Gender Differences in the Prevalence of Depression among the Working Population of Nepal

Objective To estimate the prevalence of depression in the working population; to examine if any gender disparity prevails; and to determine the sociodemographic mediators of depression. Methods Data from previous research was retrieved for this study. Only paid workers were selected (n=160). Sociodemographic variables including prevalence of moderate depression were compared between the genders using Chi square test. Significant variables were subject to logistic regression. Validated Nepali version of the Beck Depression Inventory scale (BDI-Ia) was used to determine depressive symptoms with a cutoff score of ≥20 considered as moderate depression. Result The overall prevalence of moderate depression was 15%, with higher prevalence among working women compared to men [χ2 (df) = 6.7(1), P=0.01], those practicing religions other than Hinduism [χ2 (df) = 5.5(1), P=0.01], those educated up to primary school compared to other education criteria [χ2 (df) = 9.4(4), P=0.03], those having vitamin D deficiency compared to others [χ2 (df) = 8.5(3), P=0.03], and sedentary lifestyle compared to active lifestyle [χ2 (df) = 6.7(1), P=0.009]. The OR (95% CI) for moderate depression was significantly higher in women than in men [3.2 (1.1-9.6), P= 0.03] and sedentary lifestyle [2.9(1.1-8.2), P= 0.04] even after adjusting for confounding variables. Conclusion Working women have increased odds of depression compared to men. Among various characteristics, sedentary lifestyle was the most important causative factor for depression among women.


Introduction
Over the past decade, there has been a steady increase in the proportion of women joining the labour force and at present Nepal has the third highest women labour force in the world [1]. While this participation of women in paid work brings economic empowerment, it brings forth a new challenge of balancing family and work. The pressure to fulfill each responsibility can act as a mediator of stress and mental disorder, the most common one being depression.
Numerous studies have acknowledged the relationship between depression and work [2][3][4]. The prevalence of depression in work place has been steadily increasing and was found to vary across occupations. Depression is highly prevalent in Nepal accounting for the second highest rate of "disability adjusted life years" in the world [5]. However, we have insufficient data regarding the prevalence of depression in the Nepalese working population [6,7]. Employed people often hide depression due to the presumption that the prejudice surrounding depression may cost them their job. The fear of being labeled as "mad" by the society can be taunting and even more so in women who are considered inferior to men in our patriarchal culture [8]. Furthermore, somatization may render the person unaware of the hidden mental disorder making them seek help for these symptoms alone leaving depression untreated [9,10]. Indeed, depressive symptoms are significantly associated with physical work-related symptoms in indoor workers [11] with musculoskeletal symptoms in health social care workers [12] and with skin symptoms in health care workers [13].
In this study, we estimate the prevalence of depression in the working population in Nepal; examine if there is any disparity in the prevalence of depressive symptoms between the genders; and determine sociodemographic mediators of depressive symptoms between them. We have used the term "working" to denote those individuals who hold paid employment.

2.1.
Sample. This paper is based on the data obtained from previous study [14] done in B.P. Koirala Institute of Health Sciences, Dharan, Nepal, in the year 2017. Only those individuals holding paid employment were selected while excluding people who were unemployed, retired, and housewives. We also included data from those participants who had met the exclusion criteria in the previous study (hypertensive, diabetic participants) but were engaged in paid employment ( Figure 1). Ethical approval was received from Institutional Review Committee of B.P Koirala Institute of Health Sciences.

Study Variables
3.1. Sociodemographic Characteristics. The study population was divided into two main groups: working men and working women. Sociodemographic variables were compared between the two groups (  [18], and presence of physical comorbidities (none, endocrine disorders, vitamin D deficiency, and pain) ( Table 1). The details of vitamin D estimation have been described in previous literature [14].  [20].

Statistical Analysis.
Data was entered and analyzed using Statistical Package of Social Science (SPSS) version 11.5. The sociodemographic characteristics of the study sample were expressed using descriptive statistics. Statistical comparison of sociodemographic characteristics among working men and working women was done using 2 test. A probability P<0.05 was considered statistically significant. Sociodemographic characteristics were also compared among participants with BDI(Ia) ≥ 20 and < 20 using 2 test. Those found to be statistically significant were further subject to logistic regression to calculate adjusted odds ratio (OR) and 95% confidence interval (CI) for BDI(Ia) ≥ 20 [moderate depression] and < 20). 2 test was also performed between gender and moderate depression using control variables to find which characteristic of working women and men had higher prevalence of moderate depression. The combined effect of gender and lifestyle on presence of moderate depression was analyzed using relative excess risk due to interaction (RERI) and attributable proportion due to interaction (AP). We calculated RERI and AP using the formulas RERI=OR11−OR10−OR01+1 and AP=(OR11-OR10-OR01 +1)/OR11 [21]. Regression coefficients and the asymptotic covariance matrix from multinomial logistic regression analyses were used to calculate the 95% CIs using delta method described by Anderson [22]. We considered Interaction to be significant if the 95% CI did not include 0.

Result
The sociodemographic variables between working men and women are presented in Table 1.
Out of 160 participants, 49% were men and 51 % were women. Both groups were comparable without any statistically significant difference in the following categories: age, ethnicity, religion occupation, education, BMI, marital status, socioeconomic status, sedentary lifestyle, and presence of physical comorbidity. However, we found higher prevalence of moderate depression among working women compared to working men (P = 0.01) ( Table 1). The median depression BDI 1(a) score was significantly higher in women than men ( Figure 2).
The overall prevalence of moderate depression was 15%, with higher prevalence among working women compared to men [ 2 (df) = 6.  Table 2).
The frequency of moderate depression among gender characteristics found higher prevalence of moderate depression among women of age group 41-65 years compared to men of the same age group  Table 3).

Discussion
Depression is a major challenge to the economic development of the country as it primarily affects the working population [23]. Depression leads to decline in cognitive abilities and interpersonal skills, lack of motivation, and disregard to safety measures [24] which translates to decreased work performance, and absenteeism directly impacting the labour force market. Moreover, untreated depression can lead to unemployment, poverty, substance abuse, and suicide which add further to the economic burden of the disorder [25].
In our study, we found significantly higher prevalence of moderate depression among working women compared to working men. This was in agreement to a study conducted in Psychiatry Journal 5 Italy where female radiologists had a higher rate of depression disorder compared to male colleagues [26]. Furthermore, when marital status was examined in our study, married women had significantly higher prevalence of depression compared to married men. The same did not hold true for unmarried and marital discord categories. In the context of Nepalese society, men are often exempt from domestic responsibilities while women are expected to look after the family and perform household chores even if they are employed [27,28]. The additional burden of unpaid work means women have to work longer hours than men which can cause mental and physical exhaustion. Socioeconomic status (SES) did not affect the prevalence of depression. This finding was different from previous 6 Psychiatry Journal   a Adjusted for religion, education, and physical comorbidities. b Relative Excess Risk due to Interaction (RERI) = (OR11 -OR10 -OR01 + 1). c Attributable proportions due to the interaction (AP) = (OR11 -OR10 -OR01 + 1)/OR11) were derived from the above shown odds ratios (ORs) from logistic regression adjusted for religion, education, and physical comorbidities. studies reporting lower socioeconomic status (LSES) as a risk factor for depression [29][30][31]. However, on examining each socioeconomic class among gender category, we observed women from lower middle and lower classes had significantly greater prevalence of moderate depression compared to men of the same socioeconomic status. This was similar to a study conducted in Belgium reporting higher effects of SES on depressive symptoms among women [31]. The reason behind such disproportional effect of socioeconomic status on gender remains unresolved in our study. We can only postulate it may be due to the way men and women handle everyday stressors as economic hardships associated with LSES. The prevalence of moderate depression was higher among individuals with sedentary lifestyle compared to active people. When examining genders we found that sedentary women had significantly higher rates of depression irrespective of which SES they belonged to. Another important finding was the effect of religion on the prevalence of depression. Those practicing Hinduism was buffered against depression. The mechanism behind such occurrence needs further insight into the religious and cultural aspects between Hinduism and other religions and how they affect the minds of the people.
People suffering from vitamin D deficiency had greater prevalence of moderate depression than those with other comorbidities. However, on comparing between the genders, the effect of vitamin D did not vary; instead we found women suffering from endocrine disorders had higher occurrence of depression compared to men with the same.
Logistic regression showed significant association of gender and lifestyle with moderate depression even after adjustment for confounding variables. The odds of depression were 3.2 times higher in women than in men and 2.9 times higher in people with sedentary lifestyle compared to those with active lifestyle. Among various characteristics of women, sedentary lifestyle was significantly associated with depression. The combined effect of gender and lifestyle on presence of moderate depression was additive but not significant. Our study highlights the need to identify factors including workplace dynamics contributing to gender disparity in the prevalence of depression. Furthermore, it urges the need to initiate workplace screening and intervention programs to prevent and treat depression.

Conclusion
We found the gender of women and sedentary lifestyle are independently associated with higher risk of moderate depression. Working women have threefold greater risk of developing depression than working men.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors declare that they have no conflicts of interest.