Received date: October 14, 2016; Accepted date: October 28, 2016; Published date: October 31, 2016
Citation: Ongeri L, Otieno P, Mbui J, Juma E, Mathai M (2016) Antepartum Risk Factors for Postpartum Depression: A Follow up Study among Urban Women Living in Nairobi, Kenya. J Preg Child Health 3:288. doi:10.4172/2376-127X.1000288
Copyright: © 2016 Ongeri L, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Pregnancy and Child Health
Introduction: Longitudinal studies that assess antepartum risk factors and outcome in the postpartum period can help provide a wealth of information in understanding maternal depression. In addition to collecting information on prevalence and correlates of antepartum depression, such studies reveal postpartum outcomes of depression as well as its risk factors while avoiding recall bias, a limitation frequently seen in cross sectional postpartum studies. Methodology: We consecutively recruited 188 adult women residing in an urban, resource poor setting and attending maternal and child health clinics in 2 major public hospitals in Nairobi, Kenya. A translated Kiswahili EPDS was used to screen for depressive symptoms at baseline assessment in the 3rd trimester and a follow up assessment at 6-10 weeks post-partum. A different questionnaire was administered at baseline to collect information on potential socio demographic and clinical antepartum risk variables. Study results: At a cut off of 13 or more on the EPDS, our study found a prevalence of 18% for antepartum depression. Associated correlates of antepartum depression were partner current alcohol use and partner wanting current pregnancy. Out of the 171 women we followed up at 6-12 weeks postpartum, 21% were found to have postpartum depression. Antenatal depression and conflict with partner were the strongest independent predictors of postpartum depression. In the adjusted analysis, the risk of having postpartum depression is increased six-fold in the presence of antepartum depression and ten times in the presence of conflict with partner. Conclusion: High rates of perinatal depression among women residing in Africa underscore the need for addressing this public health burden. Despite the comparably little emphasis on antenatal depression, antenatal depressive symptoms appear to be as common as postnatal depressive symptoms. Depression screening and psychosocial support that especially addresses conflict resolution during pregnancy should therefore be targeted for future interventions.
A greater focus on perinatal depression is steadily increasing in Sub Saharan Africa. Previously, it was thought that the existing sociocultural structures related to pregnancy and early motherhood would protect women in the region from developing perinatal depression. However, it is now evident that not only does perinatal depression exist in Africa but that the prevalence of affected women in Sub Saharan Africa surpasses the burden in the West . High rates of civil conflict and political instability, changing socio-cultural structures and the scourge of HIV in the region are strong contributors to the higher rates being seen in the region [2-6].
Maternal depression can present during pregnancy (antepartum depression) and or after delivery (postpartum depression). Unlike antepartum depression, postpartum depression has been a focus of attention and numerous studies have investigated its risk factors, prevention and care. However, emphasis on antepartum depression is warranted as it has been shown that significant risk factors for postpartum depression actually present during the antepartum period [7,8]. Furthermore, growing evidence points towards a multitude of detrimental effects related to antepartum depression on the mother, her partner and infant. Poor compliance with antenatal care as well as increased risk of developing insomnia, preeclampsia, preterm delivery and even a greater risk of the catastrophe of suicide are some of the documented adverse effects seen in mothers with depression during pregnancy [9-11]. Increased incidence of marital conflict leading to a general family disharmony is also evidenced in the homes of these mothers . Outcome studies observing effect on the infant have shown greater associations with low birth weight, smaller head circumference, lower Apgar Scores and a higher incidence of childhood behavioural disorders later in life [13-15].
Prevalence estimates of antepartum depression vary depending on methodology, period of gestation and with the region the study is conducted. A recent systematic review estimated a prevalence of 12% in the 3rd trimester for developed countries . In Sub-Saharan Africa rates ranging from 8.3% to 39% has been reported [17-20]. Estimates of postpartum depression range between 10-15% in developed countries  while higher rates are seen in low and middle income countries ranging between 15 and 57% [21-23].
Longitudinal studies that assess antepartum risk factors during pregnancy and outcome in the postpartum period are important in identifying predictive risk factors during the antenatal period for postpartum depression.
Antenatal risk factors that have been found to be associated with postnatal depression include marital discord, current or past emotional, physical, sexual or verbal abuse by partner, lack of social support, pregnancy unwanted by mother, stressful life events during pregnancy, past or present psychiatric disorders in the mother, antepartum depression, prenatal care not started until the third trimester and poor relationships of the mother with her parents and mother in law [24-26].
In addition to providing information on prevalence and risk factors of antepartum and postpartum depression, longitudinal studies that follow the same cohort of women during pregnancy right into the post-partum period have the advantage of reduced chance of recall bias which is a limitation frequently seen in cross sectional studies of postpartum mental health.
Predicting the risk of postpartum depression based on the antenatal depression correlates has scarcely been studied. This is the first Kenyan study that has measured antepartum risk factors for postnatal depression in a cohort of pregnant women residing in an urban resource poor setting in Nairobi, Kenya.
This was a longitudinal study with initial assessment of pregnant women in their third trimester and a follow up of this same cohort at 6-10 weeks postpartum.
The study was approved by the Kenya Medical Research Institute’s ethical and review committee and was carried out between March 2014 and December 2014. Participants were recruited from the maternal and child health clinics of 2 major public hospitals (Mathari Teaching and Referral Hospital and Mbagathi District Hospital) in Nairobi, Kenya. The maternal and child health clinics offer both antenatal and postnatal care including family planning and infant immunization. Majority of the women served by these two hospitals are from an urban resource poor catchment area.
Participant’s recruitment and screening
Inclusion criteria: pregnant women aged between 18 and 49 years in their 3rd trimester. As the study had a follow up time point to assess postnatal outcome, participating women had to be willing to seek postnatal care at the same facility. The pregnancy gestation was verified using the patient attendance card. A total sample size of 188 participants was conveniently recruited from the outpatient waiting area of the maternal and child health clinic of the two hospitals. Following registration at the MCH clinic the study nurses would inform attending women in their 3rd trimester of this on-going study survey. Voluntary informed consent was sought after a nurse trained on the protocol gave all explanation about the study. Those that fitted the eligibility criteria were enrolled into the study. Of the women who met the study inclusion criteria 7 refused to participate. The reasons given for refusal to participate were time constraints (4) and others (3) opted to first ask permission from their partners before enrolment. About 20 women who were over 18 years and in their 3rd trimester could not participate as they did not plan to continue with follow-up in the same facility (This being a hospital setting women preferred to attend antenatal clinic here for ease of delivery in a hospital setting but be followed up in a smaller clinic closer to their residence for vaccination and family planning during the post-partum period) - it is for this same reason that 16 participants (8.5%) were lost to follow-up.
Social, demographic and clinical history data were collected using a face to face interview with structured questionnaires. Depression levels were then assessed at baseline and again at 6-10 week postpartum follow up using the EPDS. The structured social, demographic, clinical history questionnaires and the EPDS screening tool were administered by trained nurses working at the facility. Nurses were trained to conduct the interviews as part of a larger study to see the feasibility and acceptability of integrating the EPDS into the MCH clinics.
The Edinburgh Postnatal Depression Scale [27-29] (EPDS or EDS when used in the antenatal period) is the most widely used scale to screen for antepartum and postpartum depression symptoms in low and middle income countries. It is a 10 item questionnaire, scored from 0-3 with scores ≥ to 10 indicating depression symptoms. The scale has been validated for detection of depression in both antepartum and postpartum samples  and used in many countries including Kenya . Prior to data collection the EPDS tool was translated and back translated through a rigorous process into Kiswahili language . A cut off of 13 or more for screening for antepartum depression and a cut off of 10 or more for screening for postpartum depression was used, as recommended by Murray and Cox in the assessment for minor antepartum depression and minor postpartum depression respectively [33,34].
Data analysis was done using STATA version 13 (Stata Cop LP, Texas 77845 USA. Continuous data were summarized using means (SD) or median (IQR) while categorical variables as proportions. Rates of Antepartum depression was determined by the proportion of women who screened 13 and more for probable depression on the EPDS at (P 36+). To determine the risk factors for postpartum depression, individual socio-demographic and clinical variables were analysed using logistic regression. All tests were two sided and the level of statistical significance was set at 5%.
At an EPDS score of 13 and more, the prevalence of antenatal depressive symptoms was found to be 18% while the postnatal depression rate a score of 10 or more among the women followed up was at 21%. Thirty seven per cent of the women with antenatal depression were found to have postnatal depression at 6-10 weeks postpartum (Figure 1).
No significant association was found with age, marital status and level of education.
The risk of having postpartum depression was increased in women found to have antepartum depression, women reporting economic stress and those who reported experiencing both physical and verbal conflict in their relationships.
|Yes||No||OR (95% CI)||p|
|<21||5 (15.6)||16 (11.5)||1|
|21-25 years||11 (34.4)||54 (38.8)||0.65 (0.20-2.15)||0.483|
|26-30 years||13 (40.6)||47 (33.8)||0.89 (0.27-2.87)||0.839|
|31-35 years||1 (3.1)||17 (12.2)||0.19 (0.02-1.79)||0.146|
|36+ years||2 (6.3)||5 (3.6)||1.28 (0.19-8.76)||0.801|
|Single||3 (9.4)||12 (8.6)||1|
|Married||29 (90.6)||121 (87.1)||0.96 (0.25-3.62)||0.95|
|No formal education||0 (0.0)||1 (0.7)||1|
|Primary||10 (31.3)||35 (25.2)||2.34 (0.73-7.51)||0.152|
|Secondary||17 (53.1)||62 (44.6)||2.25 (0.77-6.57)||0.139|
|Tertiary (College/University)||5 (15.6)||41 (29.5)||1.00 (1.00-1.00)|
|Catholic||8 (25.0)||44 (31.7)||1|
|Protestant||17 (53.1)||78 (56.1)||1.20 (0.48-3.00)||0.699|
|Others||7 (21.9)||17 (12.2)||2.26 (0.71-7.21)||0.167|
|KES2500-10 000||18 (56.3)||50 (36.0)||1|
|KES 10 000-30 000||12 (37.5)||66 (47.5)||0.51 (0.22-1.14)||0.102|
|KES 30 000-60 000||2 (6.3)||22 (15.8)||0.25 (0.05-1.18)||0.081|
|>KES 60 000||0 (0.0)||1 (0.7)||na||na|
Table 1: Demographic antenatal risk factors.
Postnatally, women reporting birth complications, persistent wound pain and those who lost their babies had an increased risk of developing postpartum depression. The participants who breastfed their babies had a reduced risk of developing postpartum depression (Tables 2 and 3).
|Yes||No||OR (95% CI)||p|
|No||21 (65.6)||119 (85.6)||1|
|Yes||11 (34.4)||20 (14.4)||3.12 (1.31-7.44)||0.01|
|Relationship with partner’s mother|
|Good||18 (56.3)||87 (62.6)||1|
|Not good but can cope||5 (15.6)||13 (9.4)||1.86 (0.59-5.87)||0.29|
|Bad and cannot cope||2 (6.3)||4 (2.9)||2.42 (0.41-14.21)||0.329|
|7 (21.9)||35 (25.2)||0.97 (0.37-2.52)||0.945|
|No||24 (75.0)||103 (74.1)||1|
|Yes||8 (25.0)||36 (25.9)||0.95 (0.39-2.31)||0.916|
|No||26 (81.3)||110 (79.1)||1|
|Yes||6 (18.8)||29 (20.9)||0.88 (0.33-2.33)||0.789|
|No||15 (46.9)||97 (69.8)||1|
|Yes||17 (53.1)||42 (30.2)||2.62 (1.20-5.73)||0.016|
|Societal stress and Violence|
|No||31 (96.9)||137 (98.6)||1|
|Yes||1 (3.1)||2 (1.4)||2.21 (0.19-25.15)||0.523|
|No conflict||11 (34.4)||100 (71.9)||1|
|Any Physical conflict||2 (6.3)||1 (0.7)||18.18 (1.52-217.09)||0.022|
|Verbal conflict||16 (50.0)||20 (14.4)||7.27 (2.94-17.99)||0|
|Yes||5 (15.6)||9 (6.5)||1|
|Dont know||13 (40.6)||44 (31.7)||0.53 (0.15-1.87)||0.325|
|No||11 (34.4)||68 (48.9)||0.29 (0.08-1.03)||0.056|
Table 2: Psychosocial antenatal risk factors for postpartum depression.
|Yes||No||OR (95% CI)||P|
|Mode of delivery|
|Normal Vaginal delivery||26 (81.3)||111 (79.9)||1|
|Assisted Vaginal delivery||1 (3.1)||0 (0.0)||na||na|
|Caesarean Section||5 (15.6)||27 (19.4)||0.79 (0.28-2.25)||0.66|
|Reported Birth complications|
|Yes||11 (34.4)||22 (15.8)||1|
|No||21 (65.6)||116 (83.5)||0.36 (0.15-0.86)||0.021|
|Persistent wound pain|
|Present||0 (0.0)||9 (6.5)||1|
|Absent||20 (62.5)||109 (78.4)||0.31 (0.13-0.72)||0.007|
|N/A||12 (37.5)||20 (14.4)||na||na|
|Low birth weight baby|
|No||28 (87.5)||124 (89.2)||1|
|Yes||4 (12.5)||13 (9.4)||1.36 (0.41-4.49)||0.611|
|No||27 (84.4)||126 (90.6)||1|
|Yes||5 (15.6)||12 (8.6)||1.94 (0.63-5.98)||0.246|
|Alive||27 (84.4)||132 (95.0)||1|
|Baby died||5 (15.6)||6 (4.3)||4.07 (1.16-14.32)||0.029|
|History of baby being unwell|
|No||5 (15.6)||24 (17.3)||1|
|Yes||27 (84.4)||114 (82.0)||1.14 (0.40-3.25)||0.811|
|Male||12 (37.5)||67 (48.2)||1|
|Female||20 (62.5)||71 (51.1)||1.57 (0.71-3.46)||0.261|
|Mother happy with birth|
|No||2 (6.3)||3 (2.2)||1|
|Yes||28 (87.5)||130 (93.5)||0.32 (0.05-2.02)||0.228|
|Partner happy with birth|
|No||2 (6.3)||5 (3.6)||1|
|Yes||26 (81.3)||118 (84.9)||0.55 (0.10-3.00)||0.49|
|No||4 (12.5)||5 (3.6)||1|
|Yes||26 (81.3)||133 (95.7)||0.24 (0.06-0.97)||0.045|
Table 3: Postnatal risk factors for postpartum depression.
The study findings show that in this setting both antenatal and postpartum depression rates are comparable. At a cut off of 13 or more on the EPDS antepartum depression rates of 18% was found while for postpartum depression at 6 weeks postpartum and at a cut off of 10,a rate of 21% was found. This is in keeping with studies that have reported similar rates of depression in antepartum, postpartum and non-pregnant women [35-37].
Antepartum depression rate of 18% is higher than estimates from developed countries (13%) , but falls within the range of rates seen in Sub Saharan Africa (8.3-39%) 6. Higher rates in Sub Saharan Africa have been attributed to higher rates of poverty, intimate partner violence and HIV infection [2,35,39-41].This study recruited women from an urban poor resource setting. Both urban living and poverty have been shown to contribute to the risk of developing perinatal depression. Majority of women from urban settings live in nuclear families hence may lack strong interpersonal support networks seen in their rural counterparts . A low socioeconomic status has been shown to increase the risk of depression through increased exposure to traumatic events and stressors [43-45].
Significant antepartum risk factors for postpartum depression were a diagnosis of antepartum depression during baseline assessment, reported economic stress and the presence of conflict with partner. The risk of having postpartum depression symptoms was increased three fold in the presence of antepartum depression. Physical and verbal conflicts increased the risk of postpartum depression 18 fold and 7 fold respectively.
Findings from a recent meta-analysis report conducted on over 14,000 subjects found depression during pregnancy as one of the strongest predictor of postpartum depression . Similarly, a follow up study looking at depression in pregnancy and in the postpartum period found significant association with both antepartum depression and history of exposure to violence and conflict [25,47,48]. Gender based violence both physical and verbal is particularly destructive during pregnancy as this is the period women are most vulnerable and dependent. Other significant associations reported in previous similar studies include stressful life events, unwanted pregnancy and having a non-supportive husband [6,46] however, these were not statistically significant in our study.
Significant correlates of postpartum depression during the postnatal period were birth complications, persistent wound pain (caesarean section wound or episiotomy wound), loss of baby and breastfeeding. Among the women who reported birth complications majority (90%) specified this complication to be postpartum haemorrhage, others reported eclampsia and vaginal tears. Obstetric and infant related factors such as intrapartum haemorrhage, prolonged labour and bad outcome of the baby have been shown to confer a small risk to developing postpartum depression [24,49-51] however, factors related to mode of delivery do not seem to carry a similar risk [52,53]. Breastfeeding was found to be protective to getting postpartum depression (OR: 0.24; 95% CI: 0.06 to 0.97). Equivocal evidence has been documented when it comes to breastfeeding as a risk factor to developing postpartum depression. Some studies suggest a bidirectional relationship between breastfeeding and depression with prenatal depression predicting less breastfeeding and breastfeeding predicting low depression [54,55] conversely, other studies have demonstrated that not breastfeeding does not increase the risk of postpartum depression. Rates of breastfeeding or attitudes towards breast feeding differ within cultures; this may impact a mother’s preference for breastfeeding. In our setting breastfeeding tends to be the norm and hence raises the question that perhaps our nonbreastfeeding mothers had a medical condition not allowing them to breastfeed and hence confounding the postpartum depression finding.
A limitation in our study was the small and non-probability convenient sampling. The size and convenience of the sample may limit the generalizability of the study.
As this was part of larger study to confirm feasibility of integrating EPDS depression screening in maternal and child health clinics, we did not include a concurrent use of a diagnostic tool to confirm depression diagnosis.
To our knowledge this is the only follow up study published in Kenya on antepartum depression with a specific focus on antepartum risk factors as predictors of postpartum depression in an urban poor resource setting. The study findings confirm antepartum depression, financial difficulties and conflict with partner as strong predictors of postpartum depression. Hence, a greater focus in implementing preventive and curative measures that address depression during pregnancy is warranted. Nurses in perinatal settings have an opportunity to screen and counsel high risk pregnant women and mothers and refer the ones in need of specialist care.
This research was supported by a grant from the Kenya Medical Research Institute. The authors are grateful to Francis Creed, for his help in editing the manuscript and the mothers who participated in the study.
Make the best use of Scientific Research and information from our 700 + peer reviewed, Open Access Journals