The Parkinson Anxiety Scale (PAS): development and validation of a new anxiety scale.

Authors:

Leentjens AFG, M.D., Ph.D.1

Dujardin K, Ph.D.2

Pontone GM, M.D.3.

Starkstein SE, M.D.4

Weintraub D, M.D., Ph.D.5

Martinez-Martin P, M.D., Ph.D. 6

1
Department of Psychiatry, Maastricht University Medical Center, Maastricht, the

Netherlands
2
Neurology and Movement Disorders Unit, Lille University Medical Center, Lille, France
3
Department of Psychiatry and Behavioral Science, Johns Hopkins University School of

Medicine, Baltimore, U.S.A

4
School of Psychiatry, University of Western Australia and Fremantle Hospital, Fremantle,

Australia
5
Departments of Psychiatry and Neurology, Perelman School of Medicine at the University

of Pennsylvania, and Parkinson's Disease Research, Education and Clinical Center,

Philadelphia Veterans Affairs Medical Center, Philadelphia, Pennsylvania, U.S.A.

6
Alzheimer Disease Research Unit and CIBERNED, CIEN Foundation, Carlos III Institute of

Health. Alzheimer Center Reina Sofia Foundation, Madrid, Spain

1
Correspondence to:

Dr. A.F.G. Leentjens,

Department of Psychiatry

Maastricht University Medical Center

P.O. Box 5800

6202 AZ Maastricht

the Netherlands

phone ++ 31 43 3877443

fax ++ 31 43 3875444

email: a.leentjens@np.unimaas.nl

Running title: the Parkinson Anxiety Scale (PAS)

Word count: 3300

Tables 6; figures: none

Financial disclosure: none of the authors had a conflict of interest related to research covered

in this article. For full financial disclosure: see end of manuscript.

2
Abstract

Background: Existing anxiety rating scales have limited construct validity in patients with

Parkinson’s disease (PD).

Aim: To develop and validate a new anxiety rating scale, the Parkinson Anxiety Scale (PAS),

that would overcome the limitations of existing scales.

Design and methods: The general structure of the PAS was based on the outcome of a Delphi

procedure. Item selection was based on a canonical correlation analysis and a Rasch analysis

of items of Hamilton Anxiety Rating Scale (HARS) and Beck Anxiety Inventory (BAI) from

a previously published study.

Validation was done in a cross-sectional international multicenter study involving 362

patients with idiopathic PD. Patients underwent a single screening session in which the PAS

was administered, along with the Hamilton Depression Rating Scale, the HARS and BAI. The

Mini International Neuropsychiatric Interview was administered to establish DSM diagnoses

of anxiety and depressive disorders.

Results: The PAS is a 12-item observer or patient-rated scale with three subscales, for

persistent, episodic anxiety, and avoidance behavior.

Properties for acceptability and reliability met predetermined criteria. The convergent and

known groups validity was good. The scale has a satisfactory factorial structure. The AUC

and Youden index of the PAS is higher than that of existing anxiety rating scales.

Conclusion: The PAS is a reliable and valid anxiety measure for use in PD patients. It is easy

and brief to administer, and has better clinimetric properties than existing anxiety rating

scales. The ‘sensitivity to change’ of the PAS remains yet to be assessed.

3
Introduction

Anxiety disorders are common in patients with Parkinson’s disease (PD). Twenty-five to 43%

of PD patients have a circumscribed anxiety disorder using DSM IV criteria 1-4, making

anxiety disorders more prevalent than depressive disorders, which have an average prevalence

of 17% 5. Four to 8% of patients suffers from panic disorder, 2 to 16% from agoraphobia

without panic, 3 to 21% from generalized anxiety disorder (GAD), and 8% to 13% from

social phobia (or social anxiety disorder) 1, 2, 4. Anxiety disorders with persisting

symptomatology, such as GAD, are thus more common than episodic anxiety disorders, such

as panic disorders. Anxiety in PD is associated with increased subjective motor symptoms,

more severe gait problems and dyskinesias, freezing and on/off fluctuations 6-8. Anxiety

symptoms in PD patients also have a negative impact on health related quality of life 6-8.

Reliable assessment and measurement of symptoms may help to establish a diagnosis, guide

the decision to start treatment for anxiety, and monitor this treatment. A recent review of

anxiety rating scales commissioned by the Movement Disorder Society (MDS) revealed that

none of the available anxiety rating scales had been validated in patients with Parkinson’s

disease (PD) 9. A subsequent validation study of the Hamilton Anxiety Rating Scale (HARS),

the Beck Anxiety Inventory (BAI) and the Hospital Anxiety and Depression Scale (HADS)

raised questions about the construct validity of these scales. It was shown that the BAI, with

its focus on symptoms of panic, identifies a different group of patients than the HARS, with

its focus on generalized anxiety. Moreover, the positive predictive value of all scales was

poor, and the negative predictive value was only moderate. In addition, it was not possible to

identify a satisfactory factorial structure for the BAI and HADS 10.

The aim of this study was to develop and validate an anxiety scale that would overcome the

limitations of existing scales and have adequate construct validity and clinimetric properties

4
in patients with PD. The scale should be brief and easy to administer, and have both an

observer-rated version, to be scored by the clinician or researcher and a self-rated version, to

be scored by the patient. The authors furthermore state that this scale should be in the public

domain and can be used free of charge for research and clinical purposes.

Patients and Methods

Development of the PAS

The authors of this paper held a consensus meeting to reach agreement on the general format

and structure of the scale. A first exploratory Delphi round focused on the general format of

the scale, including the time frame used for symptom reporting, potential subscales, number

of items, as well as scoring issues, such as the range of scoring options and the formulation of

Likert scores. Initial item selection was based both on a canonical correlation analysis (CCA)

of items of the HARS and BAI to ensure a broad enough scope of symptoms, as well as on a

Rasch analysis of these scales to ensure selection of items that would represent the whole

spectrum of severity11, 12. Decisions were guided by exploratory analysis on data of the

previously published study of the validation of the HARS and BAI {Leentjens, 2011 #1618}.

In a first step, items influenced by age, sex and severity of motor symptoms were eliminated,

as were items that were shown not to be reliable in the Rasch analysis. Items with too little

item difficulty (less than minus 2 logits) in the Rasch analysis were also eliminated, as were

items that did not cluster with recognizable anxiety symptoms in the CCA analysis. A second

Delphi round was held focusing on item selection from the remaining pool of items and item

formulation. A more detailed description of the development process is reported in the

supplementary material (Appendix 1).

5
Validation of the PAS

Design and setting

The newly developed scale was validated in a one-year cross-sectional international multi-

centre study that included 362 patients from six tertiary referral centers: two in the United

States (Johns Hopkins University School of Medicine, Baltimore, Maryland and the

University of Pennsylvania, Philadelphia, Pennsylvania), three in Europe (Lille University

Medical Center, Lille, France, the Carlos III Institute of Health, Madrid, Spain, and

Maastricht University Medical Center, Maastricht, the Netherlands) and one in Australia

(Fremantle Hospital, Fremantle). Patients were recruited from Movement Disorders clinics, as

well as from Neurology and Psychiatry outpatient clinics of the participating centers.

Enrollment occurred between March 2012 and April 2013.

Inclusion - and exclusion criteria

Patients diagnosed with idiopathic PD according to the Queen Square Brain Bank criteria, and

who gave informed consent met inclusion criteria 13. Patients with neurodegenerative

disorders other than PD, and those with severe cognitive impairment, operationalised as a

score on the Mini Mental State Examination (MMSE) < 23 were excluded 14.

Assessment

Demographic and disease related variables were recorded. Motor function, activities of daily

living (ADL) function, complications of therapy, and disease stage were assessed with the

Unified Parkinson’s Disease Rating Scale (UPDRS) sections 2,3 and 4, and the Hoehn &

Yahr staging system (H&Y) 15; assessment of cognitive abilities and instrumental ADL were

done with the MMSE, the Clinical Dementia Rating Scale (CDR) and Lawton Instrumental

6
ADL (IADL) scale 14, 16, 17. Quality of life was assessed with the eight item Parkinson’s

Disease Questionnaire (PDQ-8) 18. The presence of DSM-defined depressive and anxiety

disorders was determined with the Mini International Neuropsychiatric Inventory (MINI, a

structured interview for DSM disorders) sections for depression (A, B) and anxiety (D, E, F,

H) 19. The severity of depressive symptoms was quantified with the 17-item Hamilton

Depression Rating Scale (HAMD) 20.

The anxiety measures included the observer-rated and patient-rated versions of the PAS (for a

description: see below). Other measures included the observer-rated Clinical Global

Impression (CGI) and patient-rated Patient Global Impression (PGI) of the severity of anxiety

symptoms 21, the observer-rated Hamilton Anxiety Rating Scale (HARS) 22, and the self-rated

Beck Anxiety Inventory (BAI) 23. Patients suffering from ‘on/off’ fluctuations were only

assessed during ‘on’ states, following the advice of the MDS task force 9.

Inter-rater reliability of the observer-rated PAS was assessed by administering this scale twice

in the same session to a minimum of 10 patients in every institution, by two different raters

(total sample size = 60). For feasibility reasons, test-retest reliability was only assessed for

the self-rated version of the PAS. This was accomplished by asking patients to complete a

second set of these scales spaced one week apart from the in-person assessment and return

them by post-mail.

Power calculation

The aforementioned validation study of anxiety rating scales showed a prevalence of anxiety

disorders of 34% in 342 PD patients. As this study followed the same design and involved the

same analyses10, we expected that a sample size of 360 would be more than adequate to

perform all planned analyses. The most demanding sample size is usually related to the factor

analysis. Allowing for a minimum of 20 subjects per scale item, and a total of 13 items we

7
surpass this requirement (minimum 260 subjects) widely. To assess inter-rater reliability and

test-retest reliability, the proposed sample size of 60 exceeds the minimum sample size of 50

subjects following the generally accepted algorithm of Cicchetti 24.

Statistical analyses

All analyses were performed using SPSS version 21.0 for Windows (SPSS Inc. Chicago,

2007). Demographic and disease related variables are presented in a descriptive way. The

Kolmogorow-Smirnow (K-S) Z test statistic was used to check whether the distribution was

gaussian. In order to identify between-institute differences in enrolled patients, demographic

and disease related variables were compared with chi-square or Kruskal-Wallis test. If p <

0.05, the differences were considered significant. The prevalence of DSM- defined anxiety

disorders in this population is reported.

Acceptability was assessed in the form of percentage of missing responses, with less then 5%

considered acceptable 25. Observed versus possible score range for items and total scores were

analyzed with distribution of scores assessed with mean, SD, median, floor and ceiling

effects, and skewness. Observed mean versus median scores can be considered a measure of

distribution and a difference < 10% of the maximum possible scale score was considered

acceptable. For floor and ceiling effects, 15% was taken as maximum acceptable 26. For

skewness the accepted limits were -1.0 to +1.0 27.

As a measure of internal consistency, Cronbach’s alpha, mean inter-item correlation (or ‘item

homogeneity coefficient’) and range, as well as corrected item-total correlation are reported.

A Cronbach’s alpha ≥0.70 was considered acceptable 28, as was item-total correlation ≥0.40
29
, and an inter item coefficient of >0.30 30.

Inter-rater reliability will only be reported for the observer-rated version of the PAS using

weighted kappa (quadratic weights) to assess reproducibility for individual item scores and

8
the intraclass correlation coefficient (ICC, two-way random effect) for the total score. Test-

retest reliability was assessed for the self-rated version of the PAS only, by intraclass

correlation coefficient (ICC, one-way random effect). A kappa or ICC ≥ 0.70 was considered

satisfactory 31. Precision (standard error of measurement, SEM) is reported.

Convergent validity of the PAS with the BAI and HARS was assessed with Spearman rank

correlation coefficients. Correlation coefficients ≥0.60 were demonstrative of a high

association 32.

Known-groups validity was assessed by comparing scores on anxiety scales with scores on

the CGI and PGI using the Kruskal-Wallis test.

Dimensionality was assessed by means of exploratory factor analysis with principal

component method and promax rotation (since it was expected that the different factors would

be correlated with each other). A scree plot and Kaiser's criterion (Eigenvalue ≥ 1) were used

to determine the number of factors, if these factors were clinically interpretable.

In order to assess the properties of the scales as diagnostic tests, criterion validity was tested

against DSM diagnoses using Receiver Operating Characteristics (ROC) curves. This was

done for the full scale (any anxiety disorder versus no anxiety disorder) and for each subscale.

For subscales the diagnostic properties were assessed in relation to specific anxiety disorders:

the persistent subscale with GAD, the episodic subscale with panic disorder and the avoidance

scale with social phobia and agoraphobia combined. Sensitivity, specificity, area under the

ROC curve (AUC) and Youden index (highest sum of sensitivity and specificity) are reported.

Approval and consent

This study received ethical approval from the Institutional Review Boards of each

participating center. Patients received written information about the study and gave their

written informed consent before participating. They received a small financial compensation

9
for participating. This study was sponsored by the Michael J. Fox Foundation for Parkinson

Research (MJFF).

Results

Development of the PAS

In the consensus meeting it was agreed that the main aim of the scale was to provide a reliable

measurement of the severity of anxiety symptoms, rather than designing an instrument

specifically intended for screening or diagnosis. Although the scale would be developed

specifically for patients with PD, the study group agreed that item selection and formulation

should be such that the scale could be administered to patients with other neurological and

physical diseases as well.

Participants reached consensus that the scale should consist of three subscales: one pertaining

to persisting anxiety (five items), one to episodic anxiety (four items), and one to avoidance

behavior (three items). These first two subscales were supported by evidence from a CC

analysis 11, while ‘avoidance' was considered a third characteristic of anxiety disorders, not

captured by the other two scales. The total scale thus consists of 12 items. Items were scored

on a 5 point Likert scale, with '0' meaning 'not or never' and '4' meaning 'severe or almost

always'. It was decided to formulate items as questions, as this way the patient-rated and

clinician rated versions would be identical. The self-rated version can be completed in less

than two minutes, while the observer-rated version may take up to five minutes if answering

options for every item are read out to the patient.

The scale was first written in English. Translations into Spanish, French and Dutch were

made for the sake of the study, following the procedure of translation and back translation to

check for inconsistencies in formulation. The English, Spanish, French and Dutch scales are

10
published in the Supplementary material as Appendix 2. The scale, including existing

translations is in the public domain and can be used free of charge for clinical and

investigational purposes.

Demographic and disease related data

Demographic and disease related variables for the included patients are displayed in Table 1.

We included 362 patients, 62.7% male, with a mean age of 65.3 years, and a median H&Y

stage of 2. Age, disease duration, as well as scores on the MMSE and IADL, differed

significantly between the centers (Supplementary material, Appendix 3) but these differences

were small and not considered clinically relevant. Scores on the UPDRS section III, H&Y

stage and PD treatment variables also differed between institutions, which was most likely

due to the different spectrum of disease severity and complexity across institutions, thought to

reflect different areas of interest or expertise.

Prevalence of anxiety disorders

Table 1 also shows the prevalence of the different anxiety disorders. Twenty-seven percent of

patients were suffering from a current DSM anxiety disorder. Among those, generalized

anxiety disorder was the most prevalent diagnosis (15.2%), followed by agoraphobia (10.5%)

and social phobia (9.1%). Panic disorder was less frequent (3.4%).

Acceptability and distribution

With less than 0.01% missing data on any of the subscales, in both the observer-rated and

patient-rated version of the instrument, the scale demonstrated good acceptability. Mean,

standard deviation, and skewness are reported in Table 2. For the total scale, no floor effects

were seen, but a floor effect could be observed for the episodic and avoidant subscales, likely

11
due to the lower prevalence of episodic anxiety and avoidance behavior. None of the

subscales has significant ceiling effects. The observed range of scores was close to the

theoretical range. For all scales, the difference between observed mean and median scores

were < 10% of the maximum possible scale score. Overall, the score distributions were

acceptable. None of the scales had normally distributed scores (p< 0.01 for the K-S Z).

Individual item performance is shown in the supplementary material (Supplementary

material; Appendix 4).

Reliability and stability

Cronbach’s alpha was good for both the patient-rated and the observer-rated total scale scores.

For the subscales Cronbach’s alpha was good for the persistent subscale, and acceptable for

the episodic subscale and acceptable for the avoidance subscale (Table 3). Item-total

correlations were high for all subscales; for the total scale the item-total correlation was

slightly less than that for the subscales, which is to be expected given the scale was designed

to represent the discrete dimensions of anxiety. Both Cronbach’s alpha and item total

correlations were slightly better for the self-rated version of the PAS. For all subscales, inter-

item correlation met the predetermined criterion. Inter-rater reliability for the observer-rated

scale and test-retest reliability for the self-rated version were both excellent (Table 4).

Convergent validity and known-groups validity

Convergent validity of PAS total scale with the CGI and PGI for anxiety, HARS and the BAI

was high (Table 5). For the subscales, this correlation was moderate to high. Divergent

validity with the HAMD was also high, although there was a moderately strong correlation

between the persistent anxiety subscale and the HAMD score, indicating that generalized

anxiety and depression are overlapping concepts.

12
Known groups validity was good: total PAS scores, as well as scores in subscales of the PAS,

were significantly higher in patients with higher scores on the CGI and PGI for both the

observer-rated as the self-rated scale (Supplementary material; Appendix 5). Scores on PGI

and CGI were strongly correlated: 0.814 (p<0.001).

Dimensionality

A scree plot, as well as Eigenvalues ≥ 1 were co-incident in identifying three factors

(Supplementary material; Appendix 6). Each PAS item loaded onto a factor with the other

items from its PAS subscale, in line with the design of the scale. For the observer-rated

version, two items (B1, ‘ panic or intense fear’ and B4, ‘fear of losing control’) had a

moderately strong loading on both the persistent and the episodic factor. For the self-rated

version, all items loaded as expected and in line with the design of the scale. The three

factors explained 63% of variance for the observer-rated PAS, and 65% of variance for the

self-rated version of the PAS.

All subscales correlated strongly with the PAS total score for both the observer-rated version

and the self-rated version. For both versions correlations amongst subscales was moderate,

indicating that these subscales measure different dimensions of anxiety (Supplementary

material Appendix 7).

Properties of the scales as diagnostic tests

The diagnostic properties of the PAS were assessed with ROC curves and Youden indeces

(highest sum of specificity and sensitivity) was calculated. For both, the observer-rated and

self-rated scales, the AUC and sensitivity and specificity for any anxiety disorder was high.

The performance of subscales was assessed in relation to specific anxiety disorders: the

persistent subscale with GAD, the episodic subscale with panic disorder and the avoidance

13
scale with social phobia and agoraphobia combined. For each subscale a high AUC was

found, as well as a high sensitivity and specificity at the optimal cut-off (Table 6). The

observer-rated version had a better specificity than the self-rated scale, whereas the self-rated

scale had a better sensitivity than the observer-rated scale.

Discussion

The aim of this study was the development and validation of a new anxiety scale for PD that

would be brief, easy to administer and overcome most of the limitations of existing anxiety

scales. The general design and format was decided upon by consensus during a Delphi

procedure. Division into subscales and item selection was evidence based and guided by

analyses on the database of a published validation study of the HARS, BAI and HADS

{Leentjens, 2011 #1618}. This evidence-based approach is reflected in the satisfactory

clinimetric properties that the new scale shows. Face validity is better than existing scales,

given the division into the clinically evident and relevant symptom domains of persistent

anxiety, episodic anxiety and avoidance behavior. The scale is not biased towards persistent

symptoms, such as is the case with the HARS, or to episodic symptoms, such as is the case

with the BAI. Avoidance behavior is not assessed by existing scales. Properties for

acceptance, distribution, reliability and stability met defined criteria. The concurrent and

known groups validity is good. The scale has a plausible and satisfactory factorial structure,

which is not the case with the BAI and HADS. The AUC and Youden index of the PAS is

higher than that of the HARS, BAI and HADS. In addition, the scale is brief and easy to

administer.

This study also has limitations. Although decisions were based on evidence, judgements were

made. For instance, a decision was made about anchoring the item responses to frequency or

severity of symptoms, or both. For persistent anxiety symptoms, severity is more relevant,

14
and for episodic anxiety frequency. For avoidance behavior, both may be relevant. The

investigators ultimately opted for a dual formulation, incorporating both frequency and

severity because this makes the answering options for all scale items uniform. The authors

believed that patients would know what is relevant for each item. The results of the study

indicate that this way of formulating items does not lead to problems in the clinical use of the

scale and probably did not affect the clinimetric properties. Another potential limitation is the

fact that study samples differed per institution. This reflects a diversity of cultures,

environments, raters, and patients, which may also be regarded as an advantage for a scale to

be used internationally. For practical reasons, researchers administering the MINI to the

patients are the same as those administering the PAS and other anxiety scales, so that the

researcher administering the scales was not blinded for potential clinical diagnoses.

While not a limitation of the present study, not all clinimetric properties of the PAS were

assessed. For instance, the sensitivity to change was not assessed, since this can only be done

longitudinally in an observational or treatment study. This was outside the scope of the

present study.

Conclusion

The PAS has some clear clinimetric advantages over existing anxiety rating scales, but still

requires additional validation, especially in the area of ‘sensitivity to change’. The authors

hope that additional validation will take place and will also involve other research groups and

diverse populations of PD patients. Since the scale does not incorporate specific PD related

questions, validation could also be done in other neurological populations. The authors hope

that this scale will be used routinely in clinical care and research.

References

15
1. Dissanayaka NN, Sellbach A, Matheson S, et al. Anxiety disorders in Parkinson's
disease: prevalence and risk factors. Movement Disorders 2010;15(25):7.
2. Leentjens AFG, Dujardin K, Marsh L, Richard IH, Starkstein SE, Martinez-Martin P.
Symptomatology and markers of anxiety disorders in Parkinson’s disease: a cross-
sectional study. Movement Disorders 2011;26(3):484-492.
3. Menza MA, Robertson-Hoffman DE, Banapace AS. Parkinson's disease and anxiety:
comorbidity with depression. Biological Psychiatry 1993;34:465-470.
4. Pontone GM, Williams JR, Anderson KE, et al. Prevalence of anxiety disorders and
anxiety subtypes in patients with Parkinson's disease. Movement Disorders
2009;24(9):1333-1338.
5. Reijnders JSAM, Ehrt U, Weber WEJ, Aarsland D, Leentjens AFG. A systematic
review of prevalence studies of depression in Parkinson's disease. Movement
Disorders 2008;23(2):183-189.
6. Lauterbach EC, Freeman A, Vogel RL. Correlates of generalized anxiety and panic
attacks in dystonia and Parkinson's disease. Cognitive and Behavioral Neurology
2003;16(4):225-233.
7. Siemers ER, Shekhar A, Quaid K, Dickson H. Anxiety and motor performance in
Parkinson's disease. Movement Disorders 1993;8(4):501-506.
8. Vázquez A, Jiménez-Jiménez FJ, García-Ruiz P, García-Urra D. "Panic attcks" in
Parkinson's disease; a long-term complication of levodopatherapy. Acta Neurologica
Scandinavica 1993;87:14-18.
9. Leentjens AFG, Dujardin K, Marsh L, et al. Anxiety rating scales in Parkinson's
disease: critique and recommendations. Movement Disorders 2008;23(14):2015-2025.
10. Leentjens AFG, Dujardin K, Marsh L, Richard IH, Starkstein SE, Martinez-Martin P.
Anxiety rating scales in Parkinson’s disease: a validation study of the Hamilton
Anxiety Rating Scale, The Beck Anxiety Inventory and the Hospital Anxiety and
Depression Scale. Movement Disorders 2011;26(3):407-415.
11. Martinez-Martin P, Rojo-Abuin JM, Dujardin K, et al. Designing a new scale to
measure anxiety symptoms in Parkinson's disease: item selection based on canonical
correlation analysis. European Journal of Neurology 2013;20:1198-1203.
12. Forjaz MJ, Martinez-Martin P, Dujardin K, et al. Rasch analysis of anxiety scales in
Parkinson's disease. Journal of Psychosomatic Research 2013;74:414-419.
13. De Rijk MC, Rocca WA, Anderson DW, Melcon MO, Breteler MMB, Maraganore
DM. A population perspective on diagnostic criteria for Parkinson's disease.
Neurology 1997;48:1277-1281.
14. Folstein MF, Folstein SE, McHugh PR. Mini Mental State. A practical method for
grading the cognitive state of patients for the clinician. Journal of Psychiatric Research
1975;12:189-198.
15. Fahn S, Elton RL, and members of the UPDRS committee. Unified Parkinson's
Disease Rating Scale. In: Fahn S, Marsden CD, Goldstein M, Calne DB, eds. Recent
developments in Parkinson's Disease. New Jersey: McMillan Health Care, 1987:153-
163.
16. Lawton MP, Brody EM. Assessment of older people: Self-maintaining and
instrumental activities of daily living. Gerontologist 1969;9:179-186.
17. Morris PLP, Robinson RG, Andrezejewski P, Samuels J, Price TR. Association of
depression with 10-year post-stroke mortality. American Journal of Psychiatry
1993;150:124-129.

16
18. Jenkinson C, Fitzpatrick R, Peto V, Greenhall R, Hyman N. The PDQ-8: development
and validation of a shortform Parkinson's disease questionnaire. Psychology and
Health 1997;12:805-814.
19. Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-International
Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured
diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical
Psychiatry 1998;59 Suppl 20:22-33.
20. Hamilton M. A rating scale for depression. Journal of Neurology, Neurosurgery and
Psychiatry 1960;23:56-62.
21. Guy W. ECDEU assessment manual for psychopharmacology - revised (DHEW
publication no. ADM 76-338). Rockville, MD: U.S. Department of Health Education
and Welfare., 1976.
22. Hamilton M. The assessment of anxiety states by rating. British Journal of Medical
Psychology 1959;32:50-55.
23. Beck AT, Epstein N, Brown G, Steer RA. An inventory for measuring clinical anxiety:
psychometric properties. Journal of Consultation and Clinical Psychology
1988;56:893-897.
24. Cicchetti DV. Assessing inter-rater reliability for rating scales: resolving some basic
issues. British Journal of Psychiatry 1976;129:452-456.
25. Smith SC, Lamping DL, Banarjee S, et al. Measurement of health-related quality of
life for people with dementia: development of a new instrument (DEMQOL) and an
evaluation of current methodology.]. Health and Technology Assessment
2005;9(10):16-19.
26. McHorney CA, Tarlov AR. Individual-patient monitoring in clinical practice: are
available health status surveys adequate? Quality of Life Research 1995;4:293-307.
27. Hays RD, Anderson R, Revicki D. Psychometric considerations in evaluating health-
related quality of life measures. Quality of Life Research 1993;2:441-449.
28. Scientific Advisory Committee of the Medical Outcomes Trust. Assessing health
status and quality-of-life instruments: Attributes and review criteria. Quality of Life
Research 2002;11:193-205.
29. Ware JE, Gandek B. Methods for testing data quality, scaling assumptions, and
reliability: the IQOLA Project approach. Journal of Clinical Epidemiology
1998;51:945-952.
30. Eisen M, Ware Jr JE, Donald CA, Brook RH. Measuring components of children's
health status. Med Care 1979;17:902-921.
31. Terwee C, Bot SDM, de Boer RM, et al. Quality criteria were proposed for
measurement properties of health status questionnaires. Journal of Clinical
Epidemiology 2007; 60:34-42.
32. Fitzpatrick R, Davey C, Buxton MJ, Jones DR. Evaluating patient-based outcome
measures for use in clinical trials. Health Technolology Assessment 1998;2:1-74.
33. Leentjens AFG, Dujardin K, Marsh L, et al. Apathy and anhedonia rating scales in
Parkinson's disease: critique and recommendations. Movement Disorders
2008;23(14):2004-2014.

Acknowledgements:

17
This study was sponsored by a grant from the Michael J Fox Foundation for Parkinson

Research (MJFF; www.michaeljfox.org). We would like to thank the patients for their

participation. We would like to thank A.J.H. Moonen, A.S. Carette, J. Czeerniakowski, B.

Leibowitz, M.A. Zea-Sevila, and B. Frades-Payo for their contribution to data collection.

Author roles

1. Research project: A. Conception, B. Organization, C. Execution;

2. Statistical Analysis: A. Design, B. Execution, C. Review and Critique;

3. Manuscript Preparation: A. Writing of the first draft, B. Review and Critique

Leentjens AFG: 1 A, B, C; 2 A, B; 3 A

Dujardin K: 1 C; 2 B, C; 3 B

Pontone G: 1 C; 2 B, C; 3 B

Starkstein SE: 1 C; 2 B, C; 3 B

Weintraub D: 1 C; 2 B, C; 3 B

Martinez-Martin P: 1 C; 2 B, C; 3 B

Full financial disclosure in past year

A.F.G. Leentjens recieves payment from Elsevier Inc as Editor-in-Chief of the Journal of

Psychosomatic Research. He has received research grants from the Michael J Fox Foundation

and the Stichting Internationaal Parkinson Fonds, as well as royalties from Reed-Elsevier, de

Tijdstroom and Van Gorcum publishers.

18
K. Dujardin received research support from the Michael J. Fox Foundation for Parkinson's

Resarch and from France Parkinson.

G.M. Pontone is funded by grants from the National Institute of Health and National Institute

on Aging and received funds to support research from Acadia Pharmaceuticals and the

Michael J. Fox Foundation during the course of this project.

S.E. Starkstein is funded by grants from the National Health Medical and Research Centre

(Australia) and the Michael J. Fox Foundation. He serves as an editorial board member of

International Psychogeriatrics.

D. Weintraub Dr. Weintraub has received research funding from Michael J. Fox Foundation

for Parkinson's Research, National Institutes of Health, Novartis Pharmaceuticals, Department

of Veterans Affairs, and Alzheimer's Disease Cooperative Study; honoraria from Teva

Pharmaceuticals, Lundbeck Inc., Pfizer, Avanir Pharmaceuticals, Merck & Co., UCB, Bristol-

Myers Squibb Company, Novartis Pharmaceuticals, Eli Lilly and Company, Clintrex LLC,

Theravance, CHDI Foundation, and Alzheimer's Disease Cooperative Study; license fee

payments from the University of Pennsylvania for the QUIP and QUIP-RS; and fees for

testifying in court case related to impulse controls disorders in Parkinson's disease

P. Martinez-Martin has received honoraria for lectures or participation in scientific meetings

from Novartis, Britannia, Orion Pharma, and Abbott/AbbVie. He is current holder of grants

for research from Michael J. Fox Foundation, Parkinson UK, and the Spanish Official

Agencies FIS and IMSERSO

19
Table 1: Demographic and disease related data of the study population (continuous and
categorical variables) (n = 360)

Variable Percentage Mean SD range

Age 65.3 8.9 41.0-88.0

Sex (female) 37.3

Years of education 13.6 4.0 3.0-22.0

Marital status

- never married 5.0

- married 81.8

- widowed 3.3

- divorced 9.9

Handedness (right) 92.6

MMSE 28.5 2.1 21.0-30.0

CDR scores median 0.5 0.0-1.0

0 46

0.5 54

1.0 0.3

higher

PDQ8 (standardized) 26 5.4 0.0-84.0

IADL 10.1 4.0 4.0-28.0

HAMD 7.5 5.5 0.0-35.0

HARS 10.0 6.9 0.0-47.0

BAI 11.8 8.6 0.0-46.0

PGI median 2.0 IQR 1.0-3.0 1.0-6.0

20
1 38

2 17

3 23

4 16

5 4

6 2

7 0

CGI median 2.0 IQR 1.0-3.0 1.0-7.0

1 39

2 16

3 24

4 14

5 4

6 2

7 0.6

Major depressive dis 7.2

Dysthymia 8.0

GAD 15.2

Social phobia 9.1

Agoraphobia without 10.5

panic

Panic disorder (total) 3.4

Panic without 1.7

agoraphobia

Panic with 1.7

21
agoraphobia

no. of anxiety disord

0 73

1 17

2 7

3 3

4 or more 0

History depression 37.3

FH depression 24.0

History anxiety 32.3

FH anxiety 14.3

History other 12.7

psychiat

FH other psychiatry 11.3

FH parkinson 25.4

Antidepressant 31.8

Beta-blocker 11.0

Benzodiazepine 21.3

antipsychotic 3.9

PD disease duration 10.4 6.1 1.0-52.0

UPDRS 2 12.7 6.6 1.0-35.0

UPDRS 3 24.7 12.1 3.0-73.0

UPDRS 4 3.6 3.1 0.0-17.0

H&Y

1 17.9

22
2 51.4

3 22.9

3.5 6.1

4 1.1

Anticholinergic 6.6

Dopa-agonist 63.5

Levodopa 90.9

NMDA antagonist 21.3

COMT inhibitor 27.3

Apomorphine 2.2

Other PD medication 4.1

LEDD 811 581 0-3724

DBS

- GPI 0.8

- STN 11.6

pallidotomy 1.1

23
 Table 2. Description of PAS scores

Range, floor and ceiling effects, distribution and acceptability of the observer and selfrated

PAS.

variable Theoret Observ % floor % Mean SD median skewness kurtosis Acceptability

ical ed score ceilin (% missing

range range g scores)

score

Observer-rated PAS

PAS obs total 0-48 0-43 6.9 0 9.4 7.9 7 1.2 1.3 0.003

PAS obs 0-29 0-20 13.3 0 5.5 4.9 4 0.8 -0.2 0

persistent

PAS obs 0-16 0-11 44.2 0 1.7 2.3 1 1.7 2.8 0.003

episodic

PAS obs 0-12 0-10 34.0 0 2.3 2.5 2 1.2 1.1 0

avoidance

PAS selfrated

PAS self total 0-52 0-42 8 0 12.2 8,9 11 0.8 0.1 0

PAS self 0-24 0-20 14.9 0 6.7 4.9 6 0.5 -0.5 0

persistent

PAS self 0-16 0-12 35.1 0 2.6 2.9 2 1.0 0.3 0

episodic

PAS self 0-12 0-12 29.0 1.1 2.3 2.5 2 0.8 0.1 0

avoidance

24
Table 3. Reliability parameters of the full scale and subscales of the observer and self-rated

PAS.

parameter → Cronbach’s alpha mean inter-item mean item-total

scale ↓ correlation (range) correlation (range)

PAS observer-rated

PAS obs total 0.87 0.33 (0.06-0.78) 0.63 (0.35-0.83)

PAS obs persistent 0.91 0.60 (0.54-0.80) 0.85 (0.80-0.90)

PAS obs episodic 0.69 0.48 (0.25-0.52) 0.72 (0.69-0.78)

PAS obs avoidance 0.57 0.39 (0.15-0.47) 0.73 (0.58-0.83)

PAS self-rated

PAS self total 0.89 0.41 (0.18-0.76) 0.663 (0.53-0.82)

PAS self persistent 0.88 0.60 (0.44-0.74) 0.825 (0.76-0.87)

PAS self episodic 0.78 0.47 (0.35-0.59) 0.777 (0.76-0.81)

PAS self avoidance 0.67 0.39 (0.25-0.58) 0.769 (0.65-0.81)

25
Table 4. Inter rater reliability of the PAS, test-retest reliability, standard error of measurement

(SEM) the full scale as well as subscales of the observer-rated and self-rated PAS.

parameter → inter-rater test-retest reliability SEM

scale ↓ reliability (ICC)

(ICC)

PAS observer-rated

PAS obs total 0.92 0.46

PAS obs persistent 0.92 0.30

PAS obs episodic 0.85 0.12

PAS obs avoidance 0.84 0.13

PAS selfrated

PAS self total 0.89 0.51

PAS self persistent 0.85 0.30

PAS self episodic 0.79 0.15

PAS self avoidance 0.75 0.15

26
Table 5. Convergent validity of PAS total scales and subscales with the CGI and PGI for

anxiety, HARS and BAI, and divergent validity with the HAMD (Spearman rank correlation

coefficients . All correlation coefficients are significant (p<0.001).

CGI PGI HARS BAI HAMD

PAS observer

PAS obs total 0.79 0.71 0.68 0.71 0.61

PAS obs persistent 0.79 0.72 0.62 0.67 0.60

PAS obs episodic 0.52 0.51 0.46 0.56 0.39

PAS obs avoidance 0.47 0.38 0.50 0.41 0.37

PAS self-rated

PAS self total 0.77 0.76 0.58 0.71 0.55

PAS self persistent 0.77 0.77 0.57 0.69 0.58

PAS self episodic 0.57 0.55 0.36 0.54 0.33

PAS self avoidance 0.52 0.49 0.45 0.47 0.36

27
 Table 6: ‘ ROC curves’ values for observer-rated and self-rated total scales and subscales.

Anxiety disorders characterized by avoidance are: agoraphobia and social phobia (here taken

together as avoidant anxiety disorders). The Youden index is the highest sum of sensitivity

and specificity. The cut-off score at which the Youden index is reached is the optimal cut-off

score for dichotomisation of patients with and without anxiety disorder. For screening or

diagnosis, higher or lower cut-offs can be selected.

scale grouping variable AUC optimal sensitivity specificity Youden index

(%) cut-off at opt cut- at opt cut-

off off

PAS obs total any anxiety disorder 85.9 13/14 0.71 0.91 1.61

PAS obs persistent generalized anx dis 88.9 9/10 0.76 0.89 1.65

PAS obs epis panic disorder 96.5 3/4 1.00 0.84 1.84

PAS obs avoidance avoidant anx disorders 88.2 3/4 0.81 0.88 1.69

PAS self total any anxiety disorder 85.1 13/14 0.81 0.74 1.54

PAS self persistent generalized anx dis 89.6 10/11 0.89 0.77 1.66

PAS self epis panic disorder 95.6 5/6 1.00 0.86 1.86

PAS self avoidance avoidant anx disorders 85.0 4/5 0.70 0.84 1.54

28
Appendix 1. Development of the Parkinson Anxiety Scale (PAS): Minutes of the Delphi

procedure

On October 14, 2011, an investigator meeting was held in Madrid in which 7 investigators

participated: 5 principal investigators (AL, KD, PMM, GP, and DW) and two affiliated

investigators (Forjaz MJ and Rojo-Abuin JM). During this meeting a modified Delphi

procedure was used to construct the Parkinson Anxiety Scale. In this process use was made of

the personal opinion and expertise of the participants, as well as of previously published and

unpublished data of a prior validation study of anxiety rating scales in Parkinson’s disease

(PD) [1]. A first exploratory Delphi round focused on the general format of the scale,

including potential subscales, item number, as well as scoring issues, including who scores

(patient or researcher/clinician), the range of scoring options and the formulation of Likert

scores. After consensus had been reached, a second Delphi round was held focusing on item

selection and item formulation.

It was agreed that the principal aim of the scale was reliable measurement of the severity of

anxiety symptoms, rather than designing an instrument specifically intended for screening or

diagnosis. It was thought that prediction of ‘caseness’, in case of screening or diagnosis,

would be best served by this approach. Although the scale would be developed specifically

for patients with Parkinson’s disease, the study group agreed that item selection and

formulation would be such that the scale can be administered in patients with other

neurological and physical diseases as well.

29
Participants reached consensus that the scale should consist of three subscales: one pertaining

to persisting anxiety, one pertaining to episodic anxiety, and one pertaining to avoidance

behavior. These first two subscales are supported by evidence from a canonical correlation

(CC) [2]. This analysis revealed that CC of items of the BAI and HARS resulted in two main

components that were interpreted as persistent and episodic anxiety. Persistent anxiety is

present in disorders such as ‘generalized anxiety disorder’, while episodic anxiety is present

in ‘panic disorder’, ‘social phobia’, and ‘specific phobia’. Avoidance behavior was considered

a specific feature of anxiety, since it is listed in the DSM criteria of social phobia, specific

phobia and agoraphobia. However, this feature is not represented in the BAI or HARS, and

was considered as third factor. For research purposes a clinician rated scale was preferred, but

it was agreed that routine clinical practice was probably better served with a self-rated scale.

For this reason the study group decided to validate two versions of the new scale: one patient-

rated and one clinician rated. Since anxiety is characterized by highly personal feelings of

distress that may not be evident to close companions, the study group did not consider a

caregiver rated version. It was agreed that the scale should be concise, but still have a wide

enough representation of symptoms. For this reason it was considered desirable that the total

number of items would be between 10 and 20.

Items will be scored on a five point Likert scale with severity and/or frequency anchors: not at

all/never, very mild/very rarely, mild/sometimes, moderate/frequently and severe/most of the

time. The ‘very mild/very rarely’ anchor was introduced to make the scale more sensitive to

anxiety symptoms in the lower severity range, since Rasch analysis revealed that

items/answers reflecting a lower severity of anxiety were underrepresented in the BAI and

HARS [3].

30
Item selection was partially based on results of CC and Rasch analysis of the BAI and HARS

items [2,3]. Items influenced by age, sex and motor symptoms (such as urogenital symptoms

on the HARS, and tremor) were eliminated, as were items were shown to be not reliable in the

Rasch analysis (such as gastro-intestinal symptoms and insomnia). Items with too little item

difficulty (less than minus 2 logits) in the Rasch analysis were also eliminated (such as

concentration difficulties), as were items that in the CFA did not cluster with recognizable

anxiety symptoms (gastro-intestinal symptoms and insomnia). A second Delphi round was

held to select items for the new scale from the remaining items. Items had to be characteristic

for persistent or episodic anxiety presentations, and located across the item difficulty

spectrum in the Rasch analysis. Five items were chosen for the ‘persistent anxiety’ subscale.

Because ‘depressed mood’ was one of the symptoms most predictive of anxiety severity, it

was decided to keep this item in as well. It was agreed that depressed mood is not a symptom

of anxiety disorders and the item should eventually be deleted. However, allowing the item to

be included during the study phase would enable the study group to assess the influence of

this item on scale performance. Four items were chosen for the ‘episodic anxiety’ subscale.

For the third subscale ‘avoidance behavior’, three types of avoidance behavior were chosen as

items: avoiding public activities, avoiding specific triggers (heights, spiders, etc.), and

avoiding wide or narrowed spaces, reflecting core features of social phobia, specific phobia

and agoraphobia. These were chosen on the basis of consensus and, since they are not

included in the BAI or HARS, not based on evidence.

After item selection is was decided to formulate items as questions, because in this way the

patient-rated and clinician rated version could be most similar textually. The scale will be

formulated in English. Translations in Dutch, French and Spanish will be made for the sake of

the study, which will be backtranslated for inconsistencies in formulation.

31
After construction of the scale, an initial exploratory validation analysis was performed on

BAI and HARS items most closely resembling the items of the new scale, making use of the

database of a prior validation study[1]. This analysis revealed good internal consistency of the

subscales, good representation of scored on subscales across the range of severity of anxiety,

good correlation of the ‘persistent anxiety’ subscale with the diagnosis of ‘generalized anxiety

disorder’, and of the ‘episodic anxiety subscale’ with ‘panic disorder’. These results were

considered satisfying and were celebrated with roast lamb and a superb 2004 Rioja at the

Posada de la Ville in old Madrid.

References

1. Leentjens AFG, Dujardin K, Marsh L, Richard IH, Starkstein SE, Martinez-Martin P.

Anxiety rating scales in Parkinson’s disease: a validation study of the Hamilton
Anxiety Rating Scale, The Beck Anxiety Inventory and the Hospital Anxiety and
Depression Scale. Movement Disorders 2011;26(3):407-415.
2. Martinez-Martin P, Rojo-Abuin JM, Dujardin K, et al. Designing a new scale to
measure anxiety symptoms in Parkinson's disease: item selection based on canonical
correlation analysis. European Journal of Neurology 2013;20:1198-1203.
3. Forjaz MJ, Martinez-Martin P, Dujardin K, et al. Rasch analysis of anxiety scales in
Parkinson's disease. Journal of Psychosomatic Research 2013;74:414-419.

32
Appendix 2. English, Spanish, French and Dutch versions of the PAS

Copyright of this scale and it translations is held by the authors (Leentjens AFG, Dujardin K,

Pontone GM, Starkstein SE, Weintraub D, and Martinez-Martin P; 2014). The scale and its

translations are in the public domain and may be used without additional permission and free

of charge on the condition that it's source is referenced.

The Parkinson Anxiety Scale (PAS); English version

A. Persistent anxiety

Please mark one circle for each item below

In the past four weeks, to what extent did you experience the
following symptoms?

A.1. Feeling anxious or nervous

o Not at all, or never
o Very mild, or rarely
o Mild, or sometimes
o Moderate, or often
o Severe, or (nearly) always

A.2. Feeling tense or stressed

o Not at all, or never
o Very mild, or rarely
o Mild, or sometimes
o Moderate, or often
o Severe, or (nearly) always

A.3. Being unable to relax

o Not at all, or never
o Very mild, or rarely
o Mild, or sometimes

33
 o Moderate, or often
o Severe, or (nearly) always

A.4. Excessive worrying about everyday matters

o Not at all, or never
o Very mild, or rarely
o Mild, or sometimes
o Moderate, or often
o Severe, or (nearly) always

A.5. Fear of something bad, or even the worst, happening

o Not at all, or never
o Very mild, or rarely
o Mild, or sometimes
o Moderate, or often
o Severe, or (nearly) always

B. Episodic anxiety

Please mark one circle for each item below

In the past four weeks, did you experience episodes of the following
symptoms?

B.1. Panic or intense fear

o Never
o Rarely
o Sometimes
o Often
o Nearly always

B.2. Shortness of breath

o Never
o Rarely
o Sometimes
o Often
o Nearly always

34
B.3. Heart palpitations or heart beating fast (not related to physical effort
or activity)
o Never
o Rarely
o Sometimes
o Often
o Nearly always

B.4. Fear of losing control

o Never
o Rarely
o Sometimes
o Often
o Nearly always

C. Avoidance behavior

Please mark one circle for each item below

In the past four weeks, to what extent did you fear or avoid the
following situations?

C.1. Social situations (where one may be observed, or evaluated by others,

such as speaking in public, or talking to unknown people)
o Never
o Rarely
o Sometimes
o Often
o Nearly always

C.2. Public settings (situations from which it may be difficult or

embarrassing to escape, such as queues or lines, crowds, bridges, or public
transportation)
o Never
o Rarely
o Sometimes
o Often
o Nearly always

35
C.3. Specific objects or situations (such as flying, heights, spiders or other
animals, needles, or blood)
o Never
o Rarely
o Sometimes
o Often
o Nearly always

36
 The Parkinson Anxiety Scale (PAS); Spanish version

Escala de Ansiedad en Parkinson

A. Ansiedad rasgo

En las últimas cuatro semanas, ¿en qué grado ha padecido los siguientes
síntomas?

A.1. Sentirse ansioso o nervioso

o Para nada, ninguna vez
o Algo, pero casi nada
o A veces
o Con frecuencia
o Muchas veces, casi siempre

A.2. Sentirse tenso o estresado

o Para nada, ninguna vez
o Algo, pero casi nada
o A veces
o Con frecuencia
o Muchas veces, casi siempre

A.3. Ser incapaz de relajarse

o Para nada, ninguna vez
o Algo, pero casi nada
o A veces
o Con frecuencia
o Muchas veces, casi siempre

A.4. Excesivamente preocupado por problemas cotidianos

o Para nada, ninguna vez
o Algo, pero casi nada
o A veces
o Con frecuencia
o Muchas veces, casi siempre

37
 A.5. Tener miedo de que pase algo malo, o muy malo
o Para nada, ninguna vez
o Algo, pero casi nada
o A veces
o Con frecuencia
o Muchas veces, casi siempre

B. Ansiedad estado

En las pasadas cuatro semanas, ¿ha experimentado episodios con estos

síntomas?

B.1. Pánico o miedo intenso

o Nunca
o Rara vez
o Alguna vez
o Con frecuencia
o Casi siempre

B.2. Quedarse sin aliento

o Nunca
o Rara vez
o Alguna vez
o Con frecuencia
o Casi siempre

B.3. Taquicardias, (no relacionadas con actividad física)

o Nunca
o Rara vez
o Alguna vez
o Con frecuencia
o Casi siempre

B.4. Miedo a perder el control

o Nunca
o Rara vez
o Alguna vez
o Con frecuencia
o Casi siempre

38
 C. Conductas de evitación

En las últimas cuatro semanas, ¿en qué medida ha temido o ha evitado estas
situaciones?

C.1. Situaciones sociales( donde uno puede ser observado y/o evaluado
por los otros, tales como hablar en público, o charlar con un
desconocido)
o Nunca
o Rara vez
o A veces
o A menudo
o Casi siempre

C.2. Sitios públicos con gran afluencia de gente, donde puede ser
complicado escapar; tales como el metro en hora punta, una
manifestación, etc.
o Nunca
o Rara vez
o A veces
o A menudo
o Casi siempre

C.3. Determinados objetos o situaciones (tales como volar en avión, las

arañas, las alturas, etc.)
o Nunca
o Rara vez
o A veces
o A menudo
o Casi siempre

39
The Parkinson Anxiety Scale (PAS) (Version française)

A. Anxiété persistante
Au cours des quatre dernières semaines, dans quelle mesure avez-vous ressenti
les symptômes suivants ? Merci de cocher une réponse pour chacune des
manifestations ci-dessous.

A.1. Vous sentir anxieux ou nerveux

o Pas du tout ou jamais
o Très peu ou rarement
o Légèrement ou parfois
o Modérément ou souvent
o Fortement ou (presque) toujours

A.2. Vous sentir tendu ou stressé

o Pas du tout ou jamais
o Très peu ou rarement
o Légèrement ou parfois
o Modérément ou souvent
o Fortement ou (presque) toujours

A.3. Etre incapable de vous détendre

o Pas du tout ou jamais
o Très peu ou rarement
o Légèrement ou parfois
o Modérément ou souvent
o Fortement ou (presque) toujours

A.4. Vous faire trop de soucis pour des petits problèmes de la vie de tous les
jours
o Pas du tout ou jamais
o Très peu ou rarement
o Légèrement ou parfois
o Modérément ou souvent
o Fortement ou (presque) toujours

A.5. Craindre qu'un malheur, ou même le pire, va se produire

o Pas du tout ou jamais
o Très peu ou rarement
o Légèrement ou parfois
o Modérément ou souvent
o Fortement ou (presque) toujours

40
 B. Anxiété épisodique
Au cours des quatre dernières semaines, avez-vous ressenti par moment les
symptômes suivants ? Merci de cocher une réponse pour chacune des
manifestations ci-dessous.

B.1. Etre pris de panique ou d'une peur intense

o Jamais
o Rarement
o Parfois
o Souvent
o Presque toujours

B.2. Avoir de la difficulté à respirer

o Jamais
o Rarement
o Parfois
o Souvent
o Presque toujours
B.3. Avoir des palpitations ou le cœur qui s'emballe (sans lien avec un effort
ou une activité physique)
o Jamais
o Rarement
o Parfois
o Souvent
o Presque toujours

B.4. Avoir peur de perdre le contrôle

o Jamais
o Rarement
o Parfois
o Souvent
o Presque toujours

C. Comportements d'évitement
Au cours des quatre dernières semaines, dans quelle mesure avez-vous redouté
ou évité les situations suivantes ? Merci de cocher une réponse pour chacune des
manifestations ci-dessous.

41
C.1. Les situations sociales (où on pourrait être observé ou jugé par
d'autres, comme prendre la parole en public ou discuter avec des inconnus)
o Jamais
o Rarement
o Parfois
o Souvent
o Presque toujours

C.2. Les lieux publics (situations dont il peut être difficile ou gênant de
s'échapper, telles que les files d'attente, la foule, les ponts ou les transports
en commun)
o Jamais
o Rarement
o Parfois
o Souvent
o Presque toujours

C.3. Des objets ou des situations spécifiques (comme prendre l'avion, être
en hauteur, les araignées ou d'autres animaux, les piqûres ou la vue du
sang)
o Jamais
o Rarement
o Parfois
o Souvent
o Presque toujours

42
The Parkinson Anxiety Scale (PAS); Dutch version

De Parkinson Angst Scaal (PAS)

A. Persisterende Angst

Kruis één antwoord aan voor elk van de volgende items:

In welke mate heeft u in de afgelopen vier weken de volgende klachten gehad?

A.1. Angst of nervositeit

o helemaal niet of nooit
o in heel geringe mate of zelden
o in geringe mate of soms
o matig ernstig of vaak
o in ernstig mate of (bijna) altijd

A.2. Spanning of stress

o helemaal niet of nooit
o in heel geringe mate of zelden
o in geringe mate of soms
o matig ernstig of vaak
o in ernstig mate of (bijna) altijd

A.3. Zich niet kunnen ontspannen

o helemaal niet of nooit
o in heel geringe mate of zelden
o in geringe mate of soms
o matig ernstig of vaak
o in ernstig mate of (bijna) altijd

A.4. Overmatig piekeren over alledaagse dingen

o helemaal niet of nooit
o in heel geringe mate of zelden
o in geringe mate of soms
o matig ernstig of vaak
o in ernstig mate of (bijna) altijd

A.5. Angst dat er iets heel ergs zal gaan gebeuren

o helemaal niet of nooit
o in heel geringe mate of zelden
o in geringe mate of soms
o matig ernstig of vaak
o in ernstig mate of (bijna) altijd

43
 B. Episodische Angst

Kruis één antwoord aan voor elk van de volgende items:

In welke mate heeft u in de afgelopen vier weken de volgende symptomen

ervaren?

B.1. Paniek of intense angst

o Nooit
o Zelden
o Soms
o Vaak
o Meestal

B.2. Kortademigheid
o Nooit
o Zelden
o Soms
o Vaak
o Meestal

B.3. Hartkloppingen of versnelde hartslag (niet gerelateerd aan lichamelijke inspanning)

o Nooit
o Zelden
o Soms
o Vaak
o Meestal

B.4. Angst voor controleverlies

o Nooit
o Zelden
o Soms
o Vaak
o Meestal

C. Vermijdingsgedrag

Kruis één antwoord aan voor elk van de volgende items:

44
Hoe vaak heeft u in de afgelopen vier weken angst gehad voor de volgende
situaties of deze situaties vermeden?

C.1. Sociale situaties (waarin men door anderen geobserveerd of beoordeeld zou kunnen
worden, zoals spreken in het openbaar of spreken met onbekenden)
o Nooit
o Zelden
o Soms
o Vaak
o Meestal

C.2. Openbare gelegenheden (situaties waarin het moeilijk of gênant kan zijn om te
ontsnappen, zoals ‘in een wachtrij staan’, zich in een mensenmassa bevinden, op
bruggen lopen, of reizen met het openbaar vervoer)
o Nooit
o Zelden
o Soms
o Vaak
o Meestal

C.3. Specifieke objecten of situaties (zoals reizen met het vliegtuig, hoogtes, spinnen of
andere dieren, naalden of bloed)
o Nooit
o Zelden
o Soms
o Vaak
o Meestal

Vertaling: Moonen AJH, Leentjens AFG, 2011

45
Appendix 3: Number of included patients, demographic and disease related data by institute,

as well as the significance of differences of proportions or means across institutes by chi-

square or ANOVA.

Baltimore Perth Lille Maastricht Madrid Philadelphia p

n 60 60 61 61 60 60

Age (years) 65.0 68.0 59.7 66.8 67.2 63.1 0.00

% female 60 57 66 62 73 63 0.45

MMSE (/30) 28.8 28.5 27.4 29.0 28.5 28.5 0.02

IADL 10.4 11.4 8.2 10.7 9.3 10.1 <0.001

Disease duration 12.1 11.4 10.4 10.9 8.6 8.9 0.01

UPDRS 3 24.3 33.8 18.2 30.1 21.0 21.0 0.00

H&Y (median) 2 3 2 2.5 2 2 0.00

Major depression (%) 5 3 2 7 15 12 0.04

Dysthymia 0 0 5 2 33 13 0.00

GAD 5 8 28 2 17 32 <0.001

Social Phobia 8 2 13 2 15 15 0.02

agoraphobia 0 7 20 3 15 18 0.00

Panic (total) 2 0 10 5 2 2 0.29

HAMD 6.9 7.3 5.5 5.6 9.7 9.7 0.00

HARS 13.1 8.7 9.3 5.9 13.0 10.1 0.00

BAI 11.4 11.4 15.5 8.5 12.3 11.5 0.00

PAS obs full 8.9 8.2 15.2 8.5 11.4 11.1 0.00

PAS self full 11.0 12.8 19.1 11.7 11.4 14.5 0.00

Anticholinergic 10 5 3 10 2 10 0.22

Dopa-agonist 53 75 57 63 73 60 0.06

Levo-dopa 93 93 90 92 88 88 0.86

NMDA antagonist 18 0 18 33 33 25 0.00

46
MAO-B inhibitor 35 2 31 11 50 35 0.00

apomorphine 0 2 5 0 3 3 0.35

LEDD 996 530 1044 888 486 811 <0.001

Antidepressant 47 35 20 20 32 38 0.01

Beta blocker 15 10 8 8 3 22 0.03

benzodiazepine 20 3 20 11 42 32 0.00

antipsychotics 7 2 2 0 10 3 0.04

47
Appendix 4. Scores on individual PAS items

Mean (SD), median (IQR) and range of scores on individual items of the observer-rated and

self-rated PAS items.

Observer-rated PAS

item mean SD median range inter-rater

reliability

(ICC)

Persistent subscale

1 1.34 1.20 1 0-4 0.94

2 1.36 1.14 1 0-4 0.95

3 1.05 1.15 1 0-4 0.88

4 1.00 1.18 1 0-4 0.93

5 0.76 1.06 0 0-4 0.92

Episodic subscale

1 0.30 0.70 0 0-3 0.87

2 0.53 0.94 0 0-4 0.87

3 0.35 0.73 0 0-3 0.90

4 0.52 0.82 0 0-3 0.90

Avoidance subscale

1 0.94 1.16 0 0-4 0.91

2 0.91 1.25 0 0-4 0.93

3 0.45 0.95 0 0-4 0.76

Selfrated PAS

48
item mean SD median range Test-retest

reliability

(ICC)

Persistent subscale

1 1.48 1.18 1 0–4 0.89

2 1.60 1.14 2 0–4 0.88

3 1.28 1.19 1 0–4 0.81

4 1.38 1.25 1 0–4 0.85

5 0.96 1.23 1 0–4 0.81

Episodic subscale

1 0.52 0.85 0 0–4 0.80

2 0.79 1.03 0 0–4 0.86

3 0.54 0.85 0 0–3 0.79

4 0.80 0.98 0 0–4 0.73

Avoidance subscale

1 1.10 1.23 1 0-4 0.85

2 1.14 1.30 1 0–4 0.84

3 0.66 1.06 1 0–4 0.79

49
Appendix 5: Known groups validity: mean scores and standard deviations across severity

groups defined by CGI or PGI scores for observer and self-rated scales. For CGI = 7, the

number of included patients in this category is only n=2. For PGI = 7, no patients rated this

severity.

CGI score n PAS total obs PAS pers obs PAS epis PAS avoid

↓/scale → obs obs

1 141 3.9 1.9 0.7 1.3

2 59 6.7 4.1 1.1 1.6

3 85 10.9 6.8 1.7 2.6

4 50 18.0 10.9 3.2 3.9

5 15 23.3 13.4 5.4 4.5

6 8 28.3 15.4 6.5 6.4

7 2 32 16.5 5.0 10.5

p (Kruskal- <0.001 <0.001 <0.001 <0.001

Wallis)

CGI score n PAS total self PAS pers self PAS epis PAS avoid

↓/scale → self self

1 141 5.6 2.9 1.1 1.5

2 59 10.4 5.9 2.2 2.3

3 85 14.6 8.3 3.0 3.2

4 50 21.3 11.7 4.6 5.1

50
5 15 26.2 13.7 6.8 5.7

6 8 31.8 16.3 8.0 7.5

7 2 27.0 11.5 5.0 10.5

p (Kruskal- <0.001 <0.001 <0.001 <0.001

Wallis)

PGI score PAS total obs PAS pers obs PAS epis obs PAS avoid

↓/scale → obs

1 138 4.0 1.9 0.6 1.5

2 60 8.1 4.9 1.5 1.7

3 81 10.5 6.5 1.6 2.6

4 58 16.0 9.9 2.8 3.3

5 16 26.7 14.6 6.3 5.8

6 6 26.2 14.3 6.5 5.3

7 0

p (Kruskal- <0.001 <0.001 <0.001 <0.001

Wallis)

PGI score n PAS total self PAS pers self PAS epis PAS avoid

↓/scale → self self

1 138 5.4 2.8 1.1 1.6

2 60 10.3 6.1 2.2 2.0

51
3 81 15.2 8.5 3.2 3.5

4 58 19.3 10.7 4.1 4.5

5 16 31.3 15.7 8.3 7.3

6 6 28.0 15.2 7.0 5.8

7 0

p (Kruskal- <0.001 <0.001 <0.001 <0.001

Wallis)

52
Appendix 6: Principal Component Analysis of the observer and self-rated PAS versions.

Observer-rated PAS

KMO measure of sampling adequacy: 0.884; total variance explained : 62.6 %

item/loading persistent episodic avoidance

A1 0.86 0.2 0.14

A2 0.83 0.18 0.13

A3 0.76 0.14 0.20

A4 0.83 0.13 0.19

A5 0.77 0.19 0.13

B1 0.45 0.5 0.1

B2 0.12 0.83 0.08

B3 0.16 0.81 0.04

B4 0.46 0.35 0.3

C1 0.40 0.01 0.59

C2 0.33 0.07 0.69

C3 -0.42 0.1 0.74

% variance 33 16 13

explained

53
Self-rated PAS

KMO measure of sampling adequacy: 0.897; total variance explained : 64.9 %

item/loading persistent episodic avoidance

A1 0.8 0.28 0.20

A2 0.8 0.22 0.28

A3 0.79 0.12 0.06

A4 0.75 0.22 0.31

A5 0.63 0.36 0.27

B1 0.26 0.62 0.36

B2 0.20 0.78 -0.01

B3 0.19 0.78 0.14

B4 0.30 0.60 0.44

C1 0.27 0.14 0.75

C2 0.23 0.02 0.83

C3 0.10 0.28 0.54

% variance 27 20 18

explained

54
Appendix 7. Divergent validity of PAS subscales (Spearman correlation coefficients; all

coefficients are significant for P<0.01)

Observer-rated PAS

Total PAS Persistent Episodic Avoidance

subscale subscale subscale

Total PAS 1 0.93 0.74 0.72

Persistent subscale 1 0.55 0.51

Episodic subscale 1 0.35

Avoidance subscale 1

Self-rated PAS

Total PAS Persistent Episodic Avoidance

subscale subscale subscale

Total PAS 1 0.91 0.81 0.77

Persistent subscale 1 0.61 0.55

Episodic subscale 1 0.50

Avoidance subscale 1

55