Predicting postnatal depressive symptoms in a prospective cohort study in Rwanda: the impact of poor maternal social support.

Background: Postnatal depression is a significant public health issue that demands attention, and recent evidence indicates that rates are relatively high in low-income countries such as Rwanda. However, lack of social support is recognized as a potential risk factor for postnatal depressive symptoms. This study sought to explore the influence of poor maternal social support on postnatal depressive symptoms in a sample of women in Rwanda.

Method: A prospective cohort research design was conducted with women recruited from four different health centers in Rwanda's Southern Province. A sample of 396 pregnant women accessing antenatal care services was recruited at the baseline from their late second term or later, then followed up after giving birth. The dropout rate was 21.46%; thus, the data of 311 women were analyzed. The outcome variable was the presence of depressive symptoms (Edinburgh Postnatal Depression Scale (EPDS) (≥12 cut-off score), while predictor variables included maternal social support measured using a modified Maternal Social Support Scale (MSSS), perceived health status, socio-demographic information (marital status, wealth class, age, education, occupation), negative life events, gestational and obstetric information (parity, pregnancy intention, age at birth, children given birth, and mode of delivery). Univariate and multivariate analyses were performed.

Results: From a sample of 311 participants, over a quarter (20.9%) had elevated postnatal depressive symptoms (EPDS ≥ 12 scores). Elevated scores were predicted by poor perceived health status; respondents reporting neither poor nor good (AOR=0.28, CI = 0.11; 0.72, = 0.007) or good health (AOR=0.14, CI = 0.05; 0.37, = 0.001) were less likely to be affected. Poor maternal social support was also linked with postnatal depressive symptoms; poor partner support (AOR=4.22; CI = 1.44; 12.34; = 0.009) was associated with high risk, while good friend support (AOR = 0.47, CI = 0.23; 0.98, = 0.04) was a significant protector. Additionally, violence or negative life events were also independent predictors of postnatal depressive symptoms (AOR: 2.94, CI: 1.37-6.29, = 0.005).

Conclusion: Postnatal depressive symptoms were found to affect one in five Rwandan women. However, good maternal social support can be a strong protector. Early interventions targeting mothers in the postnatal period and strengthened social support networks for women at risk should be developed.