Mental health status of women in reproductive age group and its determinants in rural Odisha, India using GHQ 12 questionnaire

Introduction

Much of the focus of decades in Indian policy, practice, and research related to women of reproductive age has been devoted to the prevention and management of risks associated with preterm birth, low birth weight, and to some extent immediate risky postpartum signs and symptoms. India's maternal mortality ratio (MMR) has improved to 103 in 2017-19, from 113 in 2016-18, while that of Odisha is reported as 126 per lakh live births. This is according to the special bulletin on MMR released by the Registrar General of India March 14, 2022 (RGI, India report, 2022).

One in three to one in five women in developing countries, and about one in ten in developed countries, have a significant mental health problem during pregnancy and after childbirth. For example, high rates of mental health problems in pregnant women and mothers have been reported in many countries in Africa such as Ethiopia and Nigeria (World Health Organization, 2008). Women, especially those living in developing countries are more exposed to risk factors, which increase their susceptibility to developing mental health problems. Some of these include poor socioeconomic status, less valued social roles and status, unintended pregnancy, and gender-based violence (Prince, M et al., 2007; Johnson AR et al., 2015).

Core features such as lethargy, disinterest, and lack of appetite are associated with motherhood itself and/or part of the gender stereotype which is inclusive of motherhood. Hence, these women are less likely to seek and receive antenatal or postnatal care or adhere to prescribed health regimens, which is subtly suggested by research (Khanna, T et al., 2021).

Poor mental health of women also affects their ability to take care of their newborns. Maternal depression in resource-constrained settings is linked directly to lower infant birth weight, higher rates of malnutrition and stunting, and higher rates of diarrheal disease, infectious illness, hospital admission, etc. It also adversely affects the physical, cognitive, social, behavioral, and emotional development of children (WHO Meeting Report, 2008). Thus, it is very conclusive that efforts to achieve Millennium Development Goal 5, “improve maternal health”, should include measures to prevent and manage mental health problems during pregnancy, after childbirth, and even generally for women in the reproductive age group (Prince, M et al., 2007).

Simply, there can be “No health without mental health” (Prince, M et al., 2007). Programs aimed at achieving Sustainable Developmental Goals 2025 should integrate mental health approaches within their strategies for improved maternal mental health. This study aims to assess the extent of the problem primarily in the rural population in an underserved state of India and in the process validate a simple questionnaire-based community tool that can be used by ground-level health workers.

Objectives

Methods

The reference period for the study was 2015 to 2017. As vulnerability for mental ill-health is high, especially during the antenatal and postnatal period, the study proposes to categorize the selection of women into those from the antenatal period (above 12 weeks of gestation); postnatal period (up until three months after delivery), and those who do not fall into either, i.e., any women between 18-49 years.

Results

The mean GHQ score among all the 1260 subjects was 10.45 (SD: 4.48). No difference was observed between all the 3 strata of women in terms of GHQ-12, the average score being <11 in all the 3 strata and the intergroup differences were not significant, p=0.778, wherein 41.4% of pregnant and lactating women and 40.7% NPNL had higher GHQ scores. This raises concerns, even though this was not statistically significant in this study (p=0.971).

Table 2 shows the total prevalence of women with higher GHQ is 41.19% (519/1260); the mean age of women respondents in total women subjects is 26 years and the mean age of antenatal women with GHQ more than 11 is on the higher side which is highly significant; the mean age of marriage is 19.94±2.56 years for antenatal women and it is statistically very significant for higher GHQ scores. All the 3 strata reported higher GHQs for lower age at marriage. The full raw data can be found under Underlying data (Kar, S., 2022). For total subjects, as well as cross-sectionally for all the 3 subgroups, the higher GHQs were associated with lower hemoglobin values suggesting anemia. Similarly, mean iron folic acid intake (in terms of days of consumption) was reported with wide standard deviations and was significant for antenatal women, with those with higher GHQ scores consuming for 62.94±49.29 days as against the current Government of India recommendation of prophylactic IFA consumption for 180 days (note: reporting data may have a lot of bias). Birth weight which was noted for the current child in the postnatal group showed lower birth weights being significantly associated with higher GHQs.

As per Table 3, the joint family showed significance as a protective factor only for mental health of pregnant women, meaning that, in India, a pregnant women is taken care of by their in-laws and that could be the reason why in this study, being in a joint family influenced good mental health for childbearing women but after the child is born women have take up tasks in the home as well as childcare. Casual employment was seen to be significant for high GHQ scores and in the total sample OR is 2.57 (1.68-3.95, <0.001) times as compared to permanently employed women. Any form of permanent employment is protective, and the results are statistically significant in the sample as a whole but not at subgroup levels. Perhaps a larger sample would be required to draw inferences.

The absence of civic amenities such as electricity in the house was significant for high GHQ and a stressor in all groups, except in pregnant women, which is perhaps because pregnant women are best supported by the family and cannot feel the stress (Blomqvist, G, & Ternald, D, 2014). Akin to the rural mindsets, the absence of toilet facilities inside the house wasn’t seen to influence women’s mental health, instead those in homes without toilets reported better GHQ 12 scores in the whole sample and only lactating women had 1.69 times high GHQ without a toilet, which was inferred to be because of the double burden of disposing of the infant’s feces. This reinforces the need for more behavior change efforts toward toilet habits in women. The absence of water inside the house was seen as a deterrent to GHQ, the highest being for lactating women i.e., 3.22 times (1.97-5.27, <0.001), which could be a result of the need for water for childcare and as in rural scenarios women usually have to fetch water.

In Table 4 certain individual characteristics were compared in the regression model. Absence of domestic violence in the last year was seen as protective and highly significant in all groups. Absence of physical or psychological violence was protective for lactating and NPNL women which was highly significant. Pregnant women show 3.02 times higher GHQ which is significant in terms of better health access and childcare while in lactating women GHQ is 6.39 times higher (2.67-15.27; p<0.001) when the access to childcare is unsatisfactory.

Not taking IFA supplements during their current pregnancy was 3 times associated with high GHQ (1.31-6.85; p<0.009) while, interestingly, in the currently pregnant or in NPNL women, IFA consumption did not come as a significant concern for pregnant or NPNL indicating a nonchalance about this essential health intervention and this can also hint at the loose links within our health system where the IFA service delivery is taken up casually by the ground-level health workers.

The health-seeking behavior factors are seen to have some relevance only in the lactating group as giving birth makes them more conscious of their child’s and their own health. Not seeking treatment is seen to have a protective effect on women (OR 0.45; 0.31-0.70; p<0.001), which hints at the reluctance of women to seek out health care services. The protective factor could also point towards the current mandate made by the government in the state encouraging home-based care both for the infant and mother. More in-depth studies are needed to draw assertive conclusions. Going by the current trends, it is also observed in the study that an assisted or cesarean delivery is 0.42 times protective against high GHQ scores (0.24-0.72; p<0.002). Due to a lack of prerequisite care and support after a natural delivery, women especially in rural areas are shying away from this very essential natural method of delivery. The full raw data can be found under Underlying data (Kar, S, 2022).

Discussion

Odisha state boasts of 30 districts, 476 sub-districts, and 51,313 villages as per Census 2011. Female literacy continues to be 64%; and in the rural population 60.7%. Hence the study population was taken from the rural parts of the state.

High GHQ scores were noted in all the 3 subgroups i.e., antenatal, postnatal and NPNL suggesting that mental ill-health exists among all these groups and that 40% higher GHQ is also noted equally for pregnant and lactating women (41.4%) and mildly less for NPNL women (40.7%).

This indicates that there is a definite burden of mental ill-health among the women of reproductive age groups in rural areas of Odisha as the mean score of the sample was 10.45±4.48, which is very near the cutoff of 11 for normal mental health. In another study among the whole population of Somerset (Goldberg DP et al., 1997) aged 16-64 years higher GHQ scores were seen almost above 35% in all age groups among women, though there was no stratification in the study into the reproductive phases as in this study.

GHQ 12 as an instrument has been widely explored to assess early psychological distress in 32% of infertile women (Oliver MI et al., 2005, 297-301); in Polish women with Polycystic Ovarian Disease (PCOD), wherein the mean GHQ scores varied from 13.4±6.5 to 15.5±8.0, hinting at mild distress; or in women with miscarriages (Souter VL et al., 2002; Kowalczyk R et al., 2012). The instrument has been used in British women, assessed through their reproductive cycle (Toffol E, Koponen P, Partonen T, 2013), in which predicted levels of depression do not seem to vary much among the three groups of teenage mothers, 20- something mothers, and mothers aged 30 and over, with the scores peaking 1-2 years postpartum. While the oldest group of mothers have a peak of about 0-1 years postpartum. This was a limitation in our study, as we could not follow up a woman cohort across the reproductive period, due to paucity of funds and also lack of feasibility. In our study too, all three cohorts showed the same stress and GHQ scores were mildly raised for antenatal period, contrary to studies in the West which show a rise in the post-natal period. Results were however akin to the population-based and controlled study in Britain (Liao T, 2003), which suggested that women are not more at risk of common mental health disorders during early pregnancy and postpartum than those who are not pregnant or who have not delivered recently. In this study, the NPNL strata of women too had the same risk of mental distress.

Since data was collected from 7 districts, a district level variation was also calculated that could offer suggestions to the district-level program managers of the Indian National Mental Health program to improve or intensify their efforts. Interestingly a district with poor HDI showed 77.8% of the women in the normal mental health category (Kandhamal) while a good performing district Kendrapada showed 74% of women with poor mental health. The report was shared with the district level authorities to further intensify their maternal health programmes. The mental health program personnel too can utilize this as baseline information to initiate some counseling and mental health promotive strategies to help the women. Again, the best district in HDI i.e., Khordha and the average performing district, Kandhamal, both had similar GHQ scores of 8.58±3.30 and 8.22±4.40 respectively to show that both districts had the best GHQ scores among all districts. The poorest GHQ score again was Kendrapada, which was among the good performing districts i.e., 12.91±3.27.

The study brings out demographic factors that may have a bearing on GHQ scores. This was discussed as a bivariate analysis for the sample as a whole and also for individual subgroups. Religion came out to be significant for all analyses but since the predominant proportion was Hindus, this needs a larger sample with proportionate representation from other religions. Castes such as Schedule Tribes (ST) were highly significant for high GHQs in all analyses and emerged as a pertinent finding which deserves attention at the policy level (Van Bussel JC, Spitz B, Demyttenaere K, 2006).

The logistic regression analysis has offered insight into factors that are protective as well as triggers for poor GHQ scores which imply poor mental health. Having a partner who is abusing alcohol, domestic abuse (by partner or others in the family), and education emerge as very significant factors for poor mental health and can be addressed at the program level. A similar study (De A et al., 2017) comparing mental health among rural and urban women of Odisha showed that urban women were more in distress due to spousal infidelity and alcoholism.

Mental health literacy and counseling especially among women in the reproductive age group should be made mandatory as was brought out by a study done in Maharashtra (Kermode M et al., 2010), Simple community friendly and tried tested tools like GHQ-12 could be used for this purpose with an inbuilt mechanism of counseling or added social support. In poor-resource settings such as India, such tools can be used by a paramedic or a lay community worker, who can assess for mental distress and then offer non-medical psychosocial counseling (PSY), as has been tested in Nepal (Markkula N et al., 2019) among socially disadvantaged women and yielded very good results.

Data availability