Hunt et al [ 26 ] showed that the risk of death from coronary heart diseases is lower in men with feminine traits; however, femininity is related to worse mental health [ 27 ]. One potential reason for more depression in feminine individuals either male or female could be the passive responses they adopt in responding to life challenges [ 25 ] by focusing on personal concerns instead of adopting a problem solving approach.
In contrast, androgynous people adopt a more interactive response. Depending on the situation they will mobilize their masculinity and will be assertive or use their femininity trait and show expressiveness and will be yielding.
Flexibility in behavioral responses to stressful conditions results in being more adaptive and therefore more likely to enjoy better mental health [ 24 ]. This theoretical conceptualization has been supported by research. In adults, androgyny has been associated with better mental health [ 28 ] and fewer risky health behaviors [ 29 ]. Androgynous Spanish [ 21 ] and Canadian [ 30 ] older adults reported better self-rated physical health, life satisfaction, and mobility.
Most existing research on gender and aging has focused on health impacts of gender roles in separate populations of men [ 31 , 32 ] or women [ 33 ] or has looked into changes in gender roles as people age, not specifically on gender-health relationships [ 34 , 35 ]. The androgyny model for older adults is seldom considered. The main objective of the few studies [ 21 , 22 , 30 ] that have explored the androgyny models in older adults populations was validation of the gender role measures and not their health impacts.
Research in this area is growing but in spite of widespread worldwide variations in gender equality, as demonstrated by the United Nations Gender Inequality Index [ 36 ], little international research has been conducted. The International Mobility in Aging study IMIAS provides the opportunity to examine the associations between gender roles and depression in five international samples of older adults.
We hypothesized that regardless of their living environment and biological sex, older adults endorsing the androgynous gender role will have lower prevalence of depression due to better psychological adaptation and higher competence. The objective of this study was to examine differences in prevalence of depression in older adults according to their gender roles.
A secondary objective was to assess if gender roles constitute an independent risk in these older men and women. This is a prospective study conducted in five cities: Questionnaires and instructions for measurement procedures were available in five languages English, French, Albanian, Spanish, Portuguese and were administrated by trained interviewers. For this study, we made use of baseline data collected in The study population was composed of community—dwelling men and women aged 65—74 years.
We aimed for men and women at each city and stratified the sample by sex. Due to restrictions imposed by Canadian ethics committees, hindering direct contact with potential participants, Canadian potential participants were invited by a letter from their primary care physician and asked to contact our field coordinator if they would like to participate in the study.
To identify the potential participants, random samples were drawn from family practice lists of patients in the 65—74 age group.
The family practices participating in the study came from all family medicine teams covering populations of Kingston and Saint-Hyacinthe. In Saint-Hyacinthe the sample was stratified by neighbourhood, while in Kingston due to ethical issues this stratification was not possible. In Tirana, Manizales, and Natal, participants were randomly selected from the population in the 65—74 age group registered at neighbourhood health centres.
Of a total 1, participants, 1, answered the Bem sex roles questionnaire and had complete data in all considered covariates and were included in the analysis. All study participants provided written informed consent and were told they could withdraw at any time. This scale is a screening tool comprised of 20 items related to depressive symptoms such as mood, somatic symptoms, interactions with others, and psychomotor functions and has been validated in French [ 39 ], Brazilian Portuguese [ 40 ], and Spanish [ 41 ] older adults populations as well as in southern and eastern Mediterranean regions [ 42 , 43 ] and low income settings [ 44 ].
We utilized the established cut-off point of 16 as suggestive for depression [ 38 ]. As per precedents [ 21 , 22 ], we used a short version of the BSRI composed of 12 items. The validity and internal reliability of this short version of the BSRI in IMIAS participants have been examined via confirmatory factor analysis and showed to be high.
The details of psychometric properties of BSRI items and the full validation methodology have been described in a separate paper under review [ 45 ]. The common classification method of median split [ 46 ] was used to categorize the study population into four gender role groups as per existing precedents [ 21 , 22 , 30 ].
First, the median of distributions of masculinity and femininity scales were established. Then femininity and masculinity scores were compared to the median. The distributions of the masculinity and femininity scores were statistically different across research cities, thus the reliability of results could be influenced by culturally derived differences. In order to control for these differences across cities, city-specific medians of the distributions of femininity and masculinity were used as cut-off points.
Factors with potential confounding effects on the relationship between gender roles and depression were included in multivariate analyses. Women consistently report more depression symptoms and the probability that a woman expresses feminine gender role is higher, therefore biological sex was the most important potential confounder.
Age, marital status, education and income are documented risk factors for depression in older adults [ 10 ] and they could also be associated with gender roles. Education was assessed based on the highest level of schooling completed by participants. Responses were grouped into three categories: Self-reported health SRH which has been used extensively as an indicator of general health [ 30 , 47 , 48 ] was used for the same purpose in this study.
SRH has been shown to be a good predictor of mortality in older adults [ 49 , 50 ]. Our measure of SRH encompassed four categories: The number of self-reported chronic conditions has been also used as an indicator of health status and a potential confounder [ 10 ].
Statistical analysis To define the four Bem sex roles types we used the femininity and masculinity scores obtained from confirmatory factor analyses of the BSRI and applied city-specific median based cutoffs as explained above. Distributions of all variables in total and across sex and gender roles groups were estimated and differences were statistically examined using Chi-square and ANOVA tests where appropriate. We used Poisson regressions with robust variance to estimate the prevalence ratios PR of depression in different gender role groups masculine role as the reference category adjusting for sex and all other potential confounders [ 51 , 52 ].
All statistical analyses were conducted with SPSS v Results Participants consisted of men and 1, women with an average age of More than half of the participants The prevalence of depression was higher in women than in men