Social Psychiatry and Psychiatric Epidemiology

https://doi.org/10.1007/s00127-017-1475-9

ORIGINAL PAPER

Twelve-month prevalence rates of mental disorders and service use

in the Argentinean Study of Mental Health Epidemiology
Juan Carlos Stagnaro1 · Alfredo H. Cía2 · Sergio Aguilar Gaxiola3 · Néstor Vázquez4 · Sebastián Sustas4 ·
Corina Benjet5 · Ronald C. Kessler6

Received: 4 September 2017 / Accepted: 17 December 2017

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
Purpose  Community surveys of mental disorders and service use are important for public health policy and planning. There
is a dearth of information for Latin America. This is the first representative community survey in the Argentinean popula-
tion. The purpose is to estimate the 12-month prevalence and severity of mental disorders, socio-demographic correlates
and service use in a general population survey of adults from urban areas of Argentina.
Methods  The World Mental Health Composite International Diagnostic Interview was administered to 3927 individuals
aged 18 years and older participating in a multistage clustered area probability household survey. The response rate was 77%.
Results  The 12-month prevalence of any disorder was 14.8%, and a quarter of those disorders were classified as severe.
Younger participants and those with lower education had greater odds of any disorder and most classes of disorder. 11.6%
of the total population received treatment in the prior 12 months and only 30.2% of those with a severe disorder. Women
and those never married were more likely to receive or seek treatment, whereas those with low and low-average education
were less likely.
Conclusion  Most individuals with a mental disorder in the past year, even those with a severe disorder, have not received
treatment. Because low education is a barrier to treatment, initiatives aimed at mental health education might help timely
detection and treatment of these disorders in Argentina.

Keywords  Epidemiology · Psychiatric disorders · Argentina · Service use · Treatment

Introduction

General population surveys from the World Mental Health

* Juan Carlos Stagnaro Surveys Initiative across the globe estimate that 4–26% of
[email protected] the adult population experience a mental disorder in any
1
Department of Psychiatry and Mental Health, School given year and only 1–15% of the total population receive
of Medicine, University of Buenos Aires, Buenos Aires, any treatment in the same time period [1]. As would be
Argentina expected, service use is greater for those with more severe
2
Anxiety Clinic and Research Center, Buenos Aires, disorders, and yet the mental health treatment gap contin-
Argentina ues to be a public health challenge even among those cases
3
Center for Reducing Health Disparities, University with the greatest severity [1]. Such wide variations in cross
of California, Davis School of Medicine, Sacramento, CA, national estimates require an understanding of treatment
USA needs and service use in differing regions with unique pub-
4
Department of Public Health, School of Medicine, University lic health policies and resources as well as cultural attitudes
of Buenos Aires, Buenos Aires, Argentina toward mental disorders and help seeking. In Latin America,
5
Department of Epidemiology and Psychosocial Research, four countries from the World Mental Health Surveys have
National Institute of Psychiatry Ramón de la Fuente Muñiz, provided 12-month prevalence estimates of mental disor-
Mexico City, Mexico
ders, Brazil, Colombia, Mexico and Peru [2–5]. These coun-
6
Department of Health Care Policy, Harvard Medical School, tries present a more narrow range of 12-month prevalence
Boston, USA

13
Vol.:(0123456789)
 Social Psychiatry and Psychiatric Epidemiology

estimates from 12% for Mexico to 26% for Brazil. The treat- adult inhabitants in the country. Data collection was con-
ment gap is significant in Latin America as only 5–14% of ducted between February and June 2015. 5,082 individu-
the total population in these countries received treatment and als were randomly selected with the aforementioned eight
only a quarter to a third of severe cases of mental disorder geographic regions making up the first-stage sampling units,
received treatment. census areas within each geographic region comprising the
Until now, no community 12-month prevalence and ser- second-stage sampling units and five to seven randomly
vice use estimates for common mental disorders in Argen- selected households within each census area comprising the
tina have been available for public health planning. These third-stage sampling unit. The fourth and final sampling unit
estimates are important for public health policy and plan- was one randomly selected per household.
ning, because mental disorders have been shown to account The research protocol and procedures were approved
for an important proportion of the global burden of disease by the Ethics Committee of the University of Buenos
and to be costly both to the individual and society [6–8]. Aires Medical School and have therefore been performed
Argentina is the second largest country in Latin America in accordance with the ethical standards laid down in the
with the highest Human Development Index of the region 1964 Declaration of Helsinki and its later amendments and
[9] and the greatest number of psychologists per capita in the are consistent with the procedures followed in the WMH
world (198 psychologists per 1000,000 inhabitants in Argen- Surveys [13]. After reading the study objectives to the par-
tina compared to 57/100,000 in Finland, 30/100,000 in the ticipants and informing them that their participation was
USA, 11/100,000 in Colombia or 2/100,000 in Mexico) [10, voluntary and confidential, the interviewer answered all
11]. The large number of psychologists may be due to a long questions before seeking written informed consent. All inter-
and fervent tradition of psychoanalysis in Argentina which views were conducted face to face using computer-assisted
has permeated the general culture. Despite many psycholo- personal interviewing (CAPI) methods by trained lay inter-
gists in the country, there is a dearth of official statistics viewers in the selected respondents´ homes. Interview length
on mental health services, resources and expenditures [10]. was approximately 2 h.
Mental health is treated within the three healthcare sectors As in earlier WMH surveys, the survey was adminis-
of the country, the public sector (which covers 38% of the tered in two parts [14]. Part I, which was administered to all
population and consists of public hospitals and primary respondents, included assessments of core mental disorders,
healthcare centers), the private sector, and the social secu- while Part II was administered to a probability subsample of
rity sector called Obras Sociales with minimal coordination 2116 Part I respondents consisting of all those with a Part I
between them [12]. mental disorder and a randomly selected subsample of other
Therefore because of the lack of information on the prev- Part I respondents. Part II focused on correlates of disorders
alence of mental disorders and use of services to treat them and disorders of secondary interest. The Part II sample was
in Argentina, and the many similarities (language, culture), weighted to adjust for the undersampling of Part I non-cases
but also striking differences (economic development, num- so that prevalence estimates in the weighted Part II sample
ber of psychologists), between Argentina and other Latin were equivalent to those in the Part I sample.
American countries for which there are data, the aim of this
article is to estimate the 12-month prevalence and severity of Instrument
mental disorders, socio-demographic correlates and service
use to address these disorders in a general population survey Disorder assessment
of adults from urban areas of Argentina; these data fill a gap
for epidemiologic information for this region of the world. Mental disorders experienced in the last 12 months were
evaluated with the World Mental Health Composite Inter-
national Diagnostic Interview (WMH-CIDI) [15], a fully
Methods structured diagnostic interview previously used in the
World Mental Health Surveys Initiative, including the
Sample and procedures Spanish-speaking participating Latin American countries.
Diagnoses made with the WMH-CIDI have shown accept-
The Argentinean Study of Mental Health Epidemiology is able to good concordance with clinician diagnoses [16].
a representative household survey that used a multistage Disorders were assessed using the diagnostic criteria of
probability sampling design to represent the population aged the Diagnostic and Statistical Manual of Mental Disorders,
18 years and older with a fixed residence living in one of Fourth Edition (DSM-IV) [17]. Disorders were grouped as
the eight largest metropolitan areas of the country (Buenos follows: mood disorders (i.e., major depressive disorder,
Aires, Córdoba, Corrientes-Resistencia, Mendoza, Neuquén, bipolar I and II disorder and dysthymia), anxiety disorders
Rosario, Salta and Tucumán), covering roughly 50.1% of the (i.e., panic disorder, agoraphobia without panic disorder,

13
Social Psychiatry and Psychiatric Epidemiology

social phobia, specific phobia, separation anxiety disorder, Socio‑demographic correlates

obsessive–compulsive disorder, generalized anxiety disor-
der and post-traumatic stress disorder), substance use dis- The WMH-CIDI evaluated the following socio-demographic
orders (i.e., alcohol and drug abuse and dependence) and correlates: sex, age, income (classified as low, low-average,
disruptive behavior disorders (i.e., oppositional-defiant high-average and high), marital status (classified as married
disorder, conduct disorder, attention deficit hyperactivity or cohabitating, previously married (separated/divorced/wid-
disorder and intermittent explosive disorder). owed) and never married) and education (classified as low,
low-average, high-average and high).

Disorder severity Data analysis

Disorders were classified as mild, moderate or severe, Data were weighted to adjust for differential probabilities
according to the criteria previously implemented in the of selection within and between households chosen and to
World Mental Health Surveys [1]. For a mental disorder match sampling distributions to population distributions.
to be considered severe, it had to meet one of the follow- Part II samples were additionally weighted for the under-
ing criteria: (1) the presence of a bipolar I disorder, (2) sampling of Part  I respondents without core disorders.
substance dependence with a physiological dependence As a result of this complex sample design and subsequent
syndrome, (3) a suicide attempt in conjunction with any weighting, estimates of standard errors of proportions were
other mental disorder, or (4) reporting of at least two areas obtained by the Taylor Series Linearization Method [19]
with severe impairment as determined by a score of 7 or using SUDAAN release 8.0.1 for Windows [20]. Prevalence
higher on the Sheehan Disability Scales [18]. Respondents and service use in the prior 12 months were estimated as the
not ascertained as having a severe disorder were classified proportion of respondents who had a disorder and consulted
as moderate if they reported moderate impairment in any a professional during that period of time. Standard errors of
domain (i.e., a score of 4 or higher on any Sheehan Dis- estimates were obtained using the Jackknife Repeated Repli-
ability Scale), or if the respondent had substance depend- cation (JRR) [21] method implemented in a SAS macro [22].
ence without physiological dependence syndrome. All Logistic regression equations were conducted to estimate the
other disorders studied were classified as mild. socio-demographic correlates of disorder and service use.

Service use Results

The WMH-CIDI survey assessed service use by first deter- The response rate was 77% for a total sample of 3,997
mining if respondents sought attention for emotional, participants. The most common reasons for non-response
nervous, mental, or substance use problems in the prior included being absent or not at home when interviewers vis-
12 months from a long list of professionals. The type of ited (10.7%) and refusal to participate (9.3%).
service provider was classified into healthcare sector and
non-healthcare sector professionals. The healthcare sector Prevalence and severity of mental disorders
was further classified into mental health professionals and
general medical professionals. Mental health professionals The 12-month prevalence of any mental disorder was 14.8%,
consisted of psychiatrists and other mental health profes- with 10.7% having exactly one disorder, 2.4% two disorders
sionals such as psychologists, counselors, psychothera- and 1.6% three or more disorders. The 12-month prevalence
pists, mental health nurses and social workers in a mental of each diagnostic category and each individual disorder is
health specialty setting. General medical professionals shown in Table 1. The most common diagnostic category
consisted of family physicians, general practitioners and was anxiety disorders (9.4%), followed by mood disorders
other medical doctors, such as cardiologists, or gynecolo- (5.7%) and substance use disorders (2.4%), and the least
gists (for women) and urologists (for men), nurses, occu- common was disruptive behavior disorders (0.5%). The most
pational therapists, or other healthcare professionals. The common individual disorder was specific phobia (4.8%), fol-
non-healthcare sector consisted of human services and lowed by major depressive disorder (3.8%), obsessive–com-
complementary alternative medicine such as religious or pulsive disorder (2.5%) and bipolar disorder (2.0%). The
spiritual advisors, Internet use, self-help groups and any 12-month prevalence of a severe disorder was 3.7%.
other healers or alternative therapy like curanderos, a chi- The proportion of mild, moderate and severe disorders by
ropractor or a spiritualist. type of disorder is presented in Table 2. Twenty-five percent
of all disorders were classified as severe, 35.9% as moderate

13
 Social Psychiatry and Psychiatric Epidemiology

Table 1  Twelve-month prevalence of DSM-IV disorders with a greater number of disorders had a higher proportion
12-month prevalence
of severe disorders.

% SE
Disorder severity and service utilization
Anxiety disorders
 Panic disorder 0.8 0.2 The prevalence of any treatment and treatment from differ-
 Generalized anxiety disorder 1.5 0.2 ent sectors and by disorder severity is shown in Table 3. Of
 Social phobia 1.3 0.2 the total population, 11.6% received any treatment in the
 Specific phobia 4.8 0.3 prior 12 months, primarily in the healthcare sector (10.9%),
 Agoraphobia without panic 0.3 0.1 and within this sector, mostly by mental health professionals
 Post-traumatic stress disorder 1.1 0.2 (8.0%). Only 1.7% of the total population received treat-
 Adult separation anxiety disorder 0.6 0.1 ment in the non-healthcare sector. Those with mild disorders
 Obsessive–compulsive disorder 2.5 0.8 received less treatment (22.1%) than those with moderate
 Any anxiety disorder 9.4 0.5 (32.5%) or severe disorders (30.2%). A small number of
Mood disorders those not meeting the 12-month DSM-IV criteria for any
 Dysthymia 0.4 0.1 disorder also received some treatment (8.7%).
 Major depressive disorder 3.8 0.4
 Bipolar disorder (I and II) 2.0 0.2 Socio‑demographic correlates of 12‑month
 Any mood disorder 5.7 0.6 disorders
Disruptive behavior disorders
 Oppositional-defiant disorder 0.0 0.0 Table 4 presents the socio-demographic correlates (i.e., sex,
 Conduct disorder 0.1 0.1 age, income, marital status and education) of meeting cri-
 Attention deficit disorder 0.4 0.1 teria for any 12-month disorder and each class of disorder
 Intermittent explosive disorder 0.1 0.1 (i.e., mood, anxiety, disruptive and substance disorders).
 Any disruptive behavior disorder 0.5 0.1 While sex was not associated with meeting criteria for any
Substance disorders disorder, women had almost twice the odds of an anxiety
 Alcohol abuse 1.5 0.2 disorder (OR = 1.98; 95% CI = 1.32–3.0) and reduced odds
 Alcohol dependence 0.3 0.1 of a disruptive behavior (OR = 0.39; 95% CI = 0.22–0.68)
 Drug abuse 1.0 0.2 and substance use disorder (OR = 0.19; 95% CI = 0.10–0.36).
 Drug dependence 0.4 0.1 Younger age was associated with any disorder and to each
 Any substance use disorder 2.4 0.3 class of disorder with ORs ranging from 1.9 for the odds of
Any disorder an anxiety disorder among the youngest group aged 18–34 to
 Any 14.8 0.9 a high of 34.1 for the odds of a disruptive behavior disorder
 0 disorders 85.2 0.9 among the youngest group. Low and low-average educa-
 1 disorder 10.7 0.8 tion was associated with greater odds of any disorder, any
 2 disorders 2.4 0.3 mood and any anxiety disorder with ORs ranging from 1.6
 3+ disorders 1.6 0.3 for the odds of any disorder among those with low-average
Severity education to 3.1 for the odds of any mood disorder among
 Severe 3.7 0.4 those with low education. Income and marital status was not
 Moderate 5.3 0.5 associated with 12-month disorders.
 Mild 5.8 0.6
Correlates of 12‑month service use
Part I total sample size = 3927; Part II total sample size = 2116

We show in Table 5 the socio-demographic correlates of

and 39.1% as mild. Mood disorders had the greatest propor- the past 12-month treatment in the total population. Women
tion of severe disorders (38.6%), followed by substance use were more likely than men to receive/seek any treatment
disorders (35.8%) and disruptive behavior disorders (28.3%), (OR = 1.72; 95% CI = 1.18–2.51) as were the never mar-
and with the lowest proportion of severe disorders, the anxi- ried more likely than the married or cohabitating to receive/
ety disorders (22.9%). Individual disorders with the high- seek any treatment (OR = 1.64; 95% CI = 1.05–2.56). Those
est severity were drug dependence and oppositional-defiant with high-average income were less likely than those with
disorder (100%), followed by alcohol dependence (65.7%), high income to receive/seek any treatment (OR = 0.58; 95%
bipolar disorder (55.7%), drug abuse (50.2%), post-traumatic CI = 0.36–0.92). Similarly, respondents with low and low-
stress disorder (41.9%) and social phobia (41.3%). Those average education levels were less likely to receive/seek

13
Social Psychiatry and Psychiatric Epidemiology

Table 2  Proportion of mild, Severity

moderate and severe disorders
by individual disorder and Mild Moderate Severe
disorder class
% SE % SE % SE

Anxiety disorders
 Panic disorder 29.4 6.6 44.1 9.2 26.5 11.2
 Generalized anxiety disorder 39.3 7.8 38.5 7.3 22.2 5.7
 Social phobia 24.7 6.7 34.0 8.3 41.3 9.3
 Specific phobia 30.2 4.1 36.2 4.4 25.8 5.2
 Agoraphobia without panic 17.6 10.2 52.5 13.0 29.8 14.2
 Post-traumatic stress disorder 21.0 8.5 37.2 6.5 41.9 12.1
 Adult separation anxiety 22.3 9.8 34.5 10.5 35.3 9.3
 Obsessive–compulsive disorder 64.1 10.2 12.7 7.0 23.1 9.9
 Any anxiety disorder 40.3 4.8 36.8 4.1 22.9 3.4
Mood disorders
 Dysthymia 26.8 12.8 37.8 12.7 35.4 14.4
 Major depressive disorder 16.3 3.0 53.5 4.9 30.2 5.0
 Bipolar disorder (I and II) 9.1 3.4 35.2 6.2 55.7 5.7
 Any mood disorder 14.5 2.3 46.9 3.5 38.6 3.7
Disruptive behavior disorders
 Oppositional-defiant disorder 0.0 0.0 0.0 0.0 100.0 0.0
 Conduct disorder 83.5 19.5 0.0 0.0 16.5 19.5
 Attention deficit disorder 58.8 11.4 9.8 5.1 31.4 8.1
 Intermittent explosive disorder 0.0 0.0 0.0 0.0 0.0 0.0
 Any disruptive behavior disorder 64.0 10.1 7.7 4.0 28.3 8.1
Substance disorders
 Alcohol abuse 49.0 10.9 18.2 5.3 32.8 10.5
 Alcohol dependence 0.0 0.0 34.3 19.2 65.7 19.2
 Drug abuse 35.1 16.5 14.7 7.2 50.2 17.8
 Drug dependence 0.0 0.0 0.0 0.0 100.0 0.0
 Any substance use disorder 45.8 10.4 18.4 4.6 35.8 10.4
Any disorder
 Any 39.1 3.0 35.9 2.8 25.1 2.8
 0 disorders 0.0 0.0 0.0 0.0 0.0 0.0
 1 disorder 49.7 3.8 34.6 2.8 15.7 2.4
 2 disorders 14.7 3.4 45.2 6.3 40.1 6.8
 3+ disorders 4.3 2.2 30.4 6.7 65.4 7.2

Part I total sample size = 3927; Part II total sample size = 2116

Table 3  Prevalence of treatment Treatment Severe Moderate Mild None All

by disorder severity and
treatment sector % SE % SE % SE % SE % SE

Healthcare 27.8 3.6 31.3 4.8 21.0 5.1 8.2 0.8 10.9 0.8
 General medical 10.8 1.9 10.5 2.2 12.3 3.4 2.4 0.4 3.8 0.4
 Mental health 21.5 4.0 23.5 4.0 11.3 4.1 6.2 0.7 8.0 0.7
Non-healthcare 4.2 1.7 2.2 1.3 3.7 2.4 1.4 0.4 1.7 0.4
Any treatment 30.2 3.7 32.5 4.8 22.1 5.2 8.7 0.8 11.6 0.9
No treatment 69.8 3.7 67.5 4.8 77.9 5.2 91.3 0.8 88.4 0.9

Non-healthcare includes human services and complementary alternative medicine

Evaluated only on Part II sample; total Part II sample size, n = 2116

13
 Social Psychiatry and Psychiatric Epidemiology

Table 4  Socio-demographic correlates of any and each class of disorder in the prior 12 months
Any disorder Any mood disorder Any anxiety disorder Any disruptive disorder Any substance dis-
order
OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI

Sex
 Females 1.25 0.91 1.72 1.45 0.96 2.20 1.98 1.32 2.98 0.39 0.22 0.68 0.19 0.10 0.36
 Males 1.00 – – 1.00 – – 1.00 – – 1.00 – – 1.00 – –
 χ2/DF/Prob 2.11 1.00 0.15 3.40 1.00 0.07 11.78 1.00 0.01 11.71 1.00 0.01 29.15 1.00 0.01
Age
 18–34 3.31 2.05 5.35 3.70 1.81 7.56 1.93 1.16 3.23 34.11 5.20 224.01 7.01 2.45 20.10
 35–49 2.62 1.73 3.97 3.20 1.62 6.32 2.32 1.45 3.71 10.03 0.89 113.23 2.54 1.04 6.20
 50–64 2.32 1.13 4.74 1.85 0.86 4.00 2.35 1.05 5.26 2.96 0.15 59.06 1.00 – –
 65+ 1.00 – – 1.00 – – 1.00 – – 1.00 – – – – –
 χ2/DF/Prob 40.96 3.00 0.01 24.03 3.00 0.01 13.48 3.00 0.01 31.03 3.00 0.00 14.40 2.00 0.01
Income
 Low 1.40 0.94 2.07 1.28 0.75 2.19 1.47 0.92 2.33 0.68 0.09 5.16 2.36 0.74 7.49
 Low average 0.99 0.66 1.49 0.75 0.44 1.28 1.04 0.52 2.11 1.28 0.10 17.42 2.36 0.75 7.42
 High average 0.70 0.48 1.04 0.70 0.39 1.24 0.73 0.43 1.22 0.36 0.03 4.19 1.45 0.42 4.95
 High 1.00 – 1.00 – – 1.00 – – 1.00 – – 1.00 – –
 χ2/DF/prob 14.05 3.00 0.01 11.72 3.00 0.01 9.81 3.00 0.02 6.14 3.00 0.11 3.31 3.00 0.35
Marital status
 Married/ 1.00 – – 1.00 – – 1.00 – – 1.00 – – 1.00 – –
cohabiting
 Sep/wid/ 1.38 0.97 1.96 1.53 1.00 2.36 1.07 0.73 1.57 1.94 0.50 7.55 3.26 0.96 11.06
divorced
 Never married 1.26 0.93 1.71 1.15 0.74 1.78 1.27 0.84 1.93 0.56 0.15 2.04 1.36 0.60 3.29
 χ2/DF/prob 5.10 2.00 0.08 4.32 2.00 0.12 1.42 2.00 0.49 2.12 2.00 0.35 3.99 2.00 0.14
Education
 Low 2.06 1.36 3.12 1.89 1.19 3.01 2.15 1.35 3.41 5.04 0.59 43.40 1.55 0.56 4.30
 Low average 1.61 1.01 2.57 1.23 0.70 2.15 2.05 1.11 3.80 3.03 0.13 71.11 0.70 0.19 2.54
 High average 1.37 0.97 1.93 1.21 0.68 2.14 1.48 0.92 2.38 2.87 0.84 9.84 0.65 0.29 1.45
 High 1.00 – – 1.00 – – 1.00 – – 1.00 – – 1.00 – –
 χ2/DF/prob 13.80 3.00 0.01 10.19 3.00 0.02 12.56 3.00 0.01 6.59 3.00 0.09 3.90 3.00 0.27
Overall
 χ2/DF/Prob 976.82 13.00 0.01 1298.0 13.00 0.01 2866.8 13.00 0.01 10,030 13.00 0.01 737.07 12.00 0.01

Evaluated only on Part II sample; sample size for Part II is 2116

*To make the logistic model more stable, the age categories 50–64 and 65 + were collapsed for any substance disorder such that the reference
category was 50 and more

treatment (OR = 0.50; 95% CI = 0.28–0.92 and OR = 0.50; mental health surveys across 63 countries, but within the
95% CI = 0.30–0.85, respectively). interquartile range of 12.3% − 24.3% [23]. Unfortunately,
the greatest majority of those with disorder did not receive
any treatment; only 22.1% of those with a mild disorder and
Discussion 30.2% of those with severe disorders received treatment and
mostly in the healthcare sector by mental health profession-
Nearly one in seven (14.8% of) Argentinean adults in urban als. While moderate and severe disorders were more likely to
areas have experienced a mental disorder in the year prior receive treatment than mild disorders, suggesting rationality
to the Argentinean Study of Mental Health Epidemiology of service allocation, this also shows the need for improving
and a quarter of those disorders may be considered severe. timely detection such that disorders are treated when they are
This is slightly lower than the 17.6% pooled 12-month preva- mild to prevent greater severity or complications of disor-
lence of common mental disorders in a meta-analysis of 174 ders over time [24, 25]. Of those that did not meet criteria

13
Social Psychiatry and Psychiatric Epidemiology

Table 5  Socio-demographic correlates of 12-month treatment among Like Mexico and Colombia, and unlike many other parts
the total population of the world, mental disorders in Argentina were treated
Any Treatment more by mental health specialists than in general medical
practice [3, 4, 28]. This supports an emerging pattern for
OR 95% CI
Latin American countries and is at odds with the World
Sex Health Organization recommendation to treat most mental
 Females 1.72 1.18 2.51 disorders in primary care, thus giving access to a greater
 Males 1.00 – – number of people and for primary care to be a gatekeeper
 χ2/DF/prob 8.67 1.00 0.01 to more specialized services for more severe cases [29]. In
Age Argentina, though the gatekeeper approach is incorporated
 18–34 0.54 0.26 1.11 into stated public health policy, it is inconsistently imple-
 35–49 0.88 0.52 1.49 mented [12]. Very few people sought attention in the non-
 50–64 1.12 0.50 2.51 healthcare sector in Argentina.
 65+ 1.00 – – The twelve-month prevalence estimates for any mental
 χ2/DF/Prob 7.36 3.00 0.06 disorder and each class of disorder in Argentina were con-
Income sistent with the other Latin American countries. For exam-
 Low 0.63 0.40 1.01 ple, nearly 15% of Argentineans experienced any disorder,
 Low average 0.60 0.35 1.05 similar to 13.5% of Peruvians and 12.2% of Mexicans and
 High average 0.58 0.36 0.92 lower than 26% of Brazilians and 20% of Colombians [2–5].
 High 1.00 – – The diagnostic category for which Argentina differed most
 χ2/DF/prob 6.93 3.00 0.07 from the other Latin American countries was the disrup-
Marital status tive behavior disorders which were present in only 0.5% of
 Sep/widowed/divorced 1.02 0.61 1.69 Argentineans versus 1.6% of Mexicans, 3.5% of Peruvians,
 Never married 1.64 1.05 2.56 4.2% of Brazilians and 4.4% of Colombians [2–5].
 Married/cohabiting 1.00 – – We found that younger and less educated individuals had
 χ2/DF/prob 6.07 2.00 0.05 a higher prevalence of disorders. Low education may be
Education associated with disorders, in that early-onset disorders have
 Low 0.50 0.28 0.92 been shown to impact school dropout [30–32]. Women were
 Low average 0.50 0.30 0.85 found to have a higher prevalence of anxiety disorders and a
 High average 0.76 0.49 1.18 lower prevalence of substance use and disruptive behavior
 High 1.00 – – disorders compared to men as has been reported previously
 χ2/DF/prob 15.11 3.00 0.01 [33, 34].
Overall Not all correlates of disorder are correlates of treatment
 χ2/DF/prob 1043.9 13.00 0.01 use. The Behavioral Model of Health Services Use posits
Evaluated only on Part II sample; sample size for Part II is 2116
that treatment use is determined by predisposition to use
services (which can be demographic, social and beliefs),
factors which enable or impede service use (such as health
for any disorder, a small number (8.7%) also received treat- policy, financing and organization) and need for care [35].
ment. Whether this represents a misallocation of treatment We focused on individual predisposing demographic factors
resources or can be explained by subthreshold cases receiving and found women and those who have never been married
early attention or those who no longer meet criteria receiving more likely to have sought or received treatment and those
follow-up care cannot be determined from this study. Due to with low or low-average education less likely to have sought
the large number of psychologists in the country, and particu- or received treatment. These are consistent with results
larly psychoanalytic psychologists, this group might represent found in a report of 17 WMH countries, in which women
individuals in psychoanalysis for the purpose of self-actualiza- were found to use services more than men in 10 countries
tion [26]. In a prior study of 23 WMH countries, it was found (with no gender differences in the remaining), greater educa-
that of those receiving 12-month treatment, 52% met past year tion was related to greater service use in 3 countries, being
criteria for a mental disorder, an additional 18% for a lifetime married was associated with greater use in 5 countries and
disorder (but not in the prior 12 months) and an additional income was positively related to service use in 3 countries
13% had other indicators of need such as multiple subthresh- and negatively in 1 [28]. Sex differences in service utili-
old disorders, recent stressors or suicidal behaviors, leaving zation for mental disorders have been well documented in
almost 16% in treatment with no discernable need [27]. other studies as well [36–38]. It has been suggested that
women seek treatment more because they may be more

13
 Social Psychiatry and Psychiatric Epidemiology

likely to identify mental health symptoms or perceive less Compliance with ethical standards 
stigma related to mental disorders or treatment [28]. Results
from other studies on marital status are less consistent. In Conflict of interest  In the past 3 years, Dr. Kessler received support for
a systematic review of help-seeking behavior among indi- his epidemiological studies from Sanofi Aventis; was a consultant for
Johnson & Johnson Wellness and Prevention, Sage Pharmaceuticals,
viduals with major depression out of 15 studies, 4 found that Shire, Takeda; and served on an advisory board for the Johnson &
being married was negatively associated with help seeking Johnson Services Inc. Lake Nona Life Project. Kessler is a co-owner
and 1 study the opposite [39]. Though speculative, those of DataStat, Inc., a market research firm that carries out healthcare
who have never been married may be more likely to seek research. On behalf of all authors, the corresponding author states that
none of the other authors have conflicts of interest.
treatment because of a lack of social support which foments
outside help seeking. That low education, but not income,
is related to lower likelihood of treatment in Argentina may
be due to lack of information regarding mental disorders References
or negative attitudes toward treatment rather than a lack of
financial resources. Given that low education is related to 1. Demyttenaere K, Bruffaerts R, Posada-Villa J et al (2004) Preva-
both a greater prevalence of disorder and simultaneously less lence, severity, and unmet needs for treatment of mental disorders
treatment suggests greater health disparities in this group. in the World Health Organization World Mental Health Survey.
JAMA 291(21):2581–2590
This study only addressed receiving treatment, but whether 2. Andrade LH, Wang YP, Andreoni S et al (2012) Mental disorders
the treatment received is evidence-based treatment or mini- in megacities: findings from the São Paulo megacity mental health
mally adequate requires further study. survey, Brazil. PLoS One 7(2):e31879
These findings should be considered in light of the limi- 3. Posada-Villa J, Rodriguez M, Duque P et al (2008) Mental disor-
ders in Colombia: results from the World Mental Health Survey.
tations of the study. Because this survey is representative In: Kessler RC, Ustun TB (eds) The WHO World Mental Health
only of the largest urban metropolitan areas, the results can- Surveys: global perspectives on the epidemiology of mental dis-
not be generalized to individuals in rural or other areas of orders. Cambridge University Press, New York, pp 131–143
Argentina. Participants were limited to those with a per- 4. Medina-Mora ME, Borges G, Lara C, Benjet C, Blanco J, Fleiz C,
Villatoro J, Rojas E, Zambrano J (2005) Prevalence, service use,
manent residence, and therefore the prevalence may be and demographic correlates of 12-month DSM-IV psychiatric dis-
underestimated by the exclusion of homeless, hospitalized orders in Mexico: results from the Mexican National Comorbidity
or institutionalized individuals. Finally, the cross-sectional Survey. Psychol Med 35(12):1773–1784
design of the study precludes inferences regarding causality 5. Piazza M, Fiestas F (2014) Prevalencia anual de trastornos y uso
de servicios de salud mental en el Perú: resultados del estudio
or directionality of the socio-demographic correlates and mundial de salud mental, 2005. Rev Peru Med Exp Salud Pública
their association with disorders and treatment. 31(1):30–38
Despite these limitations, we report the first representative 6. Global Burden of Disease Study 2013 Collaborators (2015)
estimates of 12-month mental disorder and treatment use in Global, regional and national incidence, prevalence, and years
lived with disability for 301 acute and chronic conditions and
the general adult population of urban Argentina with the same injuries for 188 countries, 1990–2013: a systematic analy-
methodology as the World Mental Health Surveys, thus allow- sis for the global burden of disease study 2013. The Lancet
ing for comparisons with other countries within and outside 386(9995):743–800
of Latin America as well as providing data for local public 7. Alonso J, Petukhova M, Vilagut G et  al (2011) Days out of
role due to common physical and mental conditions: results
health planning and policies. These data suggest an impor- from the WHO World Mental Health surveys. Mol Psychiatry
tant treatment gap in the population; a greater use of treat- (16):1234–1246
ment in specialized care rather than primary care indicates 8. Schofield D, Shrestha RN, Percival R, Passey ME, Callander EJ,
the need for greater training of general medical practitioners Kelly SJ (2011) The personal and national costs of mental health
conditions: impacts on income, taxes, government support pay-
in the detection and care of common mild disorders; greater ments. BMC Psychiatry 11:72
disparities for those with low levels of education suggest that 9. United National Development Programme (UNDP) (2013) Human
initiatives aimed at mental health literacy might help timely development report 2013: the rise of the south: human progress
detection and treatment of these disorders in Argentina. in a diverse world. http://hdr.undp.org/sites/default/files/Country-
Profiles/ARG.pdf. Accessed Mar 2017
10. World Health Organization (2014) Mental health atlas 2014.
Acknowledgements  The Argentinean Study of Mental Health Epide-
World Health Organization, Geneva
miology was funded by the Ministerio de Salud de la Nación (Argen-
11. Modesto MA, Klinar D (2016) Los psicólogos/as en Argentina.
tinean Ministry of Health) (Grant number 2002–17270/13–5). This sur-
Relevamiento Cuantitativo 2015 [Psychologists in Argentina.
vey was carried out in conjunction with the World Health Organization
Quantitative review]. Poster presented in the IV Congreso Inter-
World Mental Health (WMH) Survey Initiative. We thank the WMH
nacional de Investigación y Práctica Profesional en Psicología;
staff for assistance with instrumentation and fieldwork.
XIX Jornada de Investigación; 8° Encuentro de Investigadores de
Psicología del MERCOSUR. Buenos Aires
12. Ase I, Burijovich J (2009) La estrategia de Atención Primaria de la
Salud:¿progresividad o regresividad en el derecho a la salud? [The

13
Social Psychiatry and Psychiatric Epidemiology

primary health care strategy: progressiveness or retrogressivness 27. Bruffaerts R, Posada-Villa J, Al-Hamzawi AO et al (2015) Propor-
in the health rights?] Salud colectiva 5(1):27–47 tion of patients without mental disorders being treated in mental
13. Pennell BE, Mneimneh ZN, Bowers A et al (2008) Implementa- health services worldwide. Br J Psychiatry 206(2):101–109
tion of the World Mental Health Surveys. In: Kessler RC, Üstün 28. Wang PS, Aguilar-Gaxiola S, Alonso J et al (2007) Worldwide
TB (eds) The WHO World Mental Health Surveys: global per- use of mental health services for anxiety, mood, and substance
spectives on the epidemiology of mental disorders. Cambridge disorders: Results from 17 countries in the WHO World Mental
University Press, New York, pp 33–57 Health (WMH) Surveys. The Lancet 370(9590):841–850
14. Heeringa SG, Wells EJ, Hubbard F et al (2008) Sample designs 29. World Health Organization (2001) The world mental health report
and sampling procedures. In: Kessler RC, Üstün TB (eds) The 2001. Mental health: new understanding, new hope. World Health
WHO World Mental Health Surveys: global perspectives on the Organization, Geneva
epidemiology of mental disorders. Cambridge University Press, 30. Borges G, Medina-Mora ME, Benjet C, Lee S, Lane M, Breslau J
New York, pp 14–32 (2011) Influence of mental disorders on school dropout in Mexico.
15. Kessler RC, Üstün TB (2004) The World Mental Health (WMH) Rev Panam Salud Publica 30(5):477–483
Survey initiative. Version of the World Health Organization 31. Esch P, Bocquet V, Pull C, Couffignal S, Lehnert T, Graas M,
(WHO) composite international diagnostic interview (CIDI). Int Fond-Harmant L, Ansseau M (2014) The downward spiral of
J Methods Psychiatr Res 13(2):93–121 mental disorders and educational attainment: a systematic review
16. Haro JM, Arbabzadeh-Bouchez S, Brugha TS et al (2006) Con- on early school leaving. BMC Psychiatry 14:237
cordance of the composite international diagnostic interview ver- 32. Lee S, Tsang A, Breslau J et al (2009) Mental disorders and termi-
sion 3.0 (CIDI 3.0) with standardized clinical assessments in the nation of education in high-income and low- and middle-income
WHO World Mental Health Surveys. Int J Methods Psychiatr Res countries: epidemiological study. Br J Psychiatry 194(5):411–417
15:167–180 33. Baxter AJ, Scott KM, Vos T, Whiteford HA (2013) Global preva-
17. American Psychiatric Association (1994) Diagnostic and statis- lence of anxiety disorders: a systematic review and meta-regres-
tical manual of mental disorders, 4 edn. American Psychiatric sion. Psychol Med 43(5):897–910
Association, Washington, DC 34. Boyd A, van de Velde S, Vilagut G, de Graf R, O´Neill S, Florescu
18. Sheehan DV, Harnett-Sheehan K, Raj BA (1996) The measure- S, Alonso J, Kovess-Masfety V, EU-WMH Investigators (2015)
ment of disability. Int Clin Psychopharmacol 11(S3):89–95 Gender differences in mental disorders and suicidality in Europe:
19. Wolter KM (1985) Introduction to Variance Estimation. Springer, results from a large cross-sectional population-based study. J
New York AffectDisord 173:245–254
20. Research Triangle Institute (2002) SUDAAN: Professional Soft- 35. Andersen RM (2008) National Health surveys and the behavioral
ware for Survey Data Analysis [computer program]. Version 8.0.1. model of health services use. Med Care 46(7):647–653
Research Triangle Institute, Research Triangle Park, NC 36. Harris MG, Baxter AJ, Reavley N et al (2016) Gender-related pat-
21. Kish L, Frankel MR. Inferences from complex samples. J R Stat terns and determinants of recent help-seeking for past-year affec-
Soc Ser A:361–337 tive, anxiety and substance use disorders: findings from a national
22. SAS Institute Inc (2001) SAS/STAT Software: changes and epidemiological survey. Epidemiol Psychiatr Sci 25(6):548–561
enhancements, release 8.2. SAS Institute Inc, Cary 37. Kovess-Masfety V, Boyd A, van de Velde S et al (2014) Are there
23. Steel Z, Marnane C, Iranpour C et al (2014) The global prevalence gender differences in service use for mental disorders across coun-
of common mental disorders: a systematic review and meta-anal- tries in the European Union? Results from the EU. World Mental
ysis 1980–2013. Int J Epidemiol 43(2):476–493 Health survey. J Epidemiol Community Health 68(7):649–656
24. Goi PR, Vianna-Sulzbah M, Silveira L et al (2015) Treatment 38. Fleury MJ, Grenier G, Bamvita JM, Perreault M, Kestens Y, Caron
delay is associated with more episodes and more severe illness J (2012) Comprehensive determinants of health service utilisa-
staging progression in patients with bipolar disorder. Psychiatry tion for mental health reasons in a Canadian catchment area. Int J
Res 227(2–3):372–373 Equity Health 11:20
25. Post RM, Weiss SR (1998) Sensitization and kindling phenomena 39. Magaard JL, Seerlalan T, Schulz H, Brutt AL (2017) Factors asso-
in mood, anxiety and obsessive-compulsive disorders: the role of ciated with help-seeking behavior among individuals with major
serotonergic mechanisms in illness progression. Biol Psychiatry depression: a systematic review. Plos One 12(5):e0176730
44(3):193–206
26. Bonnin JE (2014) Treating without diagnosis: psychoanalysis in
medical settings in Argentina. Commun Med 11(1):15–26

13