Cross-sectional comparison of the epidemiology of DSM-5 generalized anxiety disorder across the globe

TL;DR: The findings of this study show that DSM-5 GAD is more prevalent than DSM-IV GAD and is associated with substantial role impairment, particularly in high-income countries, and the public health significance of GAD across the globe is underscore.

Abstract: The WHO WMH Survey Initiative is supported by National Institute of Mental Health (NIMH) grant R01 MH070884; the John D. and Catherine T. MacArthur Foundation; the Pfizer Foundation; US Public Health Service grants R13-MH066849, R01-MH069864, and R01 DA016558; Fogarty International Center grant FIRCA R03-TW006481; the Pan American Health Organization; Eli Lilly and Company; Ortho-McNeil Pharmaceutical, Inc; GlaxoSmithKline; and Bristol-Myers Squibb. The 2007 Australian National Survey of Mental Health and Wellbeing is funded by the Australian Government Department of Health and Ageing. The Sao Paulo Megacity Mental Health Survey is supported by the State of Sao Paulo Research Foundation Thematic Project grant 03/00204-3. The Bulgarian Epidemiological Study of common mental disorders Epidemiology and Bulgaria is supported by the Ministry of Health and the National Center for Public Health Protection. The Chinese WMH Survey Initiative is supported by the Pfizer Foundation. The Shenzhen Mental Health Survey is supported by the Shenzhen Bureau of Health and the Shenzhen Bureau of Science, Technology, and Information. The Colombian National Study of Mental Health is supported by the Ministry of Social Protection. The Mental Health Study Medellin–Colombia was carried out and supported jointly by the Center for Excellence on Research in Mental Health (CES University) and the Secretary of Health of Medellin. The European Study of the Epidemiology of Mental Disorders project is funded by European Commission contracts QLG5-1999-01042, SANCO 2004123, and EAHC 20081308 (the Piedmont Region [Italy]); Fondo de Investigacion Sanitaria; Instituto de Salud Carlos III, Spain grant FIS 00/0028; Ministerio de Ciencia y Tecnologia, Spain, grant SAF 2000-158-CE; Departament de Salut, Generalitat de Catalunya, Spain; Instituto de Salud Carlos III grants CIBER CB06/02/0046 and RETICS RD06/0011 REM-TAP; and other local agencies and by an unrestricted educational grant from GlaxoSmithKline. Implementation of the Iraq Mental Health Survey (IMHS) and data entry were carried out by the staff of the Iraqi Ministry of Health and Ministry of Planning with direct support from the Iraqi IMHS team, with funding from both the Japanese and European Funds through United Nations Development Group Iraq Trust Fund. The Israel National Health Survey is funded by the Ministry of Health with support from the Israel National Institute for Health Policy and Health Services Research and the National Insurance Institute of Israel. The WMH Japan Survey is supported by grants H13-SHOGAI-023, H14-TOKUBETSU-026, and H16-KOKORO-013 for Research on Psychiatric and Neurological Diseases and Mental Health from the Japan Ministry of Health, Labour and Welfare. The Lebanese Evaluation of the Burden of Ailments and Needs of the Nation is supported by the Lebanese Ministry of Public Health, the WHO (Lebanon), National Institute of Health/Fogarty International Center grant R03 TW006481-01, anonymous private donations to Institute for Development Research Advocacy and Applied Care, Lebanon, and unrestricted grants from Algorithm, AstraZeneca, Benta, Bella Pharma, Eli Lilly, GlaxoSmithKline, Lundbeck, Novartis, Servier, Phenicia, and Union Pharmaceutique d'Orient SAL. The Mexican National Comorbidity Survey is supported by The National Institute of Psychiatry Ramon de la Fuente grant INPRFMDIES 4280 and by the National Council on Science and Technology grant CONACyT-G30544-H, with supplemental support from the Pan American Health Organization. Te Rau Hinengaro: The New Zealand Mental Health Survey is supported by the New Zealand Ministry of Health, Alcohol Advisory Council, and the Health Research Council. The Nigerian Survey of Mental Health and Wellbeing is supported by the WHO (Geneva), the WHO (Nigeria), and the Federal Ministry of Health, Abuja, Nigeria. The Northern Ireland Study of Mental Health was funded by the Health & Social Care Research & Development Division of the Public Health Agency. The Peruvian WMH Study was funded by the National Institute of Health of the Ministry of Health of Peru. The Polish project Epidemiology of Mental Health and Access to Care–EZOP Project grant PL 0256 was supported by Iceland, Liechtenstein, and Norway through funding from the European Economic Area Financial Mechanism and the Norwegian Financial Mechanism; the EZOP Project was cofinanced by the Polish Ministry of Health. The Portuguese Mental Health Study was carried out by the Department of Mental Health, Faculty of Medical Sciences, NOVA University of Lisbon, with collaboration of the Portuguese Catholic University, and was funded by Champalimaud Foundation, Gulbenkian Foundation, Foundation for Science and Technology and Ministry of Health. The Romania WMH Survey Initiative study projects Policies in Mental Health Area and National Study Regarding Mental Health and Services Use were carried out by the National School of Public Health & Health Services Management (formerly the National Institute for Research & Development in Health), with technical support of Metro Media Transilvania, the National Institute of Statistics–National Centre for Training in Statistics, Societatea Comerciala Cheyenne Services SRL, and Statistics Netherlands and were funded by the Ministry of Public Health (formerly the Ministry of Health) with supplemental support from Eli Lilly Romania SRL. The South Africa Stress and Health Study is supported by US NIMH grant R01-MH059575 and the National Institute of Drug Abuse, with supplemental funding from the South African Department of Health and the University of Michigan. The Psychiatric Enquiry to General Population in Southeast Spain–Murcia Project has been financed by the Regional Health Authorities of Murcia (Servicio Murciano de Salud and Consejeria de Sanidad y Politica Social) and Fundacion para la Formacion e Investigacion Sanitarias of Murcia. The Ukraine Comorbid Mental Disorders During Periods of Social Disruption study is funded by US NIMH grant R01-MH61905. The US National Comorbidity Survey Replication is supported by NIMH grant U01-MH60220 with supplemental support from the National Institute of Drug Abuse, the Substance Abuse and Mental Health Services Administration, Robert Wood Johnson Foundation grant 044708, and the John W. Alden Trust. Preparation of this manuscript was supported by grant R01 MH094425 from the NIMH (Dr Ruscio).

Summary

Introduction

• Measles cases have increased in all regions since 2017 (1).
• All methods to estimate the burden of unrecorded mortality come with potential biases and sources of uncertainty.
• Here the authors describe an update to the original Simons et al(6) methodology to address four simplifying assumptions of the original methodology.
• This assumption allowed fast computation, but leads to counterintuitive phenomena; specifically, the width of confidence bounds do not scale with burden and negative lower bounds are permitted (though were excluded post hoc).
• Though the explicit correlation between vaccination doses is difficult to know for all countries, numerous studies have shown that second dose opportunities (through routine or SIA doses) tend to underperform nominal coverage values because of inequities in access to care (9–11); i.e. second opportunities are disproportionately delivered to those who received a first dose.

Data

• Annual measles cases and vaccination coverage with the first and second dose of routine measles containing vaccine (MCV) from 1980 to present were taken from the WHO/UNICEF Joint Reporting Form (12).
• Using coverage data as well as vaccination coverage survey data sourced from published and grey literature, WHO and UNICEF estimate the national immunization coverage for timely fist dose of MCV (MCV1) and second dose of MCV (MCV2) administered through routine immunization services (not mass campaigns), updating the entire time series from 1980-previous year (13–15).
• For MCV2, it is calculated among one birth cohort of children at the recommended age for administration of MCV2, per the national immunization schedule.
• Again, vaccination given after the recommended age is not counted towards the estimated MCV2 coverage.
• A list of SIAs conducted between 1980-2019 was provided by WHO based on reporting to WHO-UNICEF annually by member states (16).

Model description

• Similar to the approach described in Simons et al (6), the authors present a 2-stage approach to estimating the burden of measles mortality.
• 𝑆𝑡 𝑎 and 𝐼𝑡 𝑎 are the number susceptible and infected in year t in age one-year age cohort a.
• At each year individuals either 1) remain susceptible and increase in age by one year, 2) are infected and move to the corresponding 𝐼𝑡 𝑎 class, or 3) or are immunized by vaccination and move to an immune class (which is not explicitly modeled).
• 𝐼𝑡 𝑎, where 𝑉𝑎(𝐵𝑡+1, 𝑆𝑡 𝑎) is the impact of vaccination on the susceptible population and is assumed known.
• The changes described above result in an age distribution of infections and the susceptible population that changes year-to-year depending on the attack rate and age-targeted vaccination.

Ensemble of models

• In the above model, Va(Bt+1,Sat), describes the impact of vaccination on the number of susceptible individuals in each age class.
• If the second doses (MCV2 or SIAs) are disproportionately delivered to individuals who have received the first dose (MCV1) and therefore already immune, then the effect of a given coverage level will be reduced relative to the same effect if doses were delivered independently at random (FIGURE 1).
• Depending on the timing of campaigns and the predominant seasonality of measles, the impact of SIAs on case numbers may be observed in the year of the campaign or in the year following the campaign.
• The authors take the mean and quantiles of this distribution as estimates of the mean and confidence interval for the true number of infections in each age cohort and year.

Global sums

• Following Simons et al, for 93 countries that have good surveillance quality the authors did not apply the above state-space model.
• For the remaining countries, the authors apply the state-space model described above to estimate the number of cases in each one-year age cohort in each year.
• For global totals of measles infections and deaths, the authors present 2 totals.
• First, the authors present an optimistic total, where all countries fitted with model that assumes independent doses; this assumes that all second opportunities are delivered independently and at random with respect the receipt of the first routine dose and is the assumption made in the Simons et al (6) model that has been used for the annual updates (1).
• Second, the authors present a total where the estimate for each country comes from the model, selected from the 8 candidates, with the highest likelihood.

Results

• The best fit model for most countries (55/100) assumed that second dose opportunities, through routine MCV2 or SIAs, was correlated with the first dose.
• For 11 of 15 countries that had not yet introduced MCV2, the independent doses model, which here reflects only the impact of SIAs, was the best fit model.
• For 32 out of 100 countries, the best fit model was either the SIA-MCV1 or all doses correlated model; suggesting that the effectiveness of either SIA doses or both SIA and MCV2 doses is lower than the nominal coverage level (FIGURE 2); these countries accounted for 86% of estimated measles cases in the 3 years 2017-2019.
• For 37 countries a single model accounted for >90% of the likelihood weight in the ensemble; for 56 countries, >90% of the likelihood weight was on two models; and only 7 countries required 3 models to account for 90% of likelihood weight (FIGURE 2).
• For 65 the impact of SIAs was realized as reductions in cases in the year after the SIA rather than in the year of the SIA.

Global Burden of Measles

• The mean estimate of the total burden of measles infection using the independent doses model is similar to that from the Simons model (6), which makes the same assumptions about the interaction among vaccine doses.
• Notably, because the new model assumes a binomial error, the width of the confidence interval scales with the mean burden of cases.
• The assumed age distribution in the Simons model changes only in response to routine first-dose MCV coverage.
• The country-specific best-fit estimates for the new model predict that a smaller fraction of cases are under 5 years of age (e.g. in 2017, 80/100 countries were predicted to have a smaller fraction of cases under 5 years than in the Simons model).
• Both the country-specific best-model sum and the Simons model estimate an increase in measles infections and deaths from a minimum mean value in 2017.

Discussion

• Though there have been dramatic reductions in the overall burden of measles disease and mortality over the last 2 decades, the recent resurgence of measles remains a concern.
• The methods the authors have described here result in narrower confidence intervals on estimated measles burden, which highlights that the recent increases are unlikely to be the result of random variation in reporting.
• The authors find that the majority of countries were better fit by models that assume that second dose opportunities (routine and/or SIAs) are correlated with the first dose.
• The formulation of the SIA-MCV1 and all doses correlated models are consistent with a strict interpretation of correlation among first and second dose opportunities; specifically, that second doses are delivered first to those who have received a prior first dose and remaining doses are then randomly distributed to those who have not received a first dose.
• Finally, the authors assume that vaccination coverage, due to all sources, is homogeneous at the sub-national level.

Figures

Table 3. Severity of Role Impairment AssociatedWith 12-Month Generalized Anxiety Disorder by Country Income Level

Full text

Copyright 2017 American Medical Association. All rights reserved.
Cross-sectional Comparison of the Epidemiology of DSM-5
Generalized Anxiety Disorder Across the Globe
Ayelet Meron Ruscio, PhD; Lauren S. Hallion, PhD; Carmen C. W. Lim, MSc; Sergio Aguilar-Gaxiola, MD, PhD;
Ali Al-Hamzawi, MD; Jordi Alonso, MD, PhD; Laura Helena Andrade, MD, PhD; Guilherme Borges, ScD;
Evelyn J. Bromet, PhD; Brendan Bunting, PhD; José Miguel Caldas de Almeida, MD, PhD;
Koen Demyttenaere, MD, PhD; Silvia Florescu, MD, PhD; Giovanni de Girolamo, MD;
Oye Gureje, MD, PhD, FRCPsych; Josep Maria Haro, MD, PhD; Yanling He, MD; Hristo Hinkov, MD, PhD;
Chiyi Hu, MD, PhD; Peter de Jonge, PhD; Elie G. Karam, MD; Sing Lee, MB, BS; Jean-Pierre Lepine, MD;
Daphna Levinson, PhD; Zeina Mneimneh, PhD; Fernando Navarro-Mateu, MD, PhD; José Posada-Villa, MD;
Tim Slade, PhD; Dan J. Stein, FRCP, PhD; Yolanda Torres, MPH; Hidenori Uda, MD; Bogdan Wojtyniak, ScD;
Ronald C. Kessler, PhD; Somnath Chatterji, MD; Kate M. Scott, PhD
IMPORTANCE
Generalized anxiety disorder (GAD) is poorly understood compared with other
anxiety disorders, and debates persist about the seriousness of this disorder. Few data exist
on GAD outside a small number of affluent, industrialized nations. No population-based data
exist on GAD as it is currently defined in DSM-5.
OBJECTIVE To provide the first epidemiologic data on DSM-5 GAD and explore cross-national
differences in its prevalence, course, correlates, and impact.
DESIGN, SETTING, AND PARTICIPANTS Data come from the World Health Organization World
Mental Health Survey Initiative. Cross-sectional general population surveys were carried out
in 26 countries using a consistent research protocol and assessment instrument. A total of
147 261 adults from representative household samples were interviewed face-to-face in the
community. The surveys were conducted between 2001 and 2012. Data analysis was
performed from July 22, 2015, to December 12, 2016.
MAIN OUTCOMES AND MEASURES The Composite International Diagnostic Interview was used
to assess GAD along with comorbid disorders, role impairment, and help seeking.
RESULTS Respondents were 147 261 adults aged 18 to 99 years. The surveys had a weighted
mean response rate of 69.5%. Across surveys, DSM-5 GAD had a combined lifetime
prevalence (SE) of 3.7% (0.1%), 12-month prevalence of 1.8% (0.1%), and 30-day prevalence
of 0.8% (0). Prevalence estimates varied widely across countries, with lifetime prevalence
highest in high-income countries (5.0% [0.1%]), lower in middle-income countries (2.8%
[0.1%]), and lowest in low-income countries (1.6% [0.1%]). Generalized anxiety disorder
typically begins in adulthood and persists over time, although onset is later and clinical course
is more persistent in lower-income countries. Lifetime comorbidity is high (81.9% [0.7%]),
particularly with mood (63.0% [0.9%]) and other anxiety (51.7% [0.9%]) disorders. Severe
role impairment is common across life domains (50.6% [1.2%]), particularly in high-income
countries. Treatment is sought by approximately half of affected individuals (49.2% [1.2%]),
especially those with severe role impairment (59.4% [1.8%]) or comorbid disorders (55.8%
[1.4%]) and those living in high-income countries (59.0% [1.3%]).
CONCLUSIONS AND RELEVANCE The findings of this study show that DSM-5 GAD is more
prevalent than DSM-IV GAD and is associated with substantial role impairment. The disorder
is especially common and impairing in high-income countries despite a negative association
between GAD and socioeconomic status within countries. These results underscore the
public health significance of GAD across the globe while uncovering cross-national
differences in prevalence, course, and impairment that require further investigation.
JAMA Psychiatry. 2017;74(5):465-475. doi:10.1001/jamapsychiatry.2017.0056
Published online March 15, 2017.
Supplemental content
Author Affiliations: Author
affiliations are listed at the end of this
article.
Corresponding Author: Ayelet
Meron Ruscio, PhD, Department of
Psychology, University of
Pennsylvania, The Stephen A. Levin
Building, 425 S University Ave,
Philadelphia, PA 19104
(ruscio@psych.upenn.edu).
Research
JAMA Psychiatry | Original Investigation
(Reprinted) 465
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: by a Universitat Pompeu Fabra User on 12/19/2017

Copyright 2017 American Medical Association. All rights reserved.
U
nderstanding of generalized anxiety disorder (GAD)
has lagged behind understanding of other anxiety
disorders.
1,2
Since its introduction in DSM-III,
3
the GAD
diagnosis has undergone frequent substantial revision, chal-
lenging efforts to accumulate a knowledge base for the disor-
der. Originally, GAD was considered a “wastebasket diagno-
sis” presumed to be associated with fairly modest impairment.
4,5
Although GAD has evolved into a well-defined condition char-
acterized by excessive, uncontrollable worry, the assumption
that GAD is associated with relatively minimal impairment has
persisted in some circles, with the term “worried well” some-
times used to describe affected individuals.
1,6
This assump-
tion may partially account for lower clinical and research atten-
tion to GAD than to other emotional disorders.
7,8
Attention to GAD has been especially limited outside a small
number of industrialized, affluent countries in which nearly all
research on the disorder has been conducted. There is reason
to suspect, however, that GAD may be important not only in af-
fluent countries but also in other parts of the world. In fact, GAD
might be more common and impairing in lower-income coun-
tries given the greater economic and political instability, inse-
curity of access to basic necessities, and uncertainty about the
future that tend to characterize those countries. In line with this
hypothesis, lower socioeconomic status is associated with
greater mental illness within countries
9,10
; similar mecha-
nisms could apply across countries as well.
Alternatively, GAD might be less prevalent and impairing
in lower-income countries. A GAD diagnosis requires worries
to be excessive, and this requirement may be met less often
in countries where worry corresponds to realistic everyday con-
cerns. Furthermore, other disorders have been observed to
manifest more frequently through somatic than cognitive
symptoms in non-Western countries,
11,12
raising the possibil-
ity that GAD, which is centrally defined by a cognitive symp-
tom (worry), may be less common in developing than devel-
oped countries. Finally, cross-cultural research on other chronic
mental disorders suggests that prevalence is often lower and
outcomes better in less developed countries, perhaps due to
greater provision of support by family and community.
13
We explored these competing hypotheses using data from
the World MentalHealth (WMH) SurveyInitiative,
14
a coordinated
series of general population surveys carried out under the aus-
pices of the World Health Organization (WHO). Two features of
the surveys are noteworthy for the present study. First, data were
collected in 26 countries of varying income levels. This diversity
provided an opportunity to isolate universal characteristics of
GAD from characteristics that vary across countries. Second,
by lifting the DSM-IV hierarchy rule prohibiting the diagnosis of
GAD during a mood disorder, the surveys were able to define
GAD using DSM-5 criteria. To our knowledge, these results rep-
resent the first community epidemiologic data on DSM-5 GAD.
Methods
Samples and Procedures
Respondents were 147 261 adults, ranging in age from 18 to 99
years, who participated in the WMH surveys. Data from 29 sur-
veys in 26 countries are included in this report (Table 1). The
surveys were fielded between 2001 and 2012 and had a
weighted mean response rate of 69.5% (eTable 1 in the Supple-
ment). All surveys included nationally or regionally represen-
tative samples of the household population. Sampling and
weighting methods are detailed elsewhere.
15
To summarize re-
sults across surveys, we used World Bank criteria
16
to classify
surveys into 3 country-level income groups: (1) low income and
lower-middle income, (2) upper-middle income, and (3) high
income. We refer to these groups as low, middle, and high in-
come for ease of presentation.
All surveys used the WHO Composite International Diag-
nostic Interview (CIDI),
17
a validated, fully structured, lay-
administered interview. Prior to administration in each coun-
try, the CIDI was translated, back-translated, and harmonized
using standardized procedures.
18
Consistent training and field
quality control procedures were established across countries.
19
Interviews were administered face-to-face in 2 parts. Part 1,
which assessed a core set of mental disorders including GAD,
was administered to all respondents. Part 2, which assessed
additional disorders, was administered to respondents with a
part 1 disorder plus a probability subsample of other respon-
dents. Comorbidity analyses were performed using the part 2
sample, which was weighted to adjust for differential prob-
ability of selection into part 2. All other analyses used the part
1 sample. A human subjects review board or ethics commit-
tee approved the survey protocol in each country (eAppendix
in the Supplement), and all respondents gave informed con-
sent; the mode of consent (written vs oral) varied by survey.
The presence and type of compensation also varied among
surveys.
19
Measures
Generalized Anxiety Disorder
The CIDI was used to assess lifetime, 12-month, and 30-day
GAD. In the US survey, lifetime GAD diagnoses based on the
CIDI had good concordance
20
with diagnoses based on the
clinician-administered Structured Clinical Interview for
DSM-IV (SCID).
21
A clinical reappraisal study including other
WMH surveys did not evaluate GAD in isolation but found
good concordance between CIDI and SCID diagnoses of any
12-month anxiety disorder, including GAD.
22
Following the
Key Points
Question What are the patterns and correlates of DSM-5
generalized anxiety disorder throughout the world?
Findings In general population surveys of approximately 150 000
adults in 26 countries, DSM-5 generalized anxiety disorder has a
combined lifetime prevalence of 3.7%, 12-month prevalence of
1.8%, and 30-day prevalence of 0.8%. The disorder is significantly
more prevalent and impairing in high-income countries than in
low- or middle-income countries.
Meaning DSM-5 generalized anxiety disorder is more prevalent
than DSM-IV generalized anxiety disorder and is associated with
considerable role impairment, especially in high-income countries,
underscoring its public health significance.
Research Original Investigation Comparison of the Epidemiology of DSM-5 Generalized Anxiety Disorder
466 JAMA Psychiatry May 2017 Volume 74, Number 5 (Reprinted) jamapsychiatry.com
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: by a Universitat Pompeu Fabra User on 12/19/2017

Copyright 2017 American Medical Association. All rights reserved.
lead of prior studies that modified the CIDI GAD algorithm to
evaluate specific diagnostic changes,
20,23,24
we generated
DSM-5 GAD diagnoses by removing the DSM-IV hierarchical
exclusion of a GAD diagnosis when symptoms occur exclu-
sively during a mood disorder.
25
Age of onset (AOO) of GAD
was assessed using probing methods that have been demon-
strated to improve dating accuracy.
26
Persistence of GAD
was estimated indirectly by calculating the proportion of
respondents with 12-month GAD among those with the life-
time disorder.
Impairment
Respondents with 12-month GAD were administered an ex-
panded version of the Sheehan Disability Scale
27
to assess role
impairments caused by the disorder during the worst month
in the year before the interview. Respondents rated the ex-
tent of interference with home management, work, close re-
lationships, and social life on separate 0-to-10 scales. We
grouped ratings into categories of absent (0), mild (1-3), mod-
erate (4-6), and severe (7-10) impairment for analysis. We also
assessed the number of days in the past 12 months during which
respondents reported being totally unable to work or carry out
usual activities because of GAD.
Comorbid Disorders
Mental disorders other than GAD were diagnosed using DSM-IV
criteria. They included other anxiety disorders (ie, panic dis-
order, agoraphobia, social phobia, specific phobia, posttrau-
Table 1. Prevalence of DSM-5 Generalized Anxiety Disorder in the World Mental Health Surveys
Country
Prevalence, No. (%) [SE]
a
Total
Sample,
No.
Lifetime
Prevalence
12-mo
Prevalence
30-d
Prevalence
12-mo Prevalence
Among Lifetime
Cases
Low income 596 (1.6) [0.1] 345 (0.9) [0.1] 186 (0.5) [0.1] 345 (59.5) [2.4] 36 498
Colombia 84 (1.9) [0.3] 39 (1.0) [0.2] 17 (0.4) [0.1] 39 (53.8) [7.7] 4426
Iraq 220 (5.0) [0.6] 131 (3.0) [0.4] 82 (2.0) [0.3] 131 (61.3) [4.4] 4332
Nigeria 8 (0.1) [0] 1 (0.0) [0] 0 1 (32.9) [24.0] 6752
Peru 40 (1.1) [0.1] 17 (0.5) [0.1] 2 (0.1) [0] 17 (44.2) [6.7] 3930
PRC Beijing/Shanghai 60 (1.0) [0.1] 36 (0.6) [0.1] 15 (0.3) [0.1] 36 (61.1) [7.1] 5201
PRC Shenzhen 19 (0.2) [0.1] 14 (0.1) [0.1] 0 14 (76.9) [11.3] 7132
Ukraine 165 (3.3) [0.3] 107 (2.1) [0.2] 70 (1.3) [0.1] 107 (63.1) [3.7] 4725
Middle income 875 (2.8) [0.1] 507 (1.6) [0.1] 231 (0.7) [0.1] 507 (56.1) [2.1] 28 927
Brazil 280 (5.1) [0.4] 187 (3.3) [0.3] 100 (1.8) [0.2] 187 (63.4) [3.8] 5037
Bulgaria 125 (2.3) [0.2] 57 (1.2) [0.1] 23 (0.5) [0.1] 57 (49.5) [4.8] 5318
Colombia (Medellín) 127 (3.8) [0.5] 73 (2.1) [0.3] 31 (0.9) [0.2] 73 (56.2) [5.6] 3261
Lebanon 71 (2.3) [0.3] 47 (1.5) [0.3] 22 (0.6) [0.2] 47 (64.5) [7.5] 2857
Mexico 78 (1.1) [0.2] 44 (0.6) [0.1] 19 (0.3) [0.1] 44 (48.8) [7.7] 5782
Romania 27 (1.0) [0.3] 9 (0.2) [0.1] 8 (0.2) [0.1] 9 (24.5) [8.9] 2357
South Africa 167 (3.6) [0.4] 90 (1.9) [0.3] 28 (0.6) [0.1] 90 (53.5) [4.8] 4315
High income 4417 (5.0) [0.1] 2031 (2.3) [0.1] 850 (0.9) [0] 2031 (45.9) [0.9] 81 836
Australia 710 (8.0) [0.5] 312 (3.6) [0.3] 128 (1.5) [0.2] 312 (45.4) [2.7] 8460
Belgium 75 (2.8) [0.5] 27 (0.9) [0.3] 10 (0.3) [0.1] 27 (31.2) [7.1] 2419
France 190 (6.2) [0.5] 61 (2.1) [0.3] 20 (0.6) [0.2] 61 (33.7) [3.7] 2894
Germany 58 (1.5) [0.2] 22 (0.5) [0.1] 6 (0.2) [0.1] 22 (34.8) [7.5] 3555
Israel 216 (4.4) [0.3] 148 (3.1) [0.3] 50 (1.1) [0.2] 148 (70.7) [3.1] 4859
Italy 100 (2.1) [0.3] 28 (0.6) [0.1] 9 (0.2) [0.1] 28 (28.4) [5.2] 4712
Japan 105 (2.6) [0.3] 53 (1.2) [0.2] 8 (0.2) [0.1] 53 (47.6) [5.0] 4129
New Zealand 1084 (7.9) [0.3] 441 (3.1) [0.2] 187 (1.2) [0.1] 441 (38.5) [2.0] 12 790
Northern Ireland 334 (6.4) [0.4] 150 (2.8) [0.3] 82 (1.5) [0.2] 150 (44.2) [3.2] 4340
Poland 90 (0.9) [0.1] 52 (0.5) [0.1] 25 (0.3) [0.1] 52 (58.4) [5.1] 10 081
Portugal 269 (6.1) [0.5] 145 (3.3) [0.3] 45 (1.1) [0.2] 145 (53.0) [3.2] 3849
Spain 131 (1.9) [0.2] 59 (0.8) [0.2] 33 (0.4) [0.1] 59 (43.8) [6.3] 5473
Spain (Murcia) 193 (7.0) [0.9] 111 (4.3) [0.7] 77 (3.0) [0.6] 111 (61.3) [3.3] 2621
The Netherlands 110 (3.6) [0.4] 28 (1.0) [0.2] 13 (0.4) [0.1] 28 (28.2) [5.8] 2372
United States 752 (7.8) [0.3] 394 (4.0) [0.2] 157 (1.6) [0.2] 394 (52.1) [1.8] 9282
All countries 5888 (3.7) [0.1] 2883 (1.8) [0.1] 1267 (0.8) [0] 2883 (48.8) [0.8] 147 261
Comparison between
individual countries
b
85.8 42.7 29.0 6.7
Comparison between
country income groups
c
311.3 106.5 21.7 20.4
Abbreviation: PRC, People’s Republic
of China.
a
The ratio of numerator to
denominator numbers does not
equal the reported percentages
because the percentages are
weighted.
b
χ
2
28
Test of homogeneity for
variation in prevalence estimates
across countries; all P < .001.
c
χ
2
2
Test of homogeneity for variation
in prevalence estimates across low-,
middle-, and high-income countries;
all P < .001.
Comparison of the Epidemiology of DSM-5 Generalized Anxiety Disorder Original Investigation Research
jamapsychiatry.com (Reprinted) JAMA Psychiatry May 2017 Volume 74, Number 5 467
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: by a Universitat Pompeu Fabra User on 12/19/2017

Copyright 2017 American Medical Association. All rights reserved.
matic stress disorder, and separation anxiety disorder with on-
set either in childhood or adulthood), mood disorders (ie, major
depression and bipolar spectrum disorder
28
), disruptive be-
havior disorders (ie, oppositional defiant disorder, conduct dis-
order, intermittent explosive disorder, attention-deficit/
hyperactivity disorder, bulimia nervosa, and binge eating
disorder), and substance-related disorders (ie, alcohol and drug
abuse and dependence).
Other Measures
Sociodemographic variables included respondent age, sex,
marital status, educational level, household income (strati-
fied into quartiles within country), and employment status at
the time of the interview. Treatment seeking was assessed by
asking whether respondents received treatment for any men-
tal health or substance-related problem during the past 12
months. Treatment in 4 sectors was probed: specialty mental
health, general medical health, human services, and comple-
mentary-alternative medicine.
Statistical Analysis
Cross-tabulations were used to estimate prevalence, comor-
bidity, impairment, and treatment. Logistic regression and sur-
vival analysis were used to examine sociodemographic corre-
lates. The actuarial method
29
was used to generate AOO
survival curves. Design-based SEs were estimated using the
Taylor series linearization method
30
implemented in SUDAAN
version 11.0
31
to adjust for data weighting and clustering. Sta-
tistical significance was evaluated using Wald and McNemar
χ
2
tests. All tests were 2-sided and used a significance thresh-
old of P < .05. Data analysis was performed from July 22, 2015,
to December 12, 2016.
Results
Prevalence
Across all surveys, the combined lifetime prevalence of GAD
was 3.7%, 12-month prevalence was 1.8%, and 30-day
prevalence was 0.8% (Table 1). Lifetime prevalence esti-
mates varied widely across countries, ranging from less
than 1% of the populations of Nigeria and Shenzhen, China,
to approximately 8% of the populations of Australia, New
Zealand, and the United States. Prevalence increased with
economic development: lifetime estimates (SE) were lowest
in low-income (1.6% [0.1%]), moderate in middle-income
(2.8% [0.1%]), and highest in high-income (5.0% [0.1%])
countries.
Sociodemographic Correlates
Around the world, being female, younger than 60 years, and
unmarried (previously married or never married) were asso-
ciated with GAD (eTable 2 in the Supplement). In addition, GAD
was found disproportionately in respondents with lower edu-
cational levels, lower household income, and Other employ-
ment status (mostly unemployed or disabled). These corre-
lates were relatively modest in magnitude (odds ratio [OR], 1.1;
95% CI, 1.0-1.3 to OR, 1.8; 95% CI, 1.7-2.0) except for the mark-
edly increased odds of GAD among younger cohorts (OR, 3.0;
95% CI, 2.7-3.3 to OR, 6.0; 95% CI, 5.1-7.0). The pattern of cor-
relates was similar across country groups, although GAD was
associated less consistently with educational level, house-
hold income, and employment status in middle- and low-
income countries.
Onset and Course
Generalized anxiety disorder typically begins in adulthood
(eFigure in the Supplement). Onset before puberty was rare
in these surveys, with only 5% of cases developing by age
13 years. There was a gradual, steady accumulation of new
cases over the lifespan, with 25% of all cases emerging by
25 years, 50% of the cases emerging by 39 years, and 75%
of the cases emerging by 53 years. Although the AOO distri-
bution was shifted earlier for high-income countries (me-
dian AOO, 36 years) relative to middle- and low-income
countries (median AOO, 43 years for both), the distributions
were substantively similar across country groups (χ
2
2
=1.9,
P = .38).
Consistent with conceptualizations of GAD as a chronic dis-
order, nearly half of all lifetime cases still had the disorder in
the 12 months before the interview (Table 1). Generalized anxi-
ety disorder was more persistent in low-income (59.5%) and
middle-income (56.1%) countries than in high-income coun-
tries (45.9%). Around the world, persistence was higher for ear-
lier-onset GAD cases, for individuals with lower educational
levels and family income, and for those not employed out-
side the home (ie, Other status, homemaker) (eTable 2 in the
Supplement).
Comorbidity
Most individuals with lifetime (81.9% [0.7%]) and 12-month
(70.8% [1.2%]) GAD had 1 or more comorbid DSM-IV/CIDI dis-
orders (Table 2). The odds were highest for lifetime mood and
anxiety disorders, lower for disruptive behavior disorders, and
lowest for substance-related disorders. The single most com-
mon comorbid condition was major depressive disorder, which
was found in 52.6% (0.9%) of lifetime cases and 40.9% (1.3%)
of 12-month cases of GAD worldwide.
Role Impairment
Respondents with 12-month GAD reported a mean (SE) of
41.2 days (2.4) out of role due to GAD in the past year. Half
(50.6%) of the 12-month cases reported severe functional
impairment resulting from GAD (Table 3). The rate of severe
impairment was lower yet still substantial (35.3%) among
individuals with GAD who had no comorbid disorders
(eTable 3 in the Supplement).
The proportion of participants with severe GAD-related
impairment was highest in high-income (54.5%), lower in
middle-income (42.6%), and lowest in low-income (39.0%)
countries. Generalized anxiety disorder was a more disabling
disorder in some countries than others, with severe impair-
ment reported by a small minority of persons with GAD in
China (17.3% in Shenzhen, 21.8% in Beijing/Shanghai) and
Mexico (28.7%) but by a large majority of those with GAD in
the Netherlands (80.3%) and Romania (82.3%). Neverthe-
Research Original Investigation Comparison of the Epidemiology of DSM-5 Generalized Anxiety Disorder
468 JAMA Psychiatry May 2017 Volume 74, Number 5 (Reprinted) jamapsychiatry.com
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: by a Universitat Pompeu Fabra User on 12/19/2017

Copyright 2017 American Medical Association. All rights reserved.
less, the proportion of severely impaired persons was sizable
in most countries (median, 48.6%; interquartile range,
39.4%-56.2%).
Treatment
Approximately half (49.2%) of respondents with 12-month GAD
reported receiving some form of mental health treatment dur-
ing the previous year (Table 4). Treatment was sought dispro-
portionately by those with comorbid disorders (55.8%) but was
also sought by approximately one-third (32.4%) of respon-
dents with GAD alone (eTable 4 in the Supplement). The treat-
ment rate was higher in high-income (59.0%) than in middle-
income (29.1%) or low-income (21.7%) countries. However,
across countries, the overall pattern was of increasing use of
services, particularly specialty mental health services, with in-
creasing impairment due to GAD.
Discussion
The present findings shed new light on patterns of GAD around
the globe. First, we showed that diagnostic changes from
DSM-IV to DSM-5 yielded an influx of new GAD cases, as life-
time prevalence estimates reported herein are 37% to 90%
higher than published estimates for DSM-IV GAD in the United
States.
32,33
In the current WMH surveys, lifetime prevalence
is 37% higher and 12-month prevalence is 50% higher for DSM-5
than DSM-IV GAD (Ruscio AM, Hallion LS, Demyttenaere K,
Lee S, Lim CCW. Generalized anxiety disorder. In: Scott KM,
de Jonge P, Stein DJ, Kessler RC, eds. The Cross-National Epi-
demiology of Mental Disorders: Facts and Figures From the World
Mental Health Surveys. Cambridge, England: Cambridge Uni-
versity Press; in press). Although an increase in prevalence was
Table 2. Comorbidity of DSM-5 Generalized Anxiety Disorder With DSM-IV Mental Disorders Across Countries
a
DSM-IV Disorder
Lifetime GAD 12-mo GAD
No. (%) [SE]
b
OR (95% CI)
c
No. (%) [SE]
d
OR (95% CI)
e
Anxiety disorder
Panic disorder 699 (12.3) [0.7] 9.8 (8.6-11.2) 328 (11.8) [0.9] 15.8 (13.3-18.8)
Agoraphobia 506 (9.4) [0.5] 7.0 (6.1-8.0) 246 (9.0) [0.7] 11.3 (9.3-13.7)
Social phobia 1303 (26.1) [0.9] 9.2 (8.4-10.2) 561 (22.6) [1.2] 12.3 (10.7-14.0)
Specific phobia 1171 (25.6) [0.8] 4.4 (4.0-4.8) 546 (25.2) [1.2] 5.6 (4.9-6.4)
Posttraumatic stress disorder 1208 (21.0) [0.7] 9.2 (8.4-10.2) 415 (14.4) [0.8] 10.8 (9.4-12.4)
Childhood-onset separation anxiety disorder
f
201 (8.6) [0.7] 4.9 (4.0-6.0)
Adult-onset separation anxiety disorder 390 (15.5) [0.8] 6.1 (5.3-7.1) 94 (6.9) [0.8] 8.2 (6.2-10.7)
Any anxiety disorder 2929 (51.7) [0.9] 8.0 (7.4-8.7) 1257 (44.0) [1.2] 9.3 (8.5-10.3)
Mood disorder
Major depressive disorder 3055 (52.6) [0.9] 10.6 (9.7-11.4) 1173 (40.9) [1.3] 16.3 (14.6-18.2)
Bipolar spectrum disorder
g
544 (11.4) [0.6] 7.6 (6.7-8.7) 267 (10.8) [0.8] 11.5 (9.6-13.8)
Any mood disorder
h
3657 (63.0) [0.9] 13.4 (12.3-14.4) 1467 (51.1) [1.3] 19.6 (17.6-21.8)
Disruptive behavior disorder
Oppositional defiant disorder 190 (10.7) [0.9] 4.4 (3.5-5.4) 34 (3.5) [0.7] 8.9 (5.5-14.2)
Conduct disorder 142 (7.8) [0.8] 4.3 (3.4-5.4) 11 (1.0) [0.3] 3.6 (1.8-7.2)
Intermittent explosive disorder 342 (15.0) [1.1] 5.7 (4.7-6.8) 149 (12.1) [1.4] 7.1 (5.5-9.3)
Attention-deficit/hyperactivity disorder 158 (7.5) [0.8] 4.6 (3.7-5.8) 68 (5.8) [0.9] 8.3 (6.0-11.5)
Bulimia nervosa 115 (3.8) [0.4] 5.8 (4.5-7.5) 33 (2.4) [0.5] 9.5 (6.0-15.1)
Binge eating disorder 150 (6.1) [0.6] 4.4 (3.5-5.5) 53 (4.3) [0.8] 6.8 (4.6-10.1)
Any disruptive behavior disorder 566 (10.1) [0.6] 4.0 (3.5-4.4) 122 (8.1) [0.8] 6.2 (5.2-7.4)
Substance-related disorder
Alcohol abuse 1004 (19.6) [0.8] 2.5 (2.2-2.7) 149 (5.8) [0.6] 2.9 (2.3-2.6)
Alcohol dependence 490 (8.9) [0.5] 4.6 (3.9-5.3) 106 (4.4) [0.6] 6.0 (4.5-8.1)
Drug abuse 459 (10.5) [0.6] 3.9 (3.4-4.5) 71 (3.6) [0.6] 6.9 (4.8-9.8)
Drug dependence 245 (5.3) [0.4] 5.9 (4.8-7.2) 52 (2.5) [0.5] 10.5 (6.8-16.2)
Any substance-related disorder 1141 (22.5) [0.8] 2.7 (2.5-3.0) 221 (9.1) [0.8] 3.9 (3.2-4.7)
Any mental disorder 4627 (81.9) [0.7] 12.3 (11.2-13.6) 1998 (70.8) [1.2] 14.9 (13.2-16.7)
Abbreviations: GAD, generalized anxiety disorder; OR, odds ratio.
a
The ratio of numerator to denominator numbers does not equal the reported
percentages because the percentages are weighted.
b
Proportion of respondents with lifetime DSM-5 GAD who also qualified for the
lifetime DSM-IV disorder in each row.
c
Based on separate logistic regression models using lifetime GAD to predict
each lifetime comorbid disorder; all significant at P < .001.
d
Proportion of respondents with 12-month DSM-5 GAD who also qualified for
the 12-month DSM-IV disorder in each row.
e
Based on separate logistic regression models using 12-month GAD to predict
each 12-month comorbid disorder; all significant at P < .001.
f
The surveys did not include a 12-month assessment of childhood-onset
separation anxiety disorder.
g
Includes bipolar I disorder, bipolar II disorder, or subthreshold bipolar disorder
as defined by Merikangas et al.
28
h
Includes major depressive episode and bipolar spectrum disorder.
Comparison of the Epidemiology of DSM-5 Generalized Anxiety Disorder Original Investigation Research
jamapsychiatry.com (Reprinted) JAMA Psychiatry May 2017 Volume 74, Number 5 469
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: by a Universitat Pompeu Fabra User on 12/19/2017

Frequently asked questions

Q1. What have the authors contributed in "Cross-sectional comparison of the epidemiology of dsm-5 generalized anxiety disorder across the globe" ?

Ayelet Meron Ruscio, PhD, Lauren S. Hallion, PhD ; Carmen C. W. Scott, PhD this paper 

Q2. What was the common comorbid condition in the study?

The single most common comorbid condition was major depressive disorder, which was found in 52.6% (0.9%) of lifetime cases and 40.9% (1.3%) of 12-month cases of GAD worldwide. 

Q3. What was the prevalence of GAD among the young people?

The prevalence of DSM-5 GAD was concentrated among individuals who were female, younger than 60 years, unmarried, not employed, less educated, and less affluent relativeto national standards. 

Q4. What is the funding for the study of mental health in northern Ireland?

The Northern Ireland Study of Mental Health was funded by the Health & Social Care Research & Development Division of the Public Health Agency. 

Q5. How many people with GAD met the criteria for life?

52,53 Once GAD begins, it often persists: nearly half of the individuals with lifetime GAD met 12-month criteria for the disorder. 

Q6. How many people with severe role impairment are sought for treatment?

Treatment is sought by approximately half of affected individuals (49.2% [1.2%]), especially those with severe role impairment (59.4% [1.8%]) or comorbid disorders (55.8% [1.4%]) and those living in high-income countries (59.0% [1.3%]). 

Q7. How many days did individuals with GAD have to be unable to work?

individuals reported a mean of more than 40 days in the past year when they were completely unable to work or carry out daily activities because of GAD. 

Q8. What is the main purpose of the Iraqi Mental Health Survey?

Implementation of the Iraq Mental Health Survey (IMHS) and data entry were carried out by the staff of the Iraqi Ministry of Health and Ministry of Planning with direct support from the Iraqi IMHS team, with funding from both the Japanese and European Funds through United Nations Development Group Iraq Trust Fund. 

Q9. What are the three income groups that were used to classify the surveys?

To summarize results across surveys, the authors used World Bank criteria16 to classify surveys into 3 country-level income groups: (1) low income and lower-middle income, (2) upper-middle income, and (3) high income. 

Q10. How many people with GAD received treatment during the past year?

GAD was associated with significant help seeking, with approximately half of the individuals with 12-month GAD receiving treatment during the past year. 

Q11. What was the prevalence of generalized anxiety disorder in the past year?

Approximately half (49.2%) of respondents with 12-month GAD reported receiving some form of mental health treatment during the previous year (Table 4). 

Q12. What did the authors do to modify the CIDI GAD algorithm?

lead of prior studies that modified the CIDI GAD algorithm to evaluate specific diagnostic changes,20,23,24 the authors generated DSM-5 GAD diagnoses by removing the DSM-IV hierarchical exclusion of a GAD diagnosis when symptoms occur exclusively during a mood disorder. 

Q13. What is the importance of a systematic assessment of GAD?

Given the frequent occurrence of GAD during mood episodes and the implications of comorbidity for treatment,36 these findings underscore the importance of systematic assessment and appropriate management of GAD in patients with mood disorders.