Gac Sanit. 2019;33(3):235–241 Original Mental health, family roles and employment status inside and outside the household in Spain Jorge Arias-de la Torrea,b,c,∗ , Antonio J. Molinaa , Tania Fernández-Villaa , Lucía Artazcozb,d , Vicente Martína,b a Research Group of Interaction Gen-Health-Environment, University of León, León, Spain b CIBER de Epidemiología y Salud Pública (CIBERESP), Spain c Agency for Health, Quality and Assessment of Catalonia (AQuAS), Barcelona, Spain d Agència de Salut Pública de Barcelona, Barcelona, Spain a r t i c l e i n f o a b s t r a c t Article history: Objective: To document the prevalence of poor mental health by gender and social class, and to analyze Received 4 July 2017 if poor mental health is associated with the family roles or the employment status inside and outside the Accepted 14 November 2017 household. Available online 13 February 2018 Method: A cross-sectional study based on a representative sample of the Spanish population was carried out (n = 14,247). Mental health was evaluated using GHQ-12. Employment status, marital status, family Keywords: roles (main breadwinner and the person who mainly carries out the household work) and educational Mental health level were considered as explanatory variables. Multiple logistic regression models stratified by gender Occupational health Gender and social class were fitted and adjusted odds ratios (aOR) were obtained. Socioeconomic factors Results: Gender and social class differences in the prevalence of poor mental health were observed. Health survey Unemployment was associated with higher prevalence. Among men the main breadwinner role was related to poor mental health mainly in those that belong to manual classes (aOR = 1.2). Among women, mainly among nonmanual classes, these problems were associated to marital status: widowed, separated or divorced (aOR = 1.9) and to dealing with the household work by themselves (aOR = 1.9). Conclusions: In Spain, gender and social class differences in mental health still exist. In addition, fam- ily roles and working situation, both inside and outside the household, could constitute a source of inequalities in mental health. © 2018 SESPAS. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Salud mental, rol familiar y situación laboral dentro y fuera del hogar en España r e s u m e n Palabras clave: Objetivo: Conocer la prevalencia de mala salud mental por sexo y clase social, y analizar si la salud mental Salud mental se relaciona con los roles familiares y la situación laboral fuera y dentro del hogar. Salud laboral Método: Se realizó un diseño transversal basado en una muestra representativa de la población española Género (n = 14.247). La salud mental se evaluó mediante el GHQ-12. Se consideraron como variables explicativas Factores socioeconómicos la situación laboral, el estado civil, el rol familiar (sustentador principal y persona que realiza el trabajo Encuestas de salud doméstico) y el nivel de estudios. Se ajustaron modelos de regresión logística estratificados por sexo y se obtuvieron las odds ratio ajustadas (ORa). Resultados: Se observaron diferencias en la prevalencia de mala salud mental por sexo y clase social. El desempleo se asoció con mayor prevalencia. En hombres pertenecientes a clases menos favorecidas, el rol de sustentador principal se relacionó con mala salud mental (Ora = 1,2) En mujeres que pertenecían a clases más favorecidas, el estado civil viuda, separada o divorciada (ORa = 1,9) y realizar el trabajo doméstico solas (ORa = 1,9) se relacionaron con mala salud mental. Conclusiones: En España, en salud mental continúa habiendo diferencias de sexo y clase social. Además, el rol familiar y la situación laboral fuera y dentro del ámbito doméstico podrían constituir también una fuente de desigualdad en salud mental. © 2018 SESPAS. Publicado por Elsevier España, S.L.U. Este es un artı́culo Open Access bajo la licencia CC BY-NC-ND (http://creativecommons.org/licenses/by-nc-nd/4.0/). ∗ Corresponding author. E-mail address: jorgeariasdelatorre@gmail.com (J. Arias-de la Torre). https://doi.org/10.1016/j.gaceta.2017.11.005 0213-9111/© 2018 SESPAS. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc- nd/4.0/). 236 J. Arias-de la Torre et al. / Gac Sanit. 2019;33(3):235–241 Introduction was collected through personal interviews between July 2011 and June 2012. Nowadays, one of the main factors that has led to a decline in The selection criteria for the present study was as follows: quality of life, dependency and disability worldwide are mental people between 16 and 65 who answered the GHQ-12 and health problems.1,2 In Spain, they are considered, together with whose employment status was working, unemployed, dedicated neurologic diseases, as the number one cause of disability among to domestic work or other situation (n = 15,150). Due to possi- non-infectious diseases.3 ble effects of reverse causality those people unable to work (2.4%) Prevalence of mental health problems has been assessed in sev- were excluded. Furthermore, those people for whom it was not eral studies, both in the European population and in the Spanish possible to determine their employment status (0.1%), social class population,4–10 finding in them variations according to the geo- (2.5%), marital or cohabitation status (0.1%) and the person who graphic area studied as well as across time.11,12 These facts show takes care of the housework (1.2%) were also excluded. The final the relevance of a periodic and specific assessment of this preva- sample included 14,247 people. lence for each geographic area, in order to try to decrease and prevent the consequences associated with it. Variables In relation to the factors associated with a higher prevalence of health problems, previous studies show a close relationship • Mental health: assessed using the GHQ-12, a mental health between gender social class and the prevalence of both physical screening tool for general population and non-psychiatric and mental health problems.13–17 In addition, these studies find a patients, validated for the Spanish population.26 The GHQ-12 higher prevalence of problems among women than among men, evaluates the subjective mental state of the person and detects and among people who belong to the most deprived social classes mainly two types of disorders: anxiety and depression. It consists in comparison with people belonging to the most favored social of 12 Likert-type items with a 4 point response scale. A 2-point classes. These results show the necessity of bearing these factors in scoring method has been used, assigning 0 points to answers 0 mind when carrying out population health studies, as is proposed and 1, and 1 point to answers 2 and 3, and then adding the points through the intersectionality concept, which considers gender and of the 12 items obtaining a total score between 0 and 12. A score social class as two of the main axes of physical and mental health of 3 or greater was considered to indicate poor mental health inequality.18,19 according to the proposal of the authors of the scale as well as in Furthermore, different studies have been carried out in order previous studies carried out in this population.8,22,26,27 to analyze possible gender differences in the influence on men- tal health of some socioeconomic variables such as the family role • Occupational social class: based on the person’s current or past or the civil status.20,21 The results obtained point to the impor- tance of the family roles and suggest that the paid-work related work, and classified through the categories proposed by the factors and the economic support for the family could be to a greater working group on social determinants of the Spanish Society of extent associated with men’s mental health, while factors related Epidemiology.28 The social class categories have been grouped to the domestic and family environment, could be more strongly for the ENSE-2012 in: directors and managers of establishments associated with women’s health. with 10 or more employees and university degrees (class I), With regard to the work situation and its relation to mental directors and managers of establishments with fewer than health, several studies have been carried out.22–24 Many of these 10 employees and university diplomas (class II), intermediate studies have been centred on people having a certain work situ- occupations and self-employed people (class III), supervisors ation, such as employed and unemployed, and only some make and workers in skilled technical occupations (class IV), skilled reference to alternative situations like dedication to housework workers in the primary sector and other semi-skilled workers (i.e., serving as a homemaker) or other work situations. It is appro- (class V) and unskilled workers (class VI). Due to the low number priate to take into account these situations in order to obtain a of people in some categories, the social classes were collapsed wider view of how work situation is related to mental health. into two categories: nonmanual (classes I, II and III) and manual Therefore, in this context, the aims of the present study are: to (classes IV, V and VI). analyze if mental health problems are associated with employment status as well as to family and household characteristics; and exam- • Employment status. The following situations were considered: ine if there are differences in prevalence related to the gender and working, unemployed, dedicated to domestic work (i.e., serving social class. as a homemaker) and other situations (included in this category are: people that are studying; retirees and early retirees and other situations not covered by the other categories) Methods • Family and household characteristics. Marital status (single, Study design and participants married or cohabiting and widowed or separated or divorced) and the family roles from two variables: main breadwinner (yes A cross-sectional study based on a representative sample of or no) and the person who mainly carries out the household the non-institutionalized Spanish population (N = 21,007) was car- work (yourself, yourself shared with someone else and someone ried out. Data from the National Health Survey of Spain 2011/2012 else). (ENSE-2012) were used.25 The survey’s questionnaires contain information on variables such as the occupational social class, • Education. The following categories were considered: university the employment status, the family and household characteristics, studies, secondary or high school, primary or no education and as well as data on mental health through the inclusion of the other situations (including people whose educational degree General Health Questionnaire (GHQ-12), a screening tool used to does not have correspondence with the Spanish educational detect mental health problems. The sample of the ENSE-2012 was levels and people not covered by the other categories). selected using a three-stage stratified sampling method, obtain- ing a response rate of 89%, from which, 28% represents reserve households to replace incidents in headlines.25 The information • Age: considered as a continuous variable. J. Arias-de la Torre et al. / Gac Sanit. 2019;33(3):235–241 237 Data analysis most frequent situation was that they carry out these chores by themselves. A descriptive analysis of prevalence of poor mental health Table 2 shows the prevalence of poor mental health across inde- according to the employment status and the family and household pendent variables. A higher prevalence among women, and among characteristics stratified by gender and social class was performed. those that belong to manual social classes (men: 17.5%; women: Differences by gender in the studied variables were tested using 25.2%) than nonmanual (men: 14.9%; women: 18.2%) was observed. Chi Square test for categorical variables and t test for age and In relation to the employment status, there was a higher prevalence GHQ-12 when was considered as a continuous variable. Secondly, among unemployed people in both genders, independent of the multivariate logistic regression models were fitted to analyze social class. Additionally, among women of manual social classes the relationship between mental health and employment status, higher rates were observed across all working situations. Regarding socioeconomic variables as well as family and household variables. family and household characteristics, there was a higher observed Also, a multivariate sub-analysis taking into account only the peo- prevalence, independent of the gender and the social class, among ple who was working and with marital status married or cohabiting the widowed, separated or divorced. A higher prevalence was also was performed. Adjusted odds ratios (aOR) and respective con- found among those that assume the main breadwinner role and fidence intervals of 95% (CI95%) were obtained. All models were belong to manual social classes. As for household work allocation, stratified by gender and social class individually and jointly, and higher rates of poor mental health were found among people that were adjusted by all the explanatory variables and age as a contin- carry out with this work by themselves, except among men from uous variable. The models including only working population with nonmanual social classes for whom the highest prevalence was marital status married or cohabiting were stratified by gender and observed among those that share these tasks (15.1%). social class jointly. Table 3 shows the association between employment status, domestic and family characteristics and poor mental health strat- ified by gender in the first two columns and by social class in Results the last two columns. Unemployment, the dedication to domes- tic work and other working situations were related to poor mental Table 1 shows the characteristics of the study sample by gender health in both genders in the manual social classes. In nonmanual and social class. Of the participants, more than half were women social classes, poor mental health was only related to unemploy- (50.6%) and most were from the manual social classes (men: 79.9%; ment (aOR = 1.85; 95%CI: 1.30-2.64). Furthermore, the civil status women: 79.2%). Regarding the employment status, for both gen- of widowed, separated or divorced was associated with poor men- ders the most frequent situation was working, though this figure tal health in both social classes and in women. Regarding family was slightly lower in the manual social classes. Among both gen- roles, only among men (aOR = 1.20; 95%CI: 1.03-1.39) and among ders unemployment was higher among the manual social classes the manual social classes (aOR = 1.13; 95%CI: 1.02-1.25) was a rela- than in nonmanual. The most common marital status was mar- tionship between poor mental health and main breadwinner role ried or cohabiting and about half of the people assume the role of observed, while only among women a relationship between carry- main breadwinner. In terms of household chore allocation, there ing out the household work by themselves and poor mental health are differences between men and women. While among men the (aOR = 1.29; 95%CI: 1.05-1.58) was found. As for gender and social most frequent situation, independent of the social class, was that class, an association between poor mental health and belonging to these chores were done by another person, among women the the manual social classes was only found in women. Table 1 General description of the sample by gender and social class. Nonmanual (%) Manual (%) Men Women p Men Women p (n = 1419) (n = 1497) (n = 5625) (n = 5706) Employment status 0.160 <0.001 Working 63.4 61.7 49.2 46.1 Unemployed 7.3 6.7 18.5 15.1 Domestic work 9.6 8.7 12.2 14.1 Other 19.8 22.9 20.0 24.7 Marital status Single 30.1 29.7 0.003 35.7 28.3 <0.001 Married/cohabiting 63.1 60.1 57.4 58.1 Widowed/separated/divorced 6.8 10.2 6.9 13.6 Main breadwinner 50.0 50.6 0.774 51.2 47.8 <0.001 Household work <0.001 <0.001 Yourself 16.4 43.2 18.2 55.6 Shared 40.2 39.5 35.9 33.8 Someone else 43.3 17.2 45.9 10.7 Education 0.017 0.003 University studies 48.4 50.0 7.6 8.6 Secondary or high school 28.7 24.9 29.9 29.1 Primary or no education 16.4 19.7 58.2 56.8 Other 6.5 5.4 4.3 5.5 Mean age (SD) 43.1 (12.3) 43.3 (12.2) 0.641 42.4 (13.0) 42.8 (13.1) 0.057 GHQ-12 mean (SD)a 1.1 (2.1) 1.3 (2.3) 0.003 1.2 (2.3) 1.7 (2.8) <0.001 SD: standard deviation. a GHQ-12 considered as a continuous variable. 238 J. Arias-de la Torre et al. / Gac Sanit. 2019;33(3):235–241 Table 2 Poor mental health prevalence (%) across independent variables stratified by gender and social class. Nonmanual Manual Men Women Men Women a a a a n % n % p n % n % p Poor mental health 211 14.9 273 18.2 0.015 984 17.5 1438 25.2 <0.001 Employment status 0.456 <0.001 Working 123 13.7 148 16.0 382 13.8 556 21.1 Unemployed 24 23.3 28 27.7 271 26.1 281 32.7 Domestic work 23 16.9 27 20.8 120 17.2 223 27.6 Other 41 14.6 70 20.4 211 18.8 378 26.8 Marital status 0.068 <0.001 Single 69 16.2 74 16.7 351 17.5 372 23.0 Married/Cohabiting 124 13.8 158 17.6 540 16.7 802 24.2 Widowed/separated/divorced 18 18.8 41 26.8 93 24.1 264 34.1 Main breadwinner 0.783 0.010 Yes 104 14.6 138 18.2 532 18.5 701 25.7 No 107 15.1 135 18.2 452 16.5 737 24.8 Household work <0.001 <0.001 Yourself 35 15.0 136 21.0 211 20.6 872 27.5 Shared 86 15.1 105 17.7 339 16.8 441 22.9 Someone else 90 14.6 32 12.4 434 16.8 125 20.6 Education 0.348 0.015 University studies 91 13.3 121 16.2 626 16.0 105 21.3 Secondary or high school 68 16.7 79 21.2 252 15.0 393 26.7 Primary or no education 38 16.3 62 21.0 68 19.2 850 26.2 Other 14 15.2 11 13.6 36 14.8 90 28.5 a Number of people with a GHQ-12 ≥3 punctuation. Table 3 Association between poor mental health, employment status and family and household characteristics.. Stratification by gender Stratification by social class a a MenaOR (CI95%) WomenaOR (CI95%) NonmanualaOR (CI95%)a ManualaOR (CI95%)a (n = 7044) (n = 7203) (n = 2916) (n = 11,331) Employment status Working 1.00 1.00 1.00 1.00 Unemployed 2.17 (1.83-2.57)d 1.77 (1.51-2.10)d 1.85 (1.30-2.64)c 1.97 (1.74-2.24)d Domestic work 1.37 (1.10-1.71)c 1.34 (1.11-1.61)c 1.32 (0.93-1.90) 1.36 (1.17-1.59)d Other 1.38 (1.14-1.67)d 1.19 (1.12-1.39)b 1.23 (0.91-1.67) 1.27 (1.11-1.45)d Marital status Single 1.00 1.00 1.00 1.00 Married/cohabiting 0.93 (0.78-1.10) 0.88 (0.76-1.02) 0.99 (0.76-1.29) 0.90 (0.79-1.01) Widowed/separated/divorced 1.27 (0.98-1.65) 1.29 (1.06-1.57)b 1.54 (1.05-2.26)b 1.25 (1.06-1.48)c Main breadwinner No 1.00 1.00 1.00 1.00 Yes 1.20 (1.03-1.39)b 1.07 (0.94-1.21) 1.04 (0.82-1.31) 1.13 (1.02-1.25)b Household work Someone else 1.00 1.00 1.00 1.00 Yourself 1.02 (0.83-1.25) 1.29 (1.05-1.58)b 1.31 (0.99-1.74) 1.11 (0.97-1.28) Shared 1.06 (0.91-1.22) 1.18 (0.96-1.45) 1.21 (0.94-1.57) 1.04 (0.91-1.18) Education University studies 1.00 1.00 1.00 1.00 Secondary or high school 1.00 (0.80-1.24) 1.15 (0.95-1.40) 1.28 (1.01-1.62)b 1.00 (0.83-1.21) Primary or no education 1.18 (0.95-1.47) 1.17 (0.97-1.41) 1.23 (0.92-1.64) 1.12 (0.93-1.34) Other 0.99 (0.67-1.47) 1.37 (1.00-1.88) 0.86 (0.51-1.47) 1.27 (0.95-1.69) Social class Nonmanual 1.00 1.00 − − Manual 1.01 (0.84-1.22) 1.29(1.10-1.52)c − − Gender Men − − 1.00 1.00 Women − − 1.18 (0.95-1.45) 1.52 (1.37-1.69)d aOR: adjusted odds ratio; CI95%: confidence interval of 95%. a Odds ratios adjusted by all explanatory variables and age. b p <0.05. c p <0.01. d p <0.001. J. Arias-de la Torre et al. / Gac Sanit. 2019;33(3):235–241 239 Table 4 Association between poor mental health, employment status and family and household characteristics stratified by gender and social class. Men, aOR (CI95%) Women, aOR (CI95%) Nonmanual Manual Nonmanual Manual Employment status Working 1.00 1.00 1.00 1.00 Unemployed 1.84 (1.09-3.09)b 2.3 (1.86-2.67)d 1.89 (1.17-3.07)b 1.78 (1.49-2.11)d Domestic work 1.21 (0.71-2.07) 1.40 (1.10-1.79)c 1.37 (0.84-2.24) 1.32 (1.08-1.60)c Other 0.91 (0.56-1.59) 1.59 (1.21-1.84)d 1.49 (1.01-2.20)b 1.13 (0.95-1.35) Marital status Single 1.00 1.00 1.00 1.00 Married/cohabiting 0.66 (0.44-1.01) 0.99 (0.82-1.20) 1.16 (0.81-1.66) 0.84 (0.71-0.99)b Widowed/separated/divorced 1.05 (0.56-1.58) 1.31 (0.99-1.74) 2.00 (1.22-3.29)c 1.19 (0.96-1.48) Main breadwinner No 1.00 1.00 1.00 1.00 Yes 1.03 (0.72-1.48) 1.23 (1.04-1.45)b 1.06 (0.77-1.44) 1.07 (0.94-1.23) Household work Someone else 1.00 1.00 1.00 1.00 Yourself 0.83 (0.51-1.35) 1.07 (0.86-1.34) 1.90 (1.23-2.95)c 1.14 (0.90-1.43) Shared 1.11 (0.80-1.55) 1.05 (0.89-1.22) 1.58 (1.01-2.45)b 1.06 (0.84-1.34) Education University studies 1.00 1.00 1.00 1.00 Secondary or high school 1.26 (0.89-1.79) 0.88 (0.66-1.19) 1.28 (0.72-1.77) 1.09 (0.85-1.40) Primary or no education 1.22 (0.78-190) 1.08 (0.82-1.44) 1.16 (0.79-1.71) 1.13 (0.89-1..42) Other 1.50 (0.70-3.22) 0.85 (0.53-1.37) 0.53 (0.25-1.16) 1.61 (1.12-2.32)b aOR: adjusted odds ratio; CI95%: confidence interval of 95%. a Odds ratios adjusted by all explanatory variables and age. b p <0.05. c p <0.01. d p <0.001. Table 4 shows the association between poor mental health, Gender and social class differences in mental health employment status and domestic and family characteristics stratified jointly by gender and social class. Concerning employ- In line with the results obtained in previous studies and ment status, unemployment was related to poor mental health in the context of the current economic crisis that started in independently of the gender and the social class. Also, a rela- 2009,8,13,29 the prevalence of poor mental health was higher tionship was found between the dedication to domestic work, among women regardless of the social class they belong to, show- other working situations and poor mental health in both genders ing that gender differences in mental health still exist in Spain. in manual social classes. Regarding family and domestic charac- Also, in previous studies carried out with ENSE-2006 data, lower teristics, an association between poor mental health and being prevalence among men and slightly higher prevalence among widowed, separated or divorced (aOR = 2.00; 95%CI: 1.22-3.29), women were observed,8,9 providing the results of the present doing the household work by themselves (aOR = 1.90; 95%CI: study new evidence in favour of the proposal by Bartoll et al.10 1.23-2.95) and shared (aOR = 1.58; 95%CI: 1.01-2.45) was only regarding a possible pattern of rising prevalence of poor men- found among women from nonmanual social classes, not being tal health among men and slight decrease among women in a found any statistically significant relationship in the sub-analysis context of economic crisis. This pattern could be related, as it carried out in working population with marital status married or was proposed by this study, to different factors being some of cohabiting (see Table I, online Appendix). Meanwhile, only among them the employment status, the social class and the educational men that belong to manual social classes an association between level. being the main breadwinner and poor mental health was observed Regarding social class, the prevalence of poor mental health (aOR = 1.23; 95%CI: 1.04-1.45). was higher among people belonging to the manual social classes in both genders. These results are consistent with previous findings that in both the general population and the working population, belonging to a manual social class and experiencing economic hard- Discussion ship were potential determinants of poor mental health.15,16,30,31 Also, the present results reinforce the hypotheses that taking polit- This study, based on a representative sample of the Spanish pop- ical measures in order to reduce social inequalities could have a ulation, shows new evidence of the existence of both gender and positive impact on mental health, as was suggested in previous social class differences in mental health, and also provides evidence studies.32,33 that these differences are related to family and household charac- As for the factors associated to poor mental health, differences teristics. As far as we know, this is one of the few population based both in function of gender as in function of social class were found. studies about mental health carried out in the south of Europe that These results are consistent with the hypothesis that in a coun- considers the work situation inside and outside the home setting. It try like Spain with traditional family policies, the variables related also improves on some of the limitations in previous studies by tak- to the paid work and the economic support of the family unit are ing into account both the influence of the social class and the gender mainly related to men’s mental health, and the socioeconomic vari- perspective on health, and considering the allocation of household ables and those related to the domestic work with the women’s work. mental health.17,22,33–35 240 J. Arias-de la Torre et al. / Gac Sanit. 2019;33(3):235–241 Mental health, employment status and household and family health in the Spanish population, especially considering that poor characteristics mental health is a possible factor associated with institutionali- zation. Among the different considered work situations, and in line with Finally, it should be mentioned the limitation related to the the results of previous studies,24,36 there was a higher prevalence of inclusion in the models of the interactions between gender and poor mental health among unemployed people regardless of gen- social class on one hand, and between family roles and employment der and social class, and lower prevalence among the employed. status on the other. Despite the stratification made in our study, the These results could be explained through the higher financial stress inclusion of these interactions in future research, and also have in and uncertainty about the future that the unemployed suffer with mind the possible colinearity between adjustment variables, could respect to the employed.34 allow to have a wider view of the relationship between these factors Regarding the employment status, the results suggest that its and mental health. effect on mental health could be influenced by social class, since unemployment was related to poor mental health regardless of social class, while the dedication to household work and other Conclusions working situations have only been associated with mental health problems among people belonging to the manual social classes. The The results of the current study show that in Spain there are results of the present study suggest a new hypothesis pointing up still gender and social class differences in mental health, and that to the mediating effect that the social class could have in the impact these differences could be related to the family role separation and of different work situations on mental health. In this way, belong- sexual division of work. ing to a nonmanual or a manual social class, could determine the Regarding the variables associated with poor mental health, the impact that the different work situations have on mental health, work situation could be a factor which is associated with mental as has been hypothesized in some studies.36 Further studies are health to a greater extent among people belonging to the manual necessary in order to clarify the relation between social class, work social classes. situation and mental health. Men who report being the main breadwinner are more likely Moreover, taking into account the household and family charac- to have poor mental health than men who do not report that role, teristics, it was observed that, when considering men and women mainly in the manual social classes. Women who report primary in the complete sample, men were more likely to report being the responsibility for household chores are more likely to have poor main breadwinner while women were more likely to report being mental health than women who do not report that role mainly in primarily responsible for household chores. These results, in line the nonmanual social classes. with previous studies,20,22,32,33 show that there still exists a gender All the results suggest that in addition to gender and social class, separation in work and family roles in Spain that assigns the role of family roles and the work done outside and inside the household economically supporting of the family unit to men, and the role of and family environment could constitute a source of inequalities in worker in household and family environment to women. Regarding mental health. family roles, several studies have shown a relationship between them and mental health problems.20,30,37 Also, and in spite of not Editor in charge having found any statistically significant relationship, the results found in our study in the sub-sample of working population with María del Mar García-Calvente. married or cohabiting marital status (see Table I, online Appendix) were congruent with the results found in these studies as well as in the whole study population. These results suggest that the main Transparency declaration breadwinner role could be associated with mental health problems through the stress and the worry about earning enough to cover The corresponding author on behalf of the other authors guar- household expenses that this role can have in some people. This antee the accuracy, transparency and honesty of the data and stress may be higher among the people that belong to manual social information contained in the study, that no relevant information classes in a crisis situation like the current one in Spain. Addition- has been omitted and that all discrepancies between authors have ally, the role of household and family workers assigned to women been adequately resolved and described. in a country with traditional family policy such as Spain, could be related to poor mental health, as a result of the overload that these chores could pose when they are added to paid work,21,22,34 being adequate in future studies to take into account the interac- tion between the employment status and the family roles in order What is known about the topic? to determine this relationship. Several studies have demonstrated that gender and social Limitations class are factors close related to mental health. Other factors possibly related to mental health as the family roles and the Regarding the limitations of the current study, a cross-sectional employment status inside the household have been less stud- design has been used. This design does not allow for establishing ied. the direction of the causal relation of the obtained results. How- ever, and following the results obtained in previous studies with What does this study add to the literature? relatively similar goals,20,21,30 our results could serve as starting Gender and social inequalities in mental health in Spain still point for future studies suggesting the hypothesis that the work exist. Family roles influences mental health differently accord- situation and the family and household variables could have an ing to sex and social class, being the work done outside and influence on mental health, and not the contrary. Moreover, among inside household a source of inequalities in mental health. the ENSE-2012 participants, the institutionalized population is not included.25 It would be appropriate to consider this type of popula- tion in order to get a more adjusted representation of poor mental J. Arias-de la Torre et al. / Gac Sanit. 2019;33(3):235–241 241 Authorship contributions 13. Malmusi D, Vives A, Benach J, et al. Gender inequalities in health: exploring the contribution of living conditions in the intersection of social class. Glob Health Action. 2014;7:23189. All authors conceived the design of the study. V. Martin, 14. Borrell C, Artazcoz L. [Inequalities gender on health: challenges for the future]. L. Artazcoz and A.J. Molina supervised all aspects of job per- Rev Esp Salud Publica. 2008;82:245–9. formance. J. Arias-de la Torre conducted the statistical analysis. 15. Mackenbach JP, Stirbu I, Roskam A-JR, et al. Socioeconomic inequalities in health in 22 European countries. N Engl J Med. 2008;358:2468–81. J. Arias-de la Torre, A.J. Molina and T Fernández-Villa wrote the 16. Muntaner C, Borrell C, Benach J, et al. The associations of social class and social manuscript. All authors have critically reviewed and agreed this stratification with patterns of general and mental health in a Spanish population. final version of the manuscript. Int J Epidemiol. 2003;32:950–8. 17. Borrell C, Muntaner C, Benach J, et al. Social class and self-reported health sta- tus among men and women: what is the role of work organisation, household Funding material standards and household labour? Soc Sci Med. 2004;58:1869–87. 18. Hankivsky O, Christoffersen A. Intersectionality and the determinants of health: a Canadian perspective. Crit Public Health. 2008;18:271–83. None. 19. Hankivsky O. Women’s health, men’s health, and gender and health: implica- tions of intersectionality. Soc Sci Med. 2012;74:1712–20. Conflicts of interest 20. Arcas MM, Novoa AM, Artazcoz L. Gender inequalities in the association between demands of family and domestic life and health in Spanish workers. Eur J Public Health. 2013;23:883–8. None. 21. Artazcoz L, Cortès I, Borrell C, et al. Social inequalities in the association between partner/marital status and health among workers in Spain. Soc Sci Med. 2011;72:600–7. Appendix A. Supplementary data 22. Artazcoz L, Cortes I, Puig-Barrachina V, et al. Combining employment and family in Europe: the role of family policies in health. Eur J Public Health. Supplementary data associated with this article can be found, in 2014;24:649–55. 23. Arias-de la Torre J, Artazcoz L, Molina AJ, et al. Inequalities in mental health the online version, at doi:10.1016/j.gaceta.2017.11.005. in the working population of Spain: a National Health Survey-based study. Gac Sanit. 2016;30:339–44. References 24. Artazcoz L, Benach J, Borrell C, et al. Unemployment and mental health: under- standing the interactions among gender, family roles, and social class. Am J Public Health. 2004;94:82–8. 1. World Health Organization. Mental Health Action Plan 2013-2020. WHO Libr 25. Instituto Nacional de Estadística. Ministerio de Sanidad y Consumo. Encuesta Cat DataLibrary Cat Data. 2013:1–44. Nacional de Salud 2011-2012. Madrid: INE; 2013. p. 1-12. 2. Whiteford HA, Degenhardt L, Rehm J, et al. Global burden of disease attributable 26. Sánchez-López MP, Dresch V. The 12-item General Health Questionnaire (GHQ- to mental and substance use disorders: findings from the Global Burden of 12): reliability, external validity and factor structure in the Spanish population. Disease Study 2010. Lancet. 2013;382:1575–86. Psicothema. 2008;20:839–43. 3. Gènova-Maleras R, Álvarez-Martín E, Morant-Ginestar C, et al. Measuring the 27. Goldberg DP. Manual of the General Health Questionnaire. NFER Publ. 1978. burden of disease and injury in Spain using disability-adjusted life years: an 28. Domingo-Salvany A, Bacigalupe A, Carrasco JM, et al. Propuestas de clase social updated and policy-oriented overview. Public Health. 2012;126:1024–31. neoweberiana y neomarxista a partir de la Clasificación Nacional de Ocupaciones 4. Wittchen H-U, Jacobi F. Size and burden of mental disorders in Europe — 2011. Gac Sanit. 2013;27:263–72. a critical review and appraisal of 27 studies. Eur Neuropsychopharmacol. 29. Artazcoz L, Borrell C, Benach J. Gender inequalities in health among workers: the 2005;15:357–76. relation with family demands. J Epidemiol Community Health. 2001;55:639–47. 5. Wittchen HU, Jacobi F, Rehm J, et al. The size and burden of mental disorders 30. Barbaglia MG, Have ten M, Dorsselaer S, et al. Negative socioeconomic changes and other disorders of the brain in Europe 2010. Eur Neuropsychopharmacol. and mental disorders: a longitudinal study. J Epidemiol Community Health. 2011;21:655–79. 2015;69:55–62. 6. Alonso J, Angermeyer MC, Bernert S, et al. Prevalence of mental disorders in 31. Linander I, Hammarstrom A, Johansson K. Which socio-economic measures are Europe: results from the European Study of the Epidemiology of Mental Disor- associated with psychological distress for men and women? A cohort analysis. ders (ESEMeD) project. Acta Psychiatr Scand. 2004;109:21–7. Eur J Public Health. 2015;25:231–6. 7. Haro JM, Palacín C, Vilagut G, et al. [Prevalence of mental disorders and 32. Bambra C, Pope D, Swami V, et al. Gender, health inequalities and welfare state associated factors: results from the ESEMeD-Spain study]. Med Clin (Barc). regimes: a cross-national study of 13 European countries. J Epidemiol Commu- 2006;126:445–51. nity Health. 2009;63:38–44. 8. Bones Rocha K, Pérez K, Rodríguez-Sanz M, et al. [Prevalence of mental health 33. Palència L, Malmusi D, De Moortel D, et al. The influence of gender equality poli- problems and their association with socioeconomic, work and health vari- cies on gender inequalities in health in Europe. Soc Sci Med. 2014;117:25–33. ables: findings from the Spain National Health Survey]. Psicothema. 2010;22: 34. Artazcoz L, Borrell C, Benach J, et al. Women, family demands and health: the 389–95. importance of employment status and socio-economic position. Soc Sci Med. 9. Ricci-Cabello I, Ruiz-Pérez I, Plazaola-Castaño J, et al. [Mental disease, exist- 2004;59:263–74. ence of diagnostic, use of psychotropic medication: differences by autonomous 35. De Moortel D, Palència L, Artazcoz L, et al. Neo-Marxian social class inequalities communities under the national health survey 2006]. Rev Esp Salud Publica. in the mental well-being of employed men and women: the role of European 2010;84:29–41. welfare regimes. Soc Sci Med. 2015;128:188–200. 10. Bartoll X, Palència L, Malmusi D, et al. The evolution of mental health in Spain 36. Hoven H, Siegrist J. Work characteristics, socioeconomic position and health: a during the economic crisis. Eur J Public Health. 2014;24:415–8. systematic review of mediation and moderation effects in prospective studies. 11. Dagher RK, Chen J, Thomas SB. Gender differences in mental health outcomes Occup Environ Med. 2013;70:663–9. before, during, and after the great recession. PLoS One. 2015;10:e0124103. 37. Bambra C, Eikemo TA. Welfare state regimes, unemployment and health: a com- 12. Riumallo-Herl C, Basu S, Stuckler D, et al. Job loss, wealth and depression parative study of the relationship between unemployment and self-reported during the great recession in the USA and Europe. Int J Epidemiol. 2014;43: health in 23 European countries. J Epidemiol Community Health. 2008;63:92–8. 1508–17.