Study setting

This study was conducted in outpatient section of Department of Obstetrics and Gynecology of a tertiary care hospital situated in rural part of Mandya district, Karnataka state, India from June to Nov 2012. The studied hospital is attached to a rural medical college recognized by Medical Council of India with teaching facility for both under and post graduates in obstetrics and gynecology. It caters to the need of nearby villages and talukas and linked with many government health programmes for the benefit of rural poor.

The total population of Mandya district is 1.92 million. Situated in southern part of Karnataka state, it is a prominent agricultural district blessed with the irrigation waters of river Cauvery and Hemavathi [18]. More than half of the population is involved in agriculture. Most of the industries that are run in the district depend on agriculture for their raw material. Sugar Mills, Jaggery making units, Rice Mills are the prominent industries of this district [18].

Study design and sample

Cross sectional study design was adopted to accomplish the study objective. All the women who came for postnatal follow up to studied hospital from June to July 2012 formed the sampling frame. We decided to study at least 96 postnatal mothers based on minimum estimated PND prevalence of 10% with 6% precision, 95% confidence level [19]. All the women who came for follow up from 4^th to 10^th week of postnatal period from June to Nov 2012 were included. Women with acute severe illness or cognitive impairment or not willing to consent for voluntary participation were excluded. In a day maximum of two postnatal women, satisfying inclusion and exclusion criteria were selected randomly from the outpatient list of Department of Obstetrics and Gynecology.

Study tools

A pretested semi-structured interview schedule was used to elicit information pertaining to socio-demographic profile, obstetric and pregnancy outcome. Anonymity of the study participants was maintained to enhance the participation rate and also to ensure confidentiality.

Following details were sought after; socio-demographic profile: age, religion, education, working status, economic status, husband’s occupation and family type; obstetric details: duration of married life, parity, any complications in the last pregnancy and known medical or psychiatric disease; pregnancy outcome: gender of the baby, mode and type of delivery.

Woman’s literacy status was classified as literate (if she can read and write with understanding in any language) or illiterate (can neither read nor write or can read but cannot write in any language) and literacy level was the highest level of education completed. (Census India 2011) [20].

Working status of the woman was categorized as employed (engaged in economically productive work) or unemployed. Similarly, the family structure was categorized as nuclear (consisting of a couple and/or their dependent children) or joint (a number of married couples and their children where all men are related by blood, who live under one roof, eat from common kitchen, hold property in common, and participate in common family worship) [21]. And the type of delivery was stratified as preterm (live birth before 37 completed weeks of gestation) or term (live birth between 37 and 42 weeks of gestation) delivery [22].

In India, public distribution system issues ration card to families by assessing their economic status. In this study the type of ration card possessed by the woman was taken as proxy indicator of her economic status i.e. yellow and blue colored cards for below and above poverty line families, respectively.

Edinburgh postnatal depression scale (EPDS): It is a 10 item self report scale based on 1 week recall, designed to screen PND in community [23]. Each item is rated from 0 to 3, yielding a total score of 0–30. Seven of its items are reverse-scored. In India, the Kannada (local language) version of EPDS has been validated to detect antenatal depression, and found to have a sensitivity of 100% and specificity of 84.9%, at a cut-off score of ≥13 [24]. For this study, a cut-off score of ≥13 on Kannada version of EPDS was used to diagnose PND. It was pre tested by the second author on 10 postnatal mothers under the guidance of a qualified psychiatrist to ensure that questions were easily understood by mothers and responses were correctly interpreted by the author. Psychiatrist approved the pilot study and rest of the interviews were conducted by second author only to avoid inter observer bias. Woman with EPDS score of ≥13 was referred to a psychiatrist for further evaluation.

Ethical approval

Institution review board and ethics committee of Adichunchanagiri Institute of Medical Sciences, Balagangadharanatha Nagara, Mandya District, Karnataka, India, approved the study protocol. Informed written consent was taken from all the respondents for voluntary participation in local language (kannada).

Main findings

Substantial proportion (31.4%) of lactating women who came for follow up, from 4 to 10 weeks of post-partum, to our study setting is at risk of PND. It is associated with poverty, birth of baby girl and occurrence of complications in pregnancy or known medical illness.

Interpretation

Pregnancy and childbirth are unique life events and cannot be reduced to primarily biological events since the social and cultural contexts are central to the subjective and collective experiences of women. [25] Women’s age, literacy, social economic status, religion and culture do have a significant influence on pregnancy experiences and her mental health. The present study was an attempt to explore the aforesaid and biological risk factors of PND.

Reported prevalence of PND in this study was 31.4% (95% CI 22.7–41.4%) and is quite high. Other hospital and community based studies in India have reported lower prevalence of PND ranging from 11% to 23% [5,6,7,8,26]. Very wide range of PND prevalence has been reported in studies from China (11%), United Arab Emirates (15.8%), Zimbabwe (16%), Brazil (20.7%), South Africa (34.7%), and Pakistan (40%) [27,28,29,30]. The present study may have overestimated the actual PND prevalence as EPDS is a screening tool but not confirmatory. Other reasons for variations in PND prevalence could be attributed to study tool and technique differences, socio-cultural norms and poverty levels.

Poverty was an independent predictor of PND in our study. Poverty lead financial problems may act as an additional stressor, especially at the juncture where another family member is being added [26]. This is corroborated by the reports of studies from low and middle income countries [7, 16, 31, 32]. However, such association was not seen in other studies [11,26,30].

Although, type of family, woman’s and husband’s occupations showed significant association with PND in univariate analysis but binominal logistic regression ruled out their effect. Was it a serendipitous finding or these women had other protective influences, e.g. better social-support and marital quality etc, needs further investigation [26]. Some studies [8,33] have found an association of PND with age and working status of the mother.

Present study did not show significant association between PND and planning of pregnancy, mode and type of delivery. Similar findings are reported by Hegde S et al [26] and Milgrom J et al [34]. On the contrary, unplanned and unwanted pregnancies, operative procedures and difficult labor have been reported as risk factors for PND in previous studies [16,27]

In India preference to male child is deeply entrenched. In Indian culture, girl goes to her husband’s house, changes her last name to husband’s last name and expected to continue his family lineage. Whereas male stays with parents till death and he will continue their family pedigree, support parents in their old age, conduct funeral rites and inherit ancestral property. Birth of baby girl is considered as a family and social stressor in Indian societies and hence could be a strong predictor of PND [26]. The present study and other Indian studies from Karnataka [26], Goa [7] and Tamilnadu [8] corroborate this gender bias.

Occurrence of complications or medical illness during pregnancy was also the predictors of PND in this study. However, none of the mother reported any psychiatric disease during pregnancy or in the past. This could be due to lack of knowledge or cultural norms affecting womens’ reporting of their symptoms [13,14]. Various studies have reported depression during pregnancy is an important predictor of PND [7,8,9].

There is agreement and disparity in the prevalence and various predictors of PND across India and the globe. However, it is strongly recommended that PND screening should be an integral part of postnatal care owing to simplicity of screening by EPDS and to avert devastating effects on mother and child.

In most of the rural India, postnatal care is being rendered by Auxiliary Nurse Midwife (ANM) for every 5,000 rural populations [35]. Community health volunteers called Accredited Social Health Activists (ASHAs), for every 1000 rural populations, have been engaged under the National Rural Health Mission to facilitate the ANM to render maternal and child health services and establish a link between community and the healthcare system [36]. Feasibility trials for screening PND among rural Indian women by these grass root level workers are needed.

Limitations

Unfortunately women with abortion or stillbirth did not come for follow up during the study period. Therefore we could not study their association with PND. We also did not consider other factors such as quality of social support, domestic violence, gender preference, women empowerment etc. EPDS is a screening tool for PND and not confirmatory. In rural parts of Karnataka state 64.7% of the women receive postnatal care within two weeks of delivery [37] and relatively a small sample was studied in our study. Therefore, results would apply to women attending hospitals/clinics and cannot comment on those who do not receive postnatal care. Nonetheless, the present study gives valuable hints to practicing obstetricians and primary care physicians in resource constrained rural India for early identification and apt preventive intervention for women at risk of PND.

Details of ethics approval

Institution review board of Adichunchanagiri Institute of Medical Sciences, Balagangadharanatha Nagara-571448, Karnataka, India approved the study protocol in April 2012.