Prevalence of comorbidities between mood and anxiety disorders : associated factors in a population sample of young adults in southern Brazil

This research aims to evaluate factors associated with the presence of comorbidities between mood and anxiety disorders in young adults aged 18 to 24 years, from Pelotas, Rio Grande do Sul State, Brazil. This was a cross-sectional, population-based study with a probabilistic sample by conglomerates. The Mini International Neuropsychiatric Interview (MINI) was used to assess mood and anxiety disorders. The prevalence of mental disorders in the sample (n = 1,561) was of 26.8% of which 9.7% had comorbidities between mood and anxiety disorders. The prevalence of comorbidities on mood and anxiety disorders is almost three times higher among women than men (p < 0.001). Lower education levels, socioeconomic status (p < 0.001) and a history of divorced parents (p < 0.050) was associated with comorbidities between mood and anxiety disorders. The main conclusion is that social factors are highly associated with comorbidities between mood and anxiety disorders. Prevention strategies on mental health should focus particularly on women in vulnerable social conditions. Mental Disorders; Mental Health; Young Adult Resumo Esta investigação teve como objetivo avaliar a presença de comorbidades entre transtornos de humor e ansiedade, assim como seus fatores associados, em adultos jovens entre 18 e 24 anos de idade em Pelotas, Rio Grande do Sul, Brasil. Em estudo transversal de base populacional a seleção amostral foi feita por conglomerados. O Mini International Neuropsychiatric Interview (MINI) foi utilizado para avaliar os transtornos de humor e ansiedade. A prevalência de transtornos mentais na amostra estudada (n = 1.561) foi de 26,8%, dos quais 9,7% apresentaram comorbidade entre transtornos de humor e ansiedade. A comorbidade entre transtornos de humor e ansiedade foi quase três vezes maior entre as mulheres quando comparadas aos homens (p < 0,001). Níveis mais baixos de escolaridade, nível socioeconômico (p < 0,001) e histórico de pais divorciados (p < 0,050) apresentaram-se associados com comorbidade entre transtornos de humor e ansiedade. A principal conclusão é que os fatores sociais são altamente associados com comorbidade entre transtornos de humor e ansiedade. Transtornos Mentais; Saúde Mental; Adulto Jovem 2413 ARTIGO ARTICLE http://dx.doi.org/10.1590/0102-311X00012614 Molina MRAL et al. 2414 Cad. Saúde Pública, Rio de Janeiro, 30(11):2413-2422, nov, 2014 Introduction Epidemiological studies show that millions of people in the world suffer from a mental disorder, especially in developing countries 1,2,3. Comorbidity cases are also very common in population studies with prevalence ranging from 14-23% 4,5 with significant consequences for mental health. Comorbid mood and anxiety disorders (54.3%) 6 and pure cases have a high prevalence in the population (9.5 and 18.1%) 5 resulting in significant damage to the quality of life. Studies have shown that comorbidity, especially between mood and anxiety disorders, is associated with increased severity, higher duration of illness, greater functioning impairment, and more frequent use of health services 7,8,9,10. Understanding the social factors related to comorbidity cases, especially in a highly prevalent diagnostic family of mental disorders, such as mood and anxiety disorders, may have important implications for the development of public health policies. Studies on cross family diagnostic comorbidity of mood and anxiety disorders associated with social factors are scarce, especially among young adults. The few studies that assess young adults do not focus on the ages of 18 to 24, the most frequent onset period of those mental disorders 5,11. Nonetheless, some sociodemographic factors have been previously associated with the presence of comorbid mental disorder such as female gender 6; low levels of education 6; and unmarried status 12. It is important to highlight that comorbidity was not presented in those studies, so it is still unclear if those social factors will relate similarly to the comorbidity of mood and anxiety disorders. It is noteworthy that most studies of psychopathology tend to focus on single-disorder cases and even in clinical trials comorbidity cases are not identified. It is understandable that pure cases are more often studied since it minimizes the effect of possible confounders. In fact, we often approach comorbid cases based on evidence derived from studies of single disorder cases. It is important to understand the specific characteristics of cross diagnostic family comorbidity particularly considering that the prevalence of some comorbid cases is at least equal to single disorders in the population 5. In addition, Lamers et al. 13 found that in 57% of comorbid cases, anxiety was preceded by depression; in 18% of cases, depression preceded anxiety. This reinforces the rationale of investigating not only pure cases but also the relationship between comorbidities, its impact on health, and the associated risk factors in a young adult population 13. Therefore, an assessment of the distribution of different mental disorders is necessary to allow for a better understanding of such psychiatric phenomena and to contribute to the development of preventive interventions in mental health, especially among young adults. Because of the scarcity of research involving this theme in the scientific literature, the present study aims to evaluate factors associated with the presence of comorbidities between mood and anxiety disorders in young adults aged 18 to 24, from Pelotas, Rio Grande do Sul State, Brazil.

ulation-based study with a probabilistic sample by conglomerates.The Mini International Neuropsychiatric Interview (MINI) was used to assess mood and anxiety disorders.The prevalence of mental disorders in the sample (n = 1,561) was of 26.8% of which 9.7% had comorbidities between mood and anxiety disorders.The prevalence of comorbidities on mood and anxiety disorders is almost three times higher among women than men (p < 0.001).Lower education levels, socioeconomic status (p < 0.001) and a history of divorced parents (p < 0.050) was associated with comorbidities between mood and anxiety disorders.The main conclusion is that social factors are highly associated with comorbidities between mood and anxiety disorders.Prevention strategies on mental health should focus particularly on women in vulnerable social conditions.

Introduction
Epidemiological studies show that millions of people in the world suffer from a mental disorder, especially in developing countries 1,2,3 .Comorbidity cases are also very common in population studies with prevalence ranging from 14-23% 4,5 with significant consequences for mental health.Comorbid mood and anxiety disorders (54.3%) 6 and pure cases have a high prevalence in the population (9.5 and 18.1%) 5 resulting in significant damage to the quality of life.Studies have shown that comorbidity, especially between mood and anxiety disorders, is associated with increased severity, higher duration of illness, greater functioning impairment, and more frequent use of health services 7,8,9,10 .Understanding the social factors related to comorbidity cases, especially in a highly prevalent diagnostic family of mental disorders, such as mood and anxiety disorders, may have important implications for the development of public health policies.
Studies on cross family diagnostic comorbidity of mood and anxiety disorders associated with social factors are scarce, especially among young adults.The few studies that assess young adults do not focus on the ages of 18 to 24, the most frequent onset period of those mental disorders 5,11 .Nonetheless, some sociodemographic factors have been previously associated with the presence of comorbid mental disorder such as female gender 6 ; low levels of education 6 ; and unmarried status 12 .It is important to highlight that comorbidity was not presented in those studies, so it is still unclear if those social factors will relate similarly to the comorbidity of mood and anxiety disorders.
It is noteworthy that most studies of psychopathology tend to focus on single-disorder cases and even in clinical trials comorbidity cases are not identified.It is understandable that pure cases are more often studied since it minimizes the effect of possible confounders.In fact, we often approach comorbid cases based on evidence derived from studies of single disorder cases.It is important to understand the specific characteristics of cross diagnostic family comorbidity particularly considering that the prevalence of some comorbid cases is at least equal to single disorders in the population 5 .In addition, Lamers et al. 13 found that in 57% of comorbid cases, anxiety was preceded by depression; in 18% of cases, depression preceded anxiety.This reinforces the rationale of investigating not only pure cases but also the relationship between comorbidities, its impact on health, and the associated risk factors in a young adult population 13 .
Therefore, an assessment of the distribution of different mental disorders is necessary to allow for a better understanding of such psychiatric phenomena and to contribute to the development of preventive interventions in mental health, especially among young adults.Because of the scarcity of research involving this theme in the scientific literature, the present study aims to evaluate factors associated with the presence of comorbidities between mood and anxiety disorders in young adults aged 18 to 24, from Pelotas, Rio Grande do Sul State, Brazil.

Methods
This is a cross-sectional, population-based study with young adults, aged between 18 to 24 years, living in the urban area of Pelotas, a city in southern Brazil, with approximately 330,000 inhabitants (83% of whom live in urban areas).
The sample selection was performed by clusters considering the population of 39,667 individuals in the above age range according to the current census of 448 sectors in the city (http://www.ibge.gov.br, accessed on Jun/2012).To assure the necessary sample size, 89 census-based sectors were randomly selected.The home selection in the sectors was performed using systematic sampling.The first home was considered to be the house at the corner as pre-established by the IBGE (Brazilian Institute of Geographic and Statistics) as the beginning of the sector.Given the required number of respondents and to increase representativeness, the interval of selection was determined by skipping two houses.After identifying the subjects, we explained the aims of the study and in case of agreement to participate we scheduled a home visit to conduct the interview.Psychology majors with extensive training in conducting interviews collected the data.We identified 1,762 young adults who met the criteria for inclusion; 201 (11.4%) were not found at home during three home visits or refused to complete the interview and were excluded from the total sample.Thus, a total of 1,561 respondents were analyzed.
Considering measured prevalence of comorbidity between mood and anxiety disorders (9.7%) found in the present study, a sample size estimation of 1,132 young adults was needed with a 95% confidence level and the lowest error accepted was 1.7%.So the sample was larger than what was required for our purpose.At first, participants answered a questionnaire consisting of sociodemographic questions (e.g.education, employment) and general health issues.A socioeconomic evaluation of the participants was performed using the classification from the Brazilian Association of Research Companies (http://www.ibge.gov.br, accessed on 10/Mar/2010), which is based on the total of material goods and the householder's education.This scale provides a list of consumer items and the respondent is asked to identify how many of them exist in their home.Moreover, the level of education of the household head is also considered in the total score that indicates the classification of socioeconomic classes (A, B, C, D, E); "A" refers to the highest socioeconomic class and "E" to the lowest.
Current mental disorders were assessed using the Mini International Neuropsychiatric Interview 5.0 (MINI).This short interview (15 to 30 minutes) is designed for use in clinical practice and aimed at diagnosing the interviewees according to the Diagnostic and Statistical Manual of Mental Disorders, 4 th edition (DSM-IV) and the International Statistical Classification of Diseases and Related Health Problems, 10 th revision (ICD-10), criteria.Although, the psychometric characteristics for each disorder assessed varies compared to a Structured Clinical Interview from DSM (SCID), the psychometric characteristics of MINI are considered appropriate 14 .We assessed the following modules of current disorders of MINI: major depressive episode, dysthymia, hypomania and mania episodes, panic disorder, agoraphobia, social phobia, obsessive-compulsive disorder, post-traumatic stress disorder, generalized anxiety disorder.The main dependent nominal variable in the study was categorized into subjects with no psychiatric disorders, with mood disorder, with anxiety disorder and those with comorbidity between mood and anxiety disorders.We focused on cross family diagnostic comorbidity, so to be considered a comorbidity case the participants had to present full diagnostic criteria for at least one disorder in each diagnostic family (mood and anxiety disorders).
Data entry and processing was carried out through double data entry on Epi-Info 6.04d (Centers for Disease Control and Prevention, Atlanta, U.S.A.) software.The software checked data entry automatically and also tested for inconsistencies on double data entry.For the statistical analysis, we used SPSS, version 13 (SPSS Inc., Chicago, U.S.A.).Descriptive statistics were used to characterize the study sample and to present the prevalence of comorbidity of mental disorders.
The first analytic approach to select the variables was a crude multinomial regression describing associations between the occurrence of mental disorders and the independent variables investigated.Finally, adjusted multinomial regression was used following a hierarchical framework approach 15 .Those with a p-value ≤ 0.2 in the bivariate tests were included in the model.
Young adults who presented with any psychiatric disorder and/or drug abuse were directed to psychiatric outpatient care.The project was approved by the Ethics Research Committee of Pelotas Catholic University according to process number 2009/24.

Results
As presented in Table 1, most of the sample was female (56.4%), white (73.5%), had 9 to 11 years of education, belonged to economic class C (48.1%), lived with a father and/or mother (65.8%), had parents who were not divorced (64%), was not living with a partner (28.2%), worked in the past 12 months (66.9%), and was religious (62%).
The results of the MINI interviews showed that 9.7% had comorbidity between mood and anxiety disorders and 26.8% of the sample had at least one of the conditions assessed (Figure 1).In addition, the prevalence of pure mood disorder was 5.9%, and 11.1% of pure anxiety disorder.The three most prevalent diagnoses were major depressive disorder (12.6%), agoraphobia (12.3%) and generalized anxiety disorder (9.7%) (Table 2).
Comorbidities between mood and anxiety disorders were more frequently observed among females; and among participants that did not work in the past 12 months, with low socioeconomic status, fewer years of education, and divorced parents.Skin color also showed a difference in the proportion of comorbidities between mood and anxiety disorders but with no statistical significance (Table 3).In regard to mood disorders, a statistically significant higher proportion was observed in females, unemployed people and people not living with a partner.Similarly, female gender and unemployment were related to a higher prevalence of anxiety disorders.
After the adjusted analysis, comorbidities between mood and anxiety disorders were three times higher for women than men, and inversely associated with education and socioeconomic status.Young adults without divorced parents showed 31% lower odds of presenting comorbidities than people with divorced parents (Table 4).The analysis that considered the odds ratios for mood disorders alone showed a higher prevalence among females and people who were not living with a partner.The female gender was the only independent variable correlated with anxiety disorders.

Discussion
The prevalence of mental disorders in this study was high (26.8%)as well as comorbidity rates between anxiety and mood disorders (9.7%).Other Population based studies assessing the prevalence of comorbidity of psychiatric disorders are scarce.The few studies that assesses psychiatric disorders do so using a single diagnostic family (e.g.Mood disorders) or individually (e.g.Major depression), focusing only on the most common mental disorders (e.g.Nonpsychotic depression, anxiety and somatoform symptoms) 16,17 .It is important to highlight that, in the present study specific, social factors were related only to cross family diagnostic comorbidity, and not to the pure cases of anxiety or mood disorders.
The data suggests that anxiety and mood disorders not only have high prevalences, but also show high comorbidity rates.In fact, comorbidity rates were higher than the prevalence of pure mood disorder cases in the present study.Some studies also demonstrate that comorbidity cases can be more prevalent than pure mood or anxiety disorders 6,12 .Similar results have also been found in early stages of adolescence and young adult life 18,19 .We believe that the high prevalence of comorbid anxiety and depressive disorder might be explained by the similarity of symptoms and specific social vulnerability factors for comorbidities.
It is important to highlight that socioeconomic status and education were inversely associated only with psychiatric comorbidity between mood and anxiety disorder.In regard to education, similarly to the De Graaf et al. 6 study, we found that participants with fewer years of education showed higher rates of comorbidities between mood and anxiety disorders.The present study highlights a possible influence of education and socioeconomic status in more complex and severe cases, as the comorbidity between mood and anxiety disorders.
Interestingly, we found that having divorced parents was related only to the presence of comorbidity cases in young adults.Studies show that divorced parents convey less social support in different periods of life 20 .This relationship is frequently reported in children with divorced parents, possibly as a result of the traumatic experience of the parents' separation in childhood.Experiencing trauma is a risk factor to the onset of comorbidities between depressive and anxiety disorders 6 .In adolescence, parental divorce was related to low social support from the family and high levels of depression/anxiety 20 .In fact, not living with parents in adolescence is related  to higher levels of depression and anxiety 11 .Accordingly, young adults with divorced parents might experience less social support during their lifetime.Contrary to the literature, we did not find a higher prevalence of comorbid cases or pure anxi-ety disorder among people who do not live with a partner.We only found this relationship among pure cases of mood disorder.According to Meng & D'Arcy 12 , unmarried young adults were more likely to develop anxiety or mood disorders with comorbidity with other mental conditions.In an-other study, people who do not live with a partner show a higher prevalence of comorbidity between mood and anxiety disorder 6 .It is still unclear why we did not find this relationship among comorbid cases, and pure anxiety disorder.We speculate it might be due to the younger age of our sample compared to other studies.
In regard to gender, women not only more frequently presented comorbidity between mood and anxiety disorder, but also pure cases of both disorders compared to men.In fact, similar to other studies 11,19 , we found that women are affected three times more often than men.It is noteworthy that women are more often affected by mental disorders than men in comorbidity cases 21 .Additionally women present twice the rate of comorbidity between mood and anxiety disorders than men 6 .This finding may be explained by the fact that women have worse self-rated health; therefore, women may be more willing than men to express their symptoms 22,23 .Another trait is that women are more likely to remember and report their mood swings than men 24 .
Among the limitations of this study is the cross-sectional study design that precludes the evaluation of any causal relationship between the independent variables and the number of mental disorders.A potential source of bias is that inter- views rely on self-reporting and were conducted during home visits which may lead to an unwillingness to participate and non-disclosure, for example, about the risky behaviors common in this age group.Since these are usual procedures in clinical epidemiology research, the research team was trained to mitigate these potential limitations.Also, MINI has appropriate psychometric properties, however, there is some discussion in regard to the diagnostic criteria for mental disorders in the scientific community 25,26 .Our research should be read in light of the current knowledge of the research field, in which diagnostic criteria for common mental disorders often overlap, and for that reason it is possible that comorbidities are, in fact, manifestations of the same condition.
It is important to call attention to some positive aspects of the present study, such as sample selection in which data was derived from a representative sample of young adults.The study of young adults allows early identification of the comorbidity cases profile allowing for the development of specific health policies with the goal of minimizing and preventing future impairment.
We highlight that the present results can be generalized for the same age group to other similar communities.Also, the paper adds to the literature by studying in a population based study, not only one mental disorder at a time but by focusing on comorbidities between two highly prevalent groups of mental disorders.The study of comorbidities allows a broader perspective into mental disorders in a relatively young group, contributing considerably to prevention and mental health promotion.

Conclusions
These results draw attention to the association of social factors on cross diagnostic family comor-bidity on young adults.These findings enable a new perspective in regard to prevention and mental health promotion in Brazil.Social vulnerability (e.g.female gender; fewer years of education; low socioeconomic status; unemployment; and having divorced parents) is strongly related to comorbid mental disorders especially in this particular age group.Thus, early interventions are needed for young adults with primary disorders in order to avoid psychiatric comorbidities across the lifetime.Therefore, primary care professionals should be alert to psychosocial stress in this age group given its relationship with the early development of mental disorders.We suggest the implementation of prevention strategies in primary care focusing on the insertion of mental health professionals in basic health units.
Trastonos Mentales; Salud Mental; Adulto Joven Contributors M. R. A. L. Molina contributed to the data analysis and interpretation, writing and approval of the final version.B. Spessato contributed to the data analysis and interpretation, critical review and approval of the final version.K. Jansen contributed to the study design, writing, data analysis, and approval of the final version.R. Pinheiro contributed to the study design and approval of the final version.R. Silva contributed to the study design, critical review, and approval of the final version.L. D. M. Souza contributed to the study design, data analysis and interpretation, writing, and approval of the final version.

Figure 1 Venn
Figure 1 Venn diagram with distribution of mental disorders among young adults from the city of Pelotas, southern Brazil, 2007-2008 (n = 1,561).

Table 1
Sample characteristics of young adults aged 18 to 24 from the city of Pelotas, southern Brazil, 2007-2008 (n = 1,561).

Table 2
Prevalence of mental disorders, according to the Mini International Neuropsychiatric Interview (MINI) in young adults aged 18 to 24 in the city of Pelotas, Rio Grande doSul, Brazil, 2007-2008.(n = 1,561)

Table 3
Bivariate analysis between independent variables in study and presence of mood and anxiety disorders and its comorbidity in young adults aged 18 to 24 in the city of Pelotas, Rio Grande doSul, Brazil, 2007-2008 (n = 1.561)

Table 4
Multinominal regression for presence of mood and anxiety disorders and its comorbidity in youth aged 18 to 24 in the city of Pelotas, Rio Grande do Sul,