Methodological Challenges

Author Manuscript
Neglect of Measurement Error

Consistent with the RDoC’s mission, several authors have argued that psychiatric diagnosis
should transition to a laboratory-based approach (Kihlstrom, 2002; see Widiger & Clark,
2000, for a review) and thereby bring psychiatry in line with the rest of medicine (Nemeroff,
Kilts, & Berns, 1999). Nevertheless, Kilhstrom’s (2002) vision, echoed by RDoC, may be
considerably more challenging to achieve than is commonly appreciated.
Laboratory measures are typically associated with largely unappreciated psychometric

weaknesses. As Epstein (1980) noted, psychologists have long granted such measures an
undeserved scientific cachet, often “giving them a pass” with respect to psychometric
requirements (see also Tryon, 1973). Laboratory measures of psychological constructs
frequently display low levels of temporal and cross-situational consistency, largely because
Author Manuscript
they contain substantial elements of “situational uniqueness.” Performance on such

measures can be affected by a plethora of contextual and situational factors, including the
mood and alertness of the participant, time of day, experimental instructions, nature of the
laboratory setting, perceived attitude of the experimenter, demand characteristics, and
idiosyncrasies of the laboratory paradigm itself (Kendler & Neale, 2010).
With respect to the lattermost source of variance, Mischel (1968) observed that even
Author Manuscript
seemingly trivial changes in experimental paradigms can lead to dramatic changes in

measures’ intercorrelations and correlations with other measures. Ironically, Kidd, Palmeri,
and Aslin (2013) demonstrated this point using the very paradigm that Mischel (1974) made
famous: the marshmallow test of delay of gratification in children. They showed that a
seemingly trivial alteration in the experimental set-up – namely, whether an adult who had
promised to bring a set of attractive art supplies to the child immediately prior to the task
actually did so – resulted in massive changes in outcomes. Specifically, children exposed to
the “reliable” adult waited an average of 12 minutes, whereas children exposed to the
“unreliable” adult waited an average of only 2 minutes. Kidd and colleagues interpreted this
result as suggesting that children who encounter unpredictable environments are less likely
than other children to delay gratification, as they have ample reason to doubt whether
expected rewards will materialize.
Author Manuscript
Block (1977; see also Tellegen, 1991) similarly noted that T data (test data), that is, data
derived from isolated laboratory indicators, are more unreliable and erratic in their relations
with (a) each other and (b) other measures compared with S data (self-report data) and R
data (rating data). Although S and R data possess their own psychometric limitations, these
data are typically aggregated across multiple diverse situations. Such aggregation is often
accomplished by summing items within scales, or in more sophisticated analyses, by
creating latent variables using such techniques as structural equation modeling. Such
aggregation minimizes situational error and yields more reliable and construct-valid
composites of behavior across situations (Epstein, 1980; Rushton, Brainerd, & Pressley,
1983). In contrast, T data are rarely aggregated across situations, at best being combined
only across multiple trials of the same measure.
Author Manuscript
These problems may be especially acute in the domain of neuroimaging. Few investigators
have examined the test-retest reliability of measures of functional magnetic resonance
imaging (fMRI; Bennett & Miller, 2010), even though this form of reliability is a basic
expectation of measures in other psychological domains. In an analysis of 63 studies,
Bennett and Miller (2010) found that the test-retest reliability of fMRI measures was
typically modest, with intraclass correlations (ICCs) averaging 0.50 (see also Vul, Harris,
Wienkielman, & Pashler, 2009). Furthermore, only 29% of activated voxels that were
statistically significant in one study were significant in a second study. Although test-retest
reliabilities for fMRI data were higher with briefer intervals, even back-to-back scans (taken
within one hour) exhibited an average cluster overlap of only 33%.
At the same time, because some of the neural processes examined in these studies may been
Author Manuscript
influenced substantially by state variables, these modest test-retest values may reflect
inherent short-term fluctuations in the neural processes themselves rather than measurement
error. Scanning is can be an unfamiliar experience for many participants, and initial anxiety
or discomfort in the scanner may change over repeated exposures. Moreover, repeated
exposures to cognitive tests will invariably lead to learning, even if only at the level of
greater familiarity and fluency. It would therefore be surprising if these changes were not
reflected in differences in neural responses. One recent study used a relatively large sample
to examine how activity in the ventral striatum (VS) during the feedback phase of a
rewarded instrumental conditioning task changed over time as compared with activity during
Author Manuscript
the anticipation phase (Chase et al 2015). As would be predicted by temporal-discounting

models of reinforcement learning (Schultz et al 1997), VS responses to better-than-expected
outcomes during the first scanning session were robust, but were nearly absent during the
second session. Conversely, ventral striatal responses to anticipation were absent during the
first session, but significant during the second. Moreover the magnitude of feedback related
VS activity at session predicted the magnitude of anticipatory VS activity at session 2. This
outcome-to-cue transfer in VS activity patterns suggests a change due to learning, but also
contributes to very low interclass-correlations when simply comparing the same condition
across testing sessions.
These effects may also vary across different types of tasks and conditions. For example,
Sauder, Hajcak, Angstadt, and Phan (2013) reported that the reliability of fMRI measures of
amygdala activation was adequate in response to fearful faces (ICCs ranged from 0.32 to
Author Manuscript
0.43), but inadequate in response to angry faces (ICCs ranged from −0.24 to 0.11). In
contrast, structural MRI measures appear to have considerably higher test-retest reliability
(Kendler & Neale, 2010). For example, the stability of measures of cortical thickness as
assessed by structural MRI is on the order of r=0.95 (Wonderlick et al., 2009).
Even the fMRI research center at which the study is conducted accounts for approximately 8
percent of the variance in fMRI blood oxygen-level dependent (BOLD) signal results,
suggesting that the laboratory itself is a potential source of error in analyses (Costafreda et
al., 2007). Another investigation revealed that the median ICC of fMRI findings across
different imaging centers that contained identical hardware set-ups was only 0.22 (Friedman
et al., 2008; see also Bennett & Miller, 2010).
Author Manuscript
All of these psychometric limitations may impede the replicability of functional imaging
findings unless addressed analytically. Adding to these concerns are findings that the
average statistical power of functional brain imaging studies is only about .20, which is
considerably lower than in most other domains of psychological and psychiatric research
(Button et al., 2013). Low power not only increases the chances of Type II errors (false
negative results), but also boosts the likelihood of overestimating the effect sizes of
statistically significant findings, a phenomenon known as the winner’s curse (Munafo et al.,
in press). Replicability concerns are not limited to functional brain imaging studies. Using a
fresh sample of college students as participants, Boekel, Wagenmakers, Belay, Verhagen,
Brown, and Forstmann (2014) attempted to replicate 17 published findings on the relation
between structural brain imaging findings and psychological findings derived from self-
reported data and other measures (e.g., a previously reported positive correlation between
Author Manuscript
amygdala gray matter and number of Facebook friends). Using Bayesian methods, they
found minimal support for any of the previous 17 findings.
The Limitations of Endophenotypes

Endophenotypes comprise a core intellectual progenitor for the RDoC’s dimensional
conceptualization of brain circuit dysfunction, and many RDoC studies have borrowed
heavily from endophenotype designs. Because they are presumed to be more proximal to
genes compared with exophenotypes, endophenotypes should ideally provide purer and
more construct-valid indicators of genetically influenced psychobiological systems.

Author Manuscript
Moreover, because many endophenotypes cut across traditional disorder categories,

endophenotypes accord well with RDoC’s transdiagnostic approach and emphasis on
identifying etiological processes that contribute to multiple DSM conditions (Miller &
Rockstroh, 2013). Nevertheless, RDoC is broader than the endophenotype approach, as it
does not mandate that candidate biological markers be heritable.
Despite high initial expectations and provisional successes for certain conditions, such as
schizophrenia (Cannon & Keller, 2006; Miller & Rockstroh, 2013), the endophenotypes
identified to date have not necessarily proven to be more genetically informative than
traditional exophenotypes, such as DSM criteria. Flint and Munafo (2007) examined this
issue in meta-analyses of studies of catechol O-methyl transferase (COMT), an enzyme that
metabolizes dopamine (among other neurotransmitters), and schizophrenia, a disorder
associated with dopamine overactivity in mesolimbic brain regions. They tested whether the
Author Manuscript
COMT genotype displayed higher effect sizes with candidate neuropsychological and
psychophysiological endophenotypes of schizophrenia, such as performance on the
Wisconsin Card Sorting Task, the N-Back Task, and P300 amplitude and latency, than with
DSM schizophrenia itself. Flint and Munafo found no evidence that the ostensible
endophenotypes were more highly related to the COMT genotype than was schizophrenia.
Their findings suggest that investigators should not assume that candidate endophenotypes
will necessarily yield higher effect sizes than do exophenotypes in genetic studies (but see
Tan, Callicott, & Weinberger, 2008, for a more sanguine view of the status of
endophenotypes for schizophrenia).
In contrast, Jonas and Markon (2014) reported more encouraging results. They examined the
relation between (a) three polymorphisms in the dopamine and serotonin systems that are
Author Manuscript
potentially relevant to impulsivity and (b) diagnostic (e.g., measures of attention-deficit

hyperactivity disorder), trait (e.g., self-report measures of impulsivity), neuropsychological
(e.g., continuous performance tasks), and neurobiological (e.g., functional imaging indices
of right inferior prefrontal activity) phenotypes relevant to impulsivity. Neurobiological
measures were the most highly associated with the genetic markers, with trait and
neuropsychological measures roughly tied for a distant second, and diagnostic measures last.
These analyses suggest that neurobiological measures, such as functional brain imaging
indices, may be promising endophenotypes of impulsivity. Still, as the authors observed,
neurobiological studies were characterized by considerably lower statistical power than
studies from other domains, raising the possibility that the effect sizes of the former studies
were inflated.
Author Manuscript
As noted earlier, one of two models of endophenotypes posits that they are intermediate
phenotypes, operating as mediators between genes and exophenotypes. Nevertheless, the
evidence that candidate endophenotypes mediate the relation between genes and behavioral
phenotypes is slim. In in a twin sample, Kendler, Neale, Kessler, Heath, and Eaves (1993)
found that although neuroticism was associated with elevated rates of major depression, it
did not mediate the association between genetic risk and major depression (see also Kendler
& Neale, 2010). Waldman (2005) reported inconsistent findings concerning whether scores
on the Trail Making Test mediate relations between dopamine genes and attention-deficit
hyperactivity disorder (ADHD), with partial mediation for Trails A but no significant
Author Manuscript
mediation for Trails B.
Such mediation should not be presumed, as certain putative endophenotypes may lie
causally downstream of the exophenotypes with which they are associated (Kendler &
Neale, 2010). For example, P300 amplitude appears to be a valid endophenotype for a broad
predisposition toward externalizing behavior and disinhibition (Patrick et al., 2009).
Nevertheless, P300 amplitude is exquisitely sensitive to attention (Polich, 2012). Hence, this
marker may be a consequence, not an antecedent, of the attentional and motivational deficits
associated with externalizing disorders, such antisocial personality and substance use
disorders, which display high levels of co-occurrence and covariation with ADHD
(Lilienfeld & Waldman, 1990; Torgersen, Gjervan, & Rasmussen, 2006).
One crucial assumption guiding the search for endophenotypes is that the biological
Author Manuscript
indicators in question are trait rather than state markers, as a measure of a biological
vulnerability would be expected to be stable over time. Again, the trait status of such
markers must be demonstrated empirically rather than assumed (Stoyanov & Kandilorova,
2014). For example, motion discrimination appears to be impaired both in patients with
schizophrenia and in their nonaffected relatives, suggesting that it may be a useful trait
marker of the disorder. In contrast, motion integration appears to be impaired in patients
with schizophrenia but not in their nonffected relatives, suggesting that is influenced by state
variables (Chen, Bidwell, & Norton, 2006). One major challenge for RDoC researchers will
be to demonstrate that endophenotypes, as well as other ostensible vulnerability markers of
RDoC domains are present even between disorder episodes or exacerbations.
RDoC: Logistical/Pragmatic Challenges

Author Manuscript
The dangers of premature reification of the RDoC matrix

As noted earlier, the RDoC matrix is intended to be provisional. It comprises cells directing
investigators to well-supported neural circuits linked to psychopathology while leaving open
the possibility of modifying extant cells or adding new ones on the basis of new evidence
(Cuthbert & Insel, 2013). In this regard, this matrix appears to represent a reasonable
compromise between excessive open-endedness and excessive prescriptiveness.
At the same time, there is a danger of premature reification of the matrix and of the RDoC
endeavor more broadly; it would indeed be unfortunate if the march to freedom from the
DSM’s “epistemic prison” (Hyman, 2010, p. 157) led merely to a padded cell in the matrix
penitentiary. A number of scholars have cautioned against such reification in the frequent
(mis)interpretation of DSM categories as “settled truths” (Kendler, 2014), and RDoC must
Author Manuscript
be careful to avoid the same error. In fairness, the RDoC framers have repeatedly described
their intentions for the matrix to be a “working document”. Good intentions, however, may
afford inadequate protection against ossification, which can easily insinuate itself through
the mundane, bureaucratic activities required by any large-scale administrative effort. For
instance, several recent requests for applications (RFAs) from NIMH have noted that
“applications must focus on at least one of the constructs that have been defined in these
RDoC workshops” (NIMH RFA-MH-14-050; see Shankman & Gorka, 2015). Such wording
appears to vitiate RDoC’s avowed goal of being self-correcting, as it will be difficult or

Author Manuscript
impossible to ascertain whether to add novel cells to the matrix unless these cells are
explicitly investigated. Social cognition research suggests that once mental categories are set
in place, they can quickly become implacable (McCrae & Bodenhausen, 2000). To avoid the
error of reification, RDoC should encourage researchers to examine psychobiological
constructs that have the potential to bridge, transcend or challenge current matrix
boundaries, as well as articulate an explicit process of matrix evaluation and revision.
Additionally, future RFAs should offer as much consideration to proposals that seek to
explicitly challenge matrix boundaries as they do to proposals that operate within them.
Although the RDoC matrix was thoughtfully constructed to be broad and flexible, a number
of fruitful hypotheses that may be challenging to capture within its framework. For example,
a small body of work has suggested that emotional numbing and alexithymia may be core
facets of reduced approach related-behaviors, but only in the context of response to a
Author Manuscript
traumatic experience, as in PTSD (Kashdan et al 2006). It is not entirely clear, however, how
this type of etiology-by-symptom interaction might be examined within the RDoC
framework.
Implications for grant funding in non-RDoC domains

The RDoC initiative developed out of an explicit desire to support translational neuroscience
in mental health research. It is therefore perhaps unsurprising that the RDoC has hastened
the already significant shift in NIMH extramural funding priorities away from psycho-social
research, including research on psychotherapy. Although we welcome the emphasis on
translational neuroscience, there is a legitimate concern that this approach may burn the
candle too far at both ends. On the one hand, as a consequence of RDoC basic neuroscience
efforts to understand circuit function more generally may be at risk for being underfunded,
Author Manuscript
raising the concern of insufficient knowledge available for translation. On the other hand, the
promise of neuroscience-guided treatment and diagnosis is widely acknowledged to be on
the distant horizon (e.g., Frances & Widiger, 2012; Paulus, 2015), whereas research devoted
to improvements to empirically-supported treatments for psychopathology could alleviate
the suffering of thousands in the near-term. It would therefore be beneficial for RDoC to co-
exist with research programs focused on (a) developing basic science knowledge without
necessarily requiring immediate demonstration of clear translatability and (b) clinical
outcome and process research that can yield important short-term breakthroughs in
treatment.
Translating RDoC findings into a feasible classification system

One crucial challenge for RDoC will be to delineate how the large and diverse corpus of
Author Manuscript
findings emanating from its investigations will be meaningfully synthesized into the
rudiments of an alternative model of psychiatric classification. The RDoC matrix is a
heuristic blueprint to guide research, but it is not a classification system in and of itself, nor
is it intended to be. A key unanswered question, then, is how RDoC findings will be
“translated” into a usable system that can ultimately guide clinical practice. Although it may
too early to address this question, it will soon be incumbent on RDoC architects to sketch
out an overarching plan for how RDoC research will be used to inform classification and
diagnosis. In the absence of such a rubric, it may be difficult to convert the enormous wealth
Author Manuscript
of data yielded by RDoC investigations into a concrete model that can supplement, or
perhaps eventually substitute for, the DSM-ICD.
The DSM-ICD and RDoC: Quo Vadis?

The DSMs, especially DSM-III and its progeny, were noteworthy achievements in
psychiatric classification, and they brought considerable order and diagnostic reliability to
previously disorganized territory. In certain respects, the DSM-ICD model has served us
well (cf., Greenberg, 2013), as it has greatly facilitated epidemiological and etiological
research and contributed to the development of efficacious interventions that have alleviated
the suffering of countless individuals. Moreover, by mapping onto signs and symptoms, the
DSM does a serviceable and face-valid job of capturing the essence of individuals’ distress
and impairment. It is difficult to imagine a comprehensive system of psychiatric
Author Manuscript
classification bereft of any reference to the subjective and behavioral manifestations of

people’s psychological suffering.
At the same time, it has become increasingly evident that many central features of the DSM-
ICD model do not map adequately onto the state of nature (Sanislow et al., 2010). The high
levels of covariation among putatively distinct categories, the large number of intermediate
cases, and the substantial phenotypic and etiological heterogeneity of numerous diagnostic
categories, among other vexing anomalies, suggest that something is deeply awry with at
least some core presuppositions underpinning the neo-Kraepelian model of psychiatric
classification. Moreover, these problems have proven stubbornly resistant to repeated efforts
at amelioration across multiple DSM and ICD revisions. The shortcomings of the DSM-ICD
edifice therefore appear to reflect an inherent deficiency in its architectural floorplan that
Author Manuscript
cannot be fixed merely by adjusting some of its walls.
The RDoC initiative emerged in the new millennium to address these mounting anomalies
(Insel, 2014), and it appears to be a valuable effort to ground psychopathology in well-
supported biological systems that carry important implications for adaptation and
maladaptation (Lilienfeld, 2014b). In this respect, RDoC has the potential to map more
closely than the DSM-ICD onto psychobiological reality. Moreover, even its critics
acknowledge that RDoC has already exerted one important salutary effect: loosening the
stranglehold of the DSM over research and grant funding. Such hegemony has impeded the
investigation of fruitful alternatives to classification and etiology, and we experience few
qualms in bidding it a greatly belated adieu
In this respect, RDoC should take heed of the lessons of the past. One historical feature
Author Manuscript
shared by DSM-ICD and RDoC is that neither was intended to become a fixed system in
terms of guiding research priorities. Yet, the DSM system eventually acquired such
sovereignty over psychiatric and psychological research that it became difficult for
investigators to examine clinical problems that fell outside of traditional disorder boundaries
(Hyman, 2010). RDoC must make concerted efforts to avoid the same error of reification.
As we have also noted, RDoC confronts a number of conceptual, methodological, and

Author Manuscript
logistical/pragmatic challenges, none of which appear to be insuperable (Lilienfeld, 2014b).

At the same time, many of the challenges appear to have received insufficient discussion in
RDoC documents, and will require careful consideration for RDoC to realize its
considerable scientific potential. In Table 1, we offer a baker’s dozen of recommendations
for addressing these challenges in the coming years of RDoC research.
One of the foremost challenges to RDoC’s effort to offer a viable alternative to DSM-ICD is
the fact that psychobiological predispositions can be expressed in a host of diverse
behavioral patterns, some of them largely healthy and others largely unhealthy (Harkness &
Lilienfeld, 1997). In this regard, it may be unrealistic to expect RDoC to fully supplant a
classification system, such as the DSM, that provides a remarkably rich compilation of the
variegated ways in which human behavior, both subjective and observable, can break down.
Signs and symptoms, their inevitable shortcomings notwithstanding, will always be
Author Manuscript
necessary for a complete characterization of mental disorder.
In our view, the DSM-ICD will never provide a sufficient foundation for a comprehensive
classification system, because psychiatric signs and symptoms, like fever, are inevitably
nonspecific indicators of a host of psychobiological dysfunctions. Conversely, RDoC will
similarly never be sufficient for a comprehensive classification system, because
psychobiological dysfunctions can be manifested in a host of markedly diverse signs and
symptoms as a function of innumerable moderating variables. As a consequence, a full
characterization of psychopathology will require the DSM-ICD’s remarkably astute
descriptive observations, informed by the best available basic research on neural circuitry
relevant to psychopathology. If it attends carefully to the constructive challenges posed here,
RDoC and the fruits of its labor hold the potential to complete the story that Emil Kraepelin,
Author Manuscript
the St. Louis group, and Robert Spitzer started.
References
Ackerknecht, EH. Medicine at the Paris Hospital. Johns Hopkins University Press; Baltimore, Md:
1967. p. 1794-1848.
Ackerman PL. Personality, self-concept, interests, and intelligence: Which construct doesn’t fit?
Journal of Personality. 1997; 65:171–205.
Aftimos PG, Barthelemy P, Awada A. Molecular biology in medical oncology: diagnosis, prognosis,
and precision medicine. Discovery Medicine. 2014; 17:81–91. [PubMed: 24534471]
American Psychiatric Association. Diagnostic and statistical manual of mental disorders. third. Author;
Washington, D.C.: 1980. (DSM-III)
American Psychiatric Association. Diagnostic and statistical manual of mental disorders. fifth. Author;
Washington, D.C.: 2013. (DSM-5)
Author Manuscript
Anderson IM. Selective serotonin reuptake inhibitors versus tricyclic antidepressants: a meta-analysis
of efficacy and tolerability. Journal of Affective Disorders. 2000; 58:19–36. [PubMed: 10760555]
Bennett CM, Miller MB. How reliable are the results from functional magnetic resonance imaging?
Annals of the New York Academy of Sciences. 2010; 1191(1):133–155. [PubMed: 20392279]
Berenbaum H. Classification and psychopathology research. Journal of Abnormal Psychology. 2013;
122:894–901. [PubMed: 24016025]
Blashfield, RK. The classification of psychopathology: Neo-Kraepelinian and quantitative Approaches.
Plenum; New York: 1984.
Block, J. Advancing the psychology of personality: Paradigmatic shift or improving the quality of
research?. In: Magnusson, D.; Endler, NS., editors. Personality at the crossroads: Current issues in
Author Manuscript
interactional psychology. Wiley; New York: 1977. p. 37-63.

Block J. A contrarian view of the five-factor approach to personality description. Psychological
Bulletin. 1995; 117:187–215. [PubMed: 7724687]
Burke MD. Laboratory medicine in the 21st century. American Journal of Clinical Pathology. 2000;
114:841–846. [PubMed: 11338472]
Button KS, Ioannidis JP, Mokrysz C, Nosek BA, Flint J, Robinson ES, Munafò MR. Power failure:
why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience.
2013; 14(5):365–376.
Cannon TD, Keller MC. Endophenotypes in the genetic analyses of mental disorders. Annual Review
of Clinical Psychology. 2006; 2:267–290.
Cardno AG, Gottesman II. Twin studies of schizophrenia: from bow-and-arrow concordances to star
wars Mx and functional genomics. American Journal of Medical Genetics. 2000; 97:12–17.
[PubMed: 10813800]
Carroll BJ. The dexamethasone suppression test for melancholia. British Journal of Psychiatry. 1982;
Author Manuscript
140:292–304. [PubMed: 7093598]

Carson RC. Aristotle, Galileo, and the DSM taxonomy: the case of schizophrenia. Journal of
Consulting and Clinical Psychology. 1996; 64:1133–1139. [PubMed: 8991300]
Chakos M, Lieberman J, Hoffman E, Bradford D, Sheitman B. Effectiveness of second-generation
antipsychotics in patients with treatment-resistant schizophrenia: a review and meta-analysis of
randomized trials. FOCUS. 2014; 2:111–121.
Chase HW, Fournier JC, Greenberg T, Almeida JR, Stiffler R, et al. Accounting for Dynamic
Fluctuations across Time when Examining fMRI Test-Retest Reliability: Analysis of a Reward
Paradigm in the EMBARC Study. PLoSONE. 2015:e0126326.
Chen Y, Bidwell LC, Norton D. Trait vs. state markers for schizophrenia: Identification and
characterization through visual processes. Current Psychiatry Reviews. 2006; 2:431–438.
[PubMed: 17487285]
Compton WM, Guze SB. The neo-Kraepelinian revolution in psychiatric diagnosis. European Archives
of Psychiatry and Clinical Neuroscience. 1995; 245:196–201. [PubMed: 7578281]
Author Manuscript
Costafreda SG, Brammer MJ, Vêncio RZ, Mourao ML, Portela LA, de Castro CC, Amaro E. Multisite
fMRI reproducibility of a motor task using identical MR systems. Journal of Magnetic Resonance
Imaging. 2007; 26(4):1122–1126. [PubMed: 17896376]
Coppen A, Abou-Saleh M, Milln P, Metcalfe M, Harwood J, Bailey J. Dexamethasone suppression test
in depression and other psychiatric illness. The British Journal of Psychiatry. 1983; 142:498–504.
[PubMed: 6409195]
Cross-Disorder Group of the Psychiatric Genomics Consortium. Identification of risk loci with shared
effects on five major psychiatric disorders: a genome-wide analysis. The Lancet. 2013; 381:1371–
1379.
Cuthbert BN. The RDoC framework: Facilitating transition from ICD/DSM to dimensional approaches
that integrate neuroscience and psychopathology. World Psychiatry. 2014a; 13:28–35. [PubMed:
24497240]
Cuthbert BN. Response to Lilienfeld. Behaviour Research and Therapy. 2014b; 62:140–142. [PubMed:
25294624]
Cuthbert BN, Insel TR. Toward the future of psychiatric diagnosis: The seven pillars of RDoC. BMC
Author Manuscript
medicine. 2013; 11:126. [PubMed: 23672542]

Cuthbert BN, Kozak MJ. Constructing constructs for psychopathology: The NIMH research domain
criteria. Journal of Abnormal Psychology. 2013; 122:928–937. [PubMed: 24016027]
Dempster EL, Pidsley R, Schalkwyk LC, Owens S, Georgiades A, Kane F, Mill J. Disease-associated
epigenetic changes in monozygotic twins discordant for schizophrenia and bipolar disorder.
Human Molecular Genetics. 2011; 20:86–4796.
Epstein S. The stability of behavior: II. Implications for psychological research. American
Psychologist. 1980; 35:790–806.
Feighner JP, Robins E, Guze SB, Woodruff RA, Winokur G, Munoz R. Diagnostic criteria for use in
psychiatric research. Archives of General Psychiatry. 1972; 26:57–63. [PubMed: 5009428]
Author Manuscript
Feinstein A. The pre-therapeutic classification of co-morbidity in chronic disease. Journal of Chronic

Disease. 1970; 24:455–468.
Flint J, Munafò MR. The endophenotype concept in psychiatric genetics. Psychological Medicine.
2007; 37:163–180. [PubMed: 16978446]
Frances A. The DSM-III personality disorders section: A commentary. American Journal of
Psychiatry. 1980; 137:1050–1054. [PubMed: 7425153]
Frances, A. Saving normal: An insider's revolt against out-of-control psychiatric diagnosis, DSM-5,
Big Pharma, and the medicalization of ordinary life. William Morrow; NY: 2013.
Frances A. RDoC is necessary, but very oversold. World Psychiatry. 2014; 13(1):47–49. [PubMed:
24497248]
Franklin JC, Jamieson JP, Glenn CR, Nock MK. How developmental psychopathology theory and
research can inform the Research Domain Criteria (RDoC) project. Journal of Clinical Child &
Adolescent Psychology. 2014:1–11. (ahead-of-print).
Fu CH, Costafreda SG. Neuroimaging-based biomarkers in psychiatry: Clinical opportunities of a
Author Manuscript
paradigm shift. Canadian Journal of Psychiatry. 2013; 58:499–508. [PubMed: 24099497]

Galatzer-Levy IR, Bryant RA. 636,120 Ways to Have Posttraumatic Stress Disorder. Perspectives on
Psychological Science. 2013; 8:651–662. [PubMed: 26173229]
Gangestad S, Snyder M. "To carve nature at its joints": On the existence of discrete classes in
personality. Psychological Review. 1985; 92:317–349.
Garb HN, Lilienfeld SO, Nezworski MT, Wood JM, O'Donohue WT. Can quality improvement
processes help psychological assessment meet the demands of evidence-based practice? The
Scientific Review of Mental Health Practice. 2009; 7:17–25.
Gould TD, Gottesman II. Psychiatric endophenotypes and the development of valid animal models.
Genes, Brain and Behavior. 2006; 5:113–119.
Graham, G. The disordered mind: An introduction to philosophy of mind and mental illness.
Routledge; New York: 2013.
Greenberg, G. The book of woe: The DSM and the unmaking of psychiatry. Penguin; New York: 2013.
Harkness AR, Lilienfeld SO. Individual differences science for treatment planning: Personality traits.
Author Manuscript
Psychological Assessment. 1997; 9:349–360.

Harkness AR, Lilienfeld SO. Science should drive the bus of clinical description; But how does
“Science take the wheel”?: A commentary on Markon. Journal of Personality Disorders. 2013;
27:580–589. [PubMed: 24044660]
Harkness AR, Reynolds SM, Lilienfeld SO. A review of systems for psychology and psychiatry:
adaptive systems, personality psychopathology five (PSY–5), and the DSM–5. Journal of
Personality Assessment. 2013 (ahead-of-print).
Haslam N, Holland E, Kuppens P. Categories versus dimensions in personality and psychopathology:
A quantitative review of taxometric research. Psychological Medicine. 2012; 42:903–920.
[PubMed: 21939592]
Hershenberg R, Goldfried MR. Implications of RDoC for the research and practice of psychotherapy.
Behavior Therapy. 2015; 46:156–165. [PubMed: 25645165]
Hyman SE. The diagnosis of mental disorders: the problem of reification. Annual Review of Clinical
Psychology. 2010; 6:155–179.
Author Manuscript
Iacono WG, Tuason VB, Johnson RA. Dissociation of smooth-pursuit and saccadic eye tracking in
remitted schizophrenics: An ocular reaction time task that schizophrenics perform well. Archives
of General Psychiatry. 1981; 38:991–996. [PubMed: 7283671]
Insel TR. Translating scientific opportunity into public health impact: a strategic plan for research on
mental illness. Archives of General Psychiatry. 2009; 66:128–133. [PubMed: 19188534]
Insel TR. The NIMH Research Domain Criteria (RDoC) Project: Precision medicine for psychiatry.
American Journal of Psychiatry. 2014; 171:395–397. [PubMed: 24687194]
Insel T, Cuthbert B, Garvey M, Heinssen R, Pine DS, Quinn K, Wang P. Research domain criteria
(RDoC): toward a new classification framework for research on mental disorders. American
Author Manuscript
Journal of Psychiatry. 2010; 167:748–751. [PubMed: 20595427]

Jonas KG, Markon KE. A meta-analytic evaluation of the endophenotype hypothesis: Effects of
measurement paradigm in the psychiatric genetics of impulsivity. Journal of Abnormal
Psychology. 2014; 123:660–675. [PubMed: 24978691]
Kapur S, Phillips AG, Insel TR. Why has it taken so long for biological psychiatry to develop clinical
tests and what to do about it. Molecular Psychiatry. 2012; 17:1174–1179. [PubMed: 22869033]
Kashdan TB, Elhai JD, Frueh BC. Anhedonia and emotional numbing in combat veterans with PTSD.
Behavioral Research Therapy. 2015; 44:456–467.
Kato T, Iwamoto K, Kakiuchi C, Kuratomi G, Okazaki Y. Genetic or epigenetic difference causing
discordance between monozygotic twins as a clue to molecular basis of mental disorders.
Molecular Psychiatry. 2005; 10:622–630. [PubMed: 15838537]
Kell DB, Oliver SG. Here is the evidence, now what is the hypothesis? The complementary roles of
inductive and hypothesis-driven science in the post-genomic era. Bioessays. 2004; 26:99–105.
[PubMed: 14696046]
Author Manuscript
Kendell RE. Clinical validity. Psychological Medicine. 1989; 19:45–55. [PubMed: 2657832]
Kendler KS. Toward a philosophical structure for psychiatry. American Journal of Psychiatry. 2005;
162:433–440. [PubMed: 15741457]
Kendler KS. The dappled nature of causes of psychiatric illness: replacing the organic–functional/
hardware–software dichotomy with empirically based pluralism. Molecular Psychiatry. 2012;
17:377–388. [PubMed: 22230881]
Kendler KS. DSM issues: incorporation of biological tests, avoidance of reification, and an approach to
the “Box Canyon Problem”. American Journal of Psychiatry. 2014; 171:1248–1250. [PubMed:
25756765]
Kendler KS, Neale MC. Endophenotype: a conceptual analysis. Molecular Psychiatry. 2010; 15:789–
797. [PubMed: 20142819]
Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves LJ. A longitudinal twin study of personality and
major depression in women. Archives of General Psychiatry. 1993; 50:853–862. [PubMed:
8215811]
Author Manuscript
Kidd C, Palmeri H, Aslin RN. Rational snacking: Young children’s decision-making on the
marshmallow task is moderated by beliefs about environmental reliability. Cognition. 2013;
126:109–114. [PubMed: 23063236]
Kihlstrom, JF. To honor Kraepelin…: From symptoms to pathology in the diagnosis of mental illness.
In: Beutler, LE.; Malik, ML., editors. Rethinking the DSM: Psychological perspectives. American
Psychological Association; Washington, D.C.: 2002. p. 279-303.
Kirk, SA.; Kutchins, H. The selling of DSM: The rhetoric of science in psychiatry. Transaction
Publishers; New York: 1992.
Kupfer DJ, Foster FG, Coble P, McPartland RJ, Ulrich RF. The application of EEG sleep for the
differential diagnosis of affective disorders. American Journal of Psychiatry. 1978; 135:69–74.
[PubMed: 201174]
Kupfer DJ, Regier DA. Why all of medicine should care about DSM-5. Journal of the American
Medical Association. 2010; 303:1974–1975. [PubMed: 20483976]
Lakatos, I. Philosophical Papers. Vol. 1. Cambridge University Press; Cambridge, U.K.: 1978. The
methodology of scientific research programmes.
Author Manuscript
Laska KM, Gurman AS, Wampold BE. Expanding the lens of evidence-based practice in
psychotherapy: A common factors perspective. Psychotherapy. 2014; 51:467–481. [PubMed:
24377408]
Lieberman, JA.; Ogas, O. Shrinks: The untold story of psychiatry. Little, Brown, and Company;
Boston: 2015.
Lilienfeld SO. The Dodo Bird verdict: Status in 2014. Behavior Therapist. 2014a; 37:91–95.
Lilienfeld SO. The Research Domain Criteria (RDoC): An analysis of methodological and conceptual
challenges. Behaviour Research and Therapy. 2014b; 62:129–139. [PubMed: 25156396]
Lilienfeld SO, Marino L. Essentialism revisited: evolutionary theory and the concept of mental
disorder. Journal of Abnormal Psychology. 1999; 106:400–411.
Author Manuscript
Lilienfeld, SO.; Smith, SF.; Watts, AL. Issues in diagnosis: Conceptual issues and controversies. In:
Craighead, E.; Mikllowitz, J.; Craighead, LW., editors. Psychopathology: History, diagnosis, and
empirical foundations. Wiley; New York: 2013. p. 1-35.
Lilienfeld SO, Waldman ID. The relation between childhood attention-deficit hyperactivity disorder
and adult antisocial behavior reexamined: The problem of heterogeneity. Clinical Psychology
Review. 1990; 10:699–725.
Lilienfeld SO, Waldman ID. Comorbidity and Chairman Mao. World Psychiatry. 2004; 3:26–27.
[PubMed: 16633448]
Lilienfeld SO, Waldman ID, Israel AC. A critical examination of the use of the term and concept of
comorbidity in psychopathology research. Clinical Psychology: Science and Practice. 1994; 1:71–
83.
Lilienfeld SO, Watts AL, Francis Smith S, Berg JM, Latzman RD. Psychopathy deconstructed and
reconstructed: Identifying and assembling the Personality building blocks of Cleckley's chimera.
Journal of Personality. 2014 awaiting volume and page numbers.
Author Manuscript
MacDonald AW, Krueger RF. Mapping the country within: A Special Section on reconceptualizing the
classification of mental disorders. Journal of Abnormal Psychology. 2013; 122:991–993.
Lopez-Munoz F, Alamo C. Monoaminergic neurotransmission: the history of the discovery of
antidepressants from 1950s until today. Current Pharmaceutical Design. 2009; 15(14):1563–1586.
[PubMed: 19442174]
MacDonald AW III, Krueger RF. Mapping the country within: a special section on reconceptualizing
the classification of mental disorders. Journal of Abnormal Psychology. 2013; 122:891–893.
[PubMed: 24016023]
Markon KE. Epistemological pluralism and scientific development: An argument against authoritative
nosologies. Journal of Personality Disorders. 2013; 27:554–579. [PubMed: 24044659]
Markon KE, Chmielewski M, Miller CJ. The reliability and validity of discrete and continuous
measures of psychopathology: a quantitative review. Psychological Bulletin. 2011; 137:856–879.
[PubMed: 21574681]
Macrae CN, Bodenhausen GV. Social cognition: Thinking categorically about others. Annual Review
Author Manuscript
of Psychology. 2000; 51:93–120.

Maj M. ‘Psychiatric comorbidity’: an artefact of current diagnostic systdems? The British Journal of
Meehl PE. Specific etiology and other forms of strong influence: Some quantitative meanings. Journal
of Medicine and Philosophy. 1977; 2:33–53.
Meehl PE. Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft
psychology. Journal of Consulting and Clinical Psychology. 1978; 46:806–834.
Meehl, PE.; Golden, R. Taxometric methods. In: Kendall, P.; Butcher, J., editors. Handbook of research
methods in clinical psychology. Wiley; New York: 1982. p. 127-181.
Miller GA, Rockstroh B. Endophenotypes in psychopathology research: Where do we stand? Annual
Review of Clinical Psychology. 2013; 9:177–213.
Mirnezami R, Nicholson J, Darzi A. Preparing for precision medicine. New England Journal of
Medicine. 2012; 366:489–491. [PubMed: 22256780]
Mischel, W. Personality and assessment. John Wiley & Sons; New York: 1968.
Author Manuscript
Mischel, W. Processes in delay of gratification. Academic Press; New York: 1974.

Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD. The unity and diversity of
executive functions and their contributions to complex “frontal lobe” tasks: A latent variable
analysis. Cognitive Psychology. 2000; 41(1):49–100. [PubMed: 10945922]
Molnar BE, Buka SL, Kessler RC. Child sexual abuse and subsequent psychopathology: results from
the National Comorbidity Survey. American Journal of Public Health. 2001; 91:753–760.
[PubMed: 11344883]
Monroe SM, Anderson SF. Depression The Shroud of Heterogeneity. Current Directions in
Psychological Science. 2015; 24:227–231.
Morris SE, Cuthbert BN. Research Domain Criteria: cognitive systems, neural circuits, and dimensions
of behavior. Dialogues in Clinical Neuroscience. 2012; 14(1):29–37. [PubMed: 22577302]
Author Manuscript
Moss HB, Chen CM, Yi HY. Subtypes of alcohol dependence in a nationally representative sample.
Drug and Alcohol Dependence. 2007; 91:149–158. [PubMed: 17597309]
Nemeroff CB, Kilts CD, Berns GS. Functional brain imaging: twenty-first century phrenology or
psychobiological advance for the millennium? The American Journal of Psychiatry. 1999;
156:671–673. [PubMed: 10327896]
Neimark J. Philosophical issues in psychiatry. American Journal of Psychiatry. 2009; 166:1301.
Noga JT, Aylward E, Barta PE, Pearlson GD. Cingulate gyrus in schizophrenic patients and normal
volunteers. Psychiatry Research. 1995; 6(1):201–208.
O'Connor T. Emergent properties. American Philosophical Quarterly. 1994:91–104.
Paris, J. The ideology behind DSM-5. In: Paris, J.; Phillips, J., editors. Making the DSM-5. New York;
Springer: 2013. p. 39-44.
Patrick CJ, Venables NC, Yancey JR, Hicks BM, Nelson LD, Kramer MD. A construct-network
approach to bridging diagnostic and physiological domains: Application to assessment of
externalizing psychopathology. Journal of Abnormal Psychology. 2013; 122:902–916. [PubMed:
Author Manuscript
24016026]
Paulus MP. Pragmatism instead of mechansm: a call for impactful biological psychiatry. JAMA
Petronis A, Gottesman II, Kan P, Kennedy JL, Basile VS, Paterson AD, Popendikyte V. Monozygotic
twins exhibit numerous epigenetic differences: clues to twin discordance? Schizophrenia
Bulletin. 2003; 29:169–178. [PubMed: 12908672]
Pigliucci, M.; Boudry, M., editors. Philosophy of pseudoscience: Reconsidering the demarcation
problem. University of Chicago Press; Chicago, Ill: 2013.
Pincus HA, Tew JD Jr, First MB. Psychiatric comorbidity: Is more less? World Psychiatry. 2004; 3:18–
23. [PubMed: 16633444]
Polich, J. Neuropsychology of P300. In: Lack, SJ.; Kappenman, ES., editors. Oxford handbook of
event-related potential components. Oxford University Press; NY: 2012. p. 159-188.
Popper, KR. The logic of scientific discovery. Routledge Classics; New York: 1959.
Regier D, Narrow W, Kuhl E, Kupfer D. The conceptual development of DSM-V. American Journal of
Author Manuscript

Robins E, Guze SB. Establishment of diagnostic validity in psychiatric illness: Its application to
schizophrenia. American Journal of Psychiatry. 1970; 126:983–987. [PubMed: 5409569]
Sanislow CA, Pine DS, Quinn KJ, Kozak MJ, Garvey MA, Heinssen RK, Cuthbert BN. Developing
constructs for psychopathology research: research domain criteria. Journal of Abnormal
Psychology. 2010; 119:631–639. [PubMed: 20939653]
Sauder CL, Hajcak G, Angstadt M, Phan KL. Test-retest reliability of amygdala response to emotional
faces. Psychophysiology. 2013; 50:1147–1156. [PubMed: 24128307]
Shankman SA, Gorka SM. Psychopathology research in the RDoC era: Unanswered questions and the
importance of the psychophysiological unit of analysis. International Journal of
Psychophysiology. 2015 awaiting volume and page numbers.
Shilling VM, Chetwynd A, Rabbitt PMA. Individual inconsistency across measures of inhibition: An
investigation of the construct validity of inhibition in older adults. Neuropsychologia. 2002;
40:605–619. [PubMed: 11792402]
Author Manuscript
Skodol AE, Bender DS, Morey LC, Clark LA, Oldham JM, Alarcon RD, Siever LJ. Personality
disorder types proposed for DSM-5. Journal of Personality Disorders. 2011; 25:136–169.
[PubMed: 21466247]
Sneath PH. Some thoughts on bacterial classification. Journal of General Microbiology. 1957; 17:184–
200. [PubMed: 13475685]
Spitzer RL, Endicott J, Robins E. Research diagnostic criteria: Rationale and reliability. Archives of
General Psychiatry. 1978; 35:773–782. [PubMed: 655775]
Spitzer, RL.; Frances, A. Spitzer/Frances letter to APA trustees. Psychology Today. 2010. Retrieved
from http://www.psychologytoday.com/blog/dsm5-in-distress/201012/spritzer/frances-letter-apa-
Author Manuscript
trustees
Stein DJ, Black DW, Pienaar W. Sexual disorders not otherwise specified: compulsive, addictive, or
impulsive? CNS Spectrums. 2000; 5(01):60–66. [PubMed: 18311101]
Stoyanov ST, Kandilavora S. State versus trait diagnostic biomarkers in psychiatry and the issue of
translation. Austin Biomarkers and Diagnosis. 2014; 1(2):1–5.
Tan HY, Callicott JH, Weinberger DR. Intermediate phenotypes in schizophrenia genetics redux: Is it a
no brainer? Molecular Psychiatry. 2008; 13:233–238. [PubMed: 18285755]
Tellegen, A. Personality traits: Issues of definition, evidence, and assessment. In: Cicchetti, D.; Grove,
WM., editors. Personality and psychopathology. Vol. 2. University of Minnesota Press;
Minneapolis, MN: 1991. p. 10-35.Thinking clearly about psychology: Essays in honor of Paul E
Meehl
Torgersen T, Gjervan B, Rasmussen K. ADHD in adults: a study of clinical characteristics, impairment
and comorbidity. Nordic Journal of Psychiatry. 2006; 60:38–43. [PubMed: 16500798]
Tsay CJ. Julius Wagner-Jauregg and the Legacy of Malarial therapy for the treatment of General
Author Manuscript
Paresis of the insane. The Yale Journal of Biology and Medicine. 2013; 86:245–254. [PubMed:
23766744]
Tsuang MT, Stone WS, Faraone SV. Towards the prevention of schizophrenia. Biological Psychiatry.
2000; 48:349–356. [PubMed: 10978718]
Tryon RC. Basic unpredictability of individual responses to discrete stimulus presentations.
Multivariate Behavioral Research. 1973; 8:275–295. [PubMed: 26828655]
Verheul R, Widiger TA. A meta-analysis of the prevalence and usage of the personality disorder not
otherwise specified (PDNOS) diagnosis. Journal of Personality Disorders. 2004; 18:309–319.
[PubMed: 15342320]
Vul E, Harris C, Winkielman P, Pashler H. Puzzlingly high correlations in fMRI studies of emotion,
personality, and social cognition. Perspectives on Psychological Science. 2009; 4:274–290.
[PubMed: 26158964]
Wakefield JC. The concept of mental disorder: on the boundary between biological facts and social
values. American Psychologist. 1992; 47:373–388. [PubMed: 1562108]
Author Manuscript
Wakefield JC. Wittgenstein's nightmare: why the RDoC grid needs a conceptual dimension. World
Psychiatry. 2014; 13(1):38–40. [PubMed: 24497242]
Waldman ID. Statistical approaches to complex phenotypes: evaluating neuropsychological
endophenotypes for attention-deficit/hyperactivity disorder. Biological Psychiatry. 2005;
57:1347–1356. [PubMed: 15950007]
Weinberger DR, Goldberg TE. RDoCs redux. World Psychiatry. 2014:36. [PubMed: 24497241]
Westen D. Prototype diagnosis of psychiatric syndromes. World Psychiatry. 2012; 11:16–21. [PubMed:
22294998]
Whooley, O.; Horwitz, AV. The paradox of professional success: Grand ambition, furious resistance,
and the derailment of the DSM-5 revision process. In: Paris, J.; Phillips, J., editors. Making the
DSM-5. Springer; New York: 2013. p. 75-92.
Widiger TA, Clark LA. Toward DSM-V and the classification of psychopathology. Psychological
Bulletin. 2000; 126:946–963. [PubMed: 11107884]
Widiger TA, Trull TJ. Plate tectonics in the classification of personality disorder: Shifting to a
dimensional model. American Psychologist. 2007; 62:71–83. [PubMed: 17324033]
Author Manuscript
Wilson M. DSM-III and the transformation of American psychiatry: a history. American Journal of
Wonderlick JS, Ziegler DA, Hosseini-Varnamkhasti P, Locascio JJ, Bakkour A, Van Der Kouwe A,
Dickerson BC. Reliability of MRI-derived cortical and subcortical morphometric measures:
effects of pulse sequence, voxel geometry, and parallel imaging. Neuroimage. 2009; 44:1324–
1333. [PubMed: 19038349]
Author Manuscript
Author Manuscript
Figure 1.
The Provisional Matrix for the Research Domain Criteria (RDoC)
Author Manuscript
Author Manuscript
Table 1
A Baker’s Dozen Recommendations for RDoC

Author Manuscript
(1) Lay out explicit benchmarks for ascertaining what findings, or patterns of findings, could
falsify the RDoC research framework or at least suggest that it is not making adequate scientific
progress
(2) Examine how RDoC is performing relative to DSM-ICD for statistically predicting important
external criteria, such as response to treatment and natural history
(3) Explicate alternative models for dysfunction in psychobiological systems (e.g.., lesion,
hyperactivity or hypoactivity in the functioning of these systems, a network model in which there
are bidirectional causal relations among diagnostic signs and symptoms)
(4) Accord adequate attention to environmental (e.g., social and cultural context) and
developmental (e.g., the effects of developmental timing) influences, and incorporate such
influences explicitly into the RDoC matrix to encourage their investigation
(5) Adopt no a priori assumptions regarding what measures will be optimal or detecting
individual differences in neural circuitry relevant to psychopathology, and be guided exclusively
by data in this regard
(6) Accord adequate attention to the role of measurement error inherent to laboratory measures
(“T data”) and capitalize on the power of statistical aggregation across indices.
Author Manuscript
(7) When conducting research on endophenotypes, test statistical models of pleiotropy and
mediation, and do not presume that such phenotypes are necessarily more heritable or
straightforward in their genetic architecture compared with exophenotypes
(8) Do not assume that endophenotypes or other candidate biomarkers are more trait-like than
state-like; the assertion that such indicators are stable over time must be demonstrated
empirically
(9) Remain cognizant of the possibility that psychobiological predispositions can be manifested
in a host of diverse behavioral manifestations
(10) Avoid premature reification of the RDoC matrix and remain open to research that examines
cells not explicitly represented in the current matrix
(11) Develop broad guidelines for beginning to translate the growing body of RDoC findings into
a usable system of psychiatric classification
(12) Seek ways in which RDoC can co-exist with both (a) basic research on neural circuitry and
(b) applied research on empirically supported treatments tied to DSM categories
(13) Explore the possibility of a combined system that incorporates both diagnostic
sign/symptoms and psychobiological predispositions
Author Manuscript
Author Manuscript

DSM Versus RDoC

Uploaded by

Copyright:

Available Formats

DSM Versus RDoC

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

DSM Versus RDoC

Uploaded by

Copyright:

Available Formats

HHS Public Access

Published in final edited form as:

Clashing Diagnostic Approaches: DSM-ICD versus RDoC

-Sir Isaac Newton

DSM-ICD and Kraepelin: Hints of a Paradigm Shift

Costafreda, 2013) or at least a “reboot” (Kendler, 2015) in psychiatric classification have

possibility of a rapprochement between these two overarching approaches.

DSM and ICD: Origins and Assumptions

modest inter-rater reliabilities of their diagnostic categories, the American Psychiatric

Scientifically arbitrary diagnostic cut-offs—The DSM is technically agnostic with

Heterogeneity—The polythetic model of DSM-III-R and later DSM editions, sometimes

compatible with a diagnosis of borderline personality disorder. As an even more extreme

Large number of intermediate cases—As noted earlier, an optimal classification

Genetic and environmental nonspecificity.—If DSM-ICD conditions were largely

Similarly, there is compelling evidence for substantial environmental nonspecificity in the

DSM-ICD anomalies: Taking stock

phrase (see Gangestad & Snyder, 1985). .

RDoC: Intellectual Antecedents

Laboratory methods in medicine

Early triumphs in psychopathology research

disorder, and obsessive-compulsive disorder, would similarly be traceable to a single

and dementia (Coppen et al., 1983). As a consequence, a state of disenchantment regarding

The slow progress of psychopharmacology

growth” of psychopharmacology, witnessed the discovery of the first tricyclic

The Birth of RDoC: Fundamental Presuppositions and Structure

supported model of neural circuitry. RDoC conceptualizes mental disorders as dysfunctions

that many superficially different conditions are underpinned by similar psychobiological

Consistent with the view of science as a self-correcting enterprise, RDoC is conceptualized

The Scientific Promise of RDoC

and Markon, 2013). Moreover, RDoC’s adoption of a dimensional approach to

Furthermore, RDoC promises to capitalize on the burgeoning corpus of knowledge

RDoC: Conceptual Challenges

RDoC is a flawed endeavor that is not leading to a feasible alternative model of

Potential overemphasis on biological “dysfunction” at the level of an individual

responses to trauma–or gradual, such as the slow accumulation of glucocorticoid resistance

socially traditional environment.

reported that scores on self-report measures of an externalizing propensity and aggression

Biological predispositions versus their behavioral manifestations

RDoC: Methodological Challenges

Neglect of Measurement Error

Laboratory measures are typically associated with largely unappreciated psychometric

they contain substantial elements of “situational uniqueness.” Performance on such

seemingly trivial changes in experimental paradigms can lead to dramatic changes in

the anticipation phase (Chase et al 2015). As would be predicted by temporal-discounting

The Limitations of Endophenotypes

more construct-valid indicators of genetically influenced psychobiological systems.

Moreover, because many endophenotypes cut across traditional disorder categories,

potentially relevant to impulsivity and (b) diagnostic (e.g., measures of attention-deficit

mediation for Trails B.

RDoC: Logistical/Pragmatic Challenges

The dangers of premature reification of the RDoC matrix

appears to vitiate RDoC’s avowed goal of being self-correcting, as it will be difficult or

Implications for grant funding in non-RDoC domains

Translating RDoC findings into a feasible classification system

The DSM-ICD and RDoC: Quo Vadis?

classification bereft of any reference to the subjective and behavioral manifestations of

cannot be fixed merely by adjusting some of its walls.

As we have also noted, RDoC confronts a number of conceptual, methodological, and

logistical/pragmatic challenges, none of which appear to be insuperable (Lilienfeld, 2014b).

necessary for a complete characterization of mental disorder.