About one in four women in a study from Borama, Somaliland, met the study’s definition of postpartum depressive symptoms. Among the 442 women included in the analysis, 109 were classified as symptom-positive, or 24.7%, with an uncertainty range of 20.9% to 28.9%.
The same study found a separate gap in maternal care continuity. Although facility delivery was common, only 152 women, or 34.4%, completed all three parts of the study’s care measure: at least four antenatal visits, delivery in a health facility and some maternal postnatal care. The 95% confidence interval for full completion was 30.1% to 38.9%.
A gap between delivery and follow-up
The contrast was clearest in the individual services. Facility delivery was reported by 408 of the 442 women, or 92.3%. Any maternal postnatal contact was reported by 365 women, or 82.6%, but only 126 women, or 28.5%, reported maternal postnatal care after returning home. Four or more antenatal visits were reported by 176 women, or 39.8%.
The researchers combined those three elements into a continuum-of-care score from 0 to 3, awarding one point for each component. A score of 3 counted as full continuity. The measure captures whether the three reported components were present.
The analysis found higher adjusted care scores among women with university or technical education, those who began antenatal care early, and those who had received information about danger-sign referrals. These were associations in the study data, not evidence that any of those factors caused better continuity. Other or informal income had a negative estimate, but the uncertainty range crossed the point at which no association would be expected.
Symptoms often stayed undisclosed
Among the 109 women classified as symptom-positive, 41, or 37.6%, said they had spoken to no one about what they were experiencing. Only seven women, or 6.4%, said a health worker was the first person they spoke to.
After accounting for other measured factors, higher odds of the symptom outcome were observed among divorced or widowed single mothers, women reporting labor complications, women with secondary or university or technical education, and women who rated their health as good, fair or poor rather than excellent. The largest reported association was among women rating their health as fair or poor, whose adjusted odds were 6.66 times those in the excellent-health reference group, with a 95% confidence interval from 2.62 to 16.90.
The pattern for support from a spouse was less decisive. Adjusted symptom probabilities fell from 31.8% among women who rarely reported support to 24.9% among those reporting support sometimes and 22.9% among those reporting it most of the time. But the overall statistical test was not significant, with a p-value of 0.430, so the study does not establish a protective effect or a dose-response relationship.
What the findings can and cannot say
The researchers conducted a community-based cross-sectional study and used interviews to examine psychosocial, structural, clinical and health-system factors linked with the two outcomes. The analysis used logistic regression for the symptom outcome and robust linear regression for the care score, with an ordered-logistic analysis used as a sensitivity check. The stated sample-size requirement was 440, and 442 women were enrolled; the researchers reported 80% power to detect adjusted odds ratios of 2.0 or greater.
The mental-health measure was a direct binary self-report. Women were classified as positive only if they endorsed all three listed symptoms. It was not a clinician-confirmed diagnosis, so the result should be read as the prevalence of the study-defined symptom outcome rather than a count of diagnosed postpartum depression.
Because the data were collected at one point in time, the study cannot establish which factors came first or whether any factor caused symptoms or changed care use. Missing sampling documentation and survey weights also meant the analysis was unweighted and should not be treated as representative of all women in Borama District. The timing of the postnatal-care component could not be verified uniformly.
There are also statistical reasons for caution. The logistic model did not show evidence of poor calibration, but the care-score model showed unequal variance and a significant Ramsey RESET test, meaning its linear form should not carry the inference on its own. These diagnostics add caution to interpretation of the care-score associations.
Paper data and sources
Original title: Psychosocial and structural factors associated with postpartum depressive symptoms and maternal continuum of care in Borama, Somaliland: a cross-sectional study
Authors: Moumin A, Ali Yusuf A, Moumin A et al.
Journal/Repository: Global health action
Status: Peer-reviewed
First online: 2026-10-09
DOI: Not available
Original paper · Full text