Postpartum mood among universally screened high and low socioeconomic status patients during COVID-19 social restrictions in New York City

Postpartum depression is common and associated with significant adverse outcomes for mothers, families, and child development. Early in the COVID-19 pandemic, concerns grew regarding its impact on postpartum mental health. Existing literature suggested that women in lower socioeconomic status (SES) are at increased risk for postpartum mood disturbances due to factors including limited childcare, reduced social and partner support, distrust of healthcare, racism, and health literacy barriers. However, prior studies during COVID-19 often relied on self-selected samples and did not examine differential effects across SES, limiting generalizability. This study aimed to (1) determine changes in postpartum mood symptomatology before versus during COVID-19-related social restrictions in New York City and (2) assess whether changes differed between high- and low-SES patients undergoing universal postpartum depression screening as part of routine care. The work seeks to inform maternal health policy by identifying SES-related differences in postpartum mood during pandemic restrictions.

The paper reviews the robust association between SES and health, noting the income gradient and the vulnerability of minority women in low SES to postpartum mood disruptions. It discusses barriers to optimal postpartum care among low SES populations, including suboptimal treatment, distrust of providers, racism, religious and cultural factors, perceived negative caregiver traits, and health literacy. Social determinants such as inadequate childcare, limited partner/social support, and time constraints linked to financial obligations also contribute to poorer outcomes. Early COVID-19 studies reported increased anxiety and depression in maternal populations but were limited by recruitment methods prone to sampling bias (social media ads, unsolicited emails, random phone calls) and did not stratify by SES. The authors situate their study within this gap, emphasizing universal screening and SES comparison to improve external validity and relevance for policy.

Design and setting: Observational analysis of universally screened postpartum patients receiving routine care in the Mount Sinai Health System (New York City) between January 2, 2020 and June 30, 2020. Sites included community ambulatory practices (Mount Sinai Hospital Obstetrics and Gynecology Associates, Mount Sinai West, Mount Sinai Queens) and Mount Sinai Faculty Practices. SES classification: Patients were categorized as low SES if seen at the community practices (serving predominantly minority, Medicaid-enrolled or pending-enrollment populations) and high SES if seen at Faculty Practices (commercial insurance/private pay). Participants: Entire clinical cohort attending in-person or virtual postpartum appointments during the study period (n=516). Community practice demographics: majority Hispanic and African-American (≈90%), age 16–40 (mean 27), parity 1–10 (21% primiparous), cesarean rate 33%. Faculty practice demographics: age 19–48 (mean 33), parity 1–9 (10% primiparous), cesarean rate 44%. Measures: Edinburgh Postnatal Depression Scale (EPDS), a 10-item self-report (0–30). Universal administration in primary language at visit; during early pandemic, EPDS administered by phone by clinical social workers with scoring entered into the EHR. EPDS cutoff guidance: ≥9 suggests possible depression; ≥12 suggests probable depression (sensitivity/specificity ranges cited). Data collection: EPDS scores and dates extracted from the Mount Sinai network and imported into SPSS 27.0. Analytic plan: To assess changes pre- versus during restrictions, patients were split by date relative to March 13, 2020 (ban on large gatherings). Because EPDS is ordinal, Wilcoxon–Mann–Whitney tests were used after confirming equivalent distributions. For SES comparisons, Wilcoxon–Mann–Whitney tests examined differences between low and high SES both before and during restrictions. Within-SES pre/post comparisons were also conducted. Ethical considerations: Protocol approved by Icahn School of Medicine at Mount Sinai IRB (ISMMS IRB #20-03633); procedures followed HIPAA security rules and the Helsinki Declaration. Informed consent: Waived under the Common Rule (45 CFR 46.116).

- Sample: 516 postpartum patients screened with EPDS between 1/2/2020 and 6/30/2020; 418 (71.9%) reported some mood change (EPDS ≥1). Sixty-four (12.4%) scored ≥9 (possible depression) and 33 (6.4%) scored ≥12 (probable depression). Overall EPDS: mean 3.8, median 3.0, range 0–25, SD 4.4. - By SES (overall, 1/2–6/30/2020): Low SES (n=323): mean 3.41, median 2, range 0–24, SD 4.37; EPDS≥1: 250 (77.4%); EPDS≥9: 43 (13.3%); EPDS≥12: 24 (7.4%). High SES (n=193): mean 4.42, median 3, range 0–25, SD 4.33; EPDS≥1: 168 (87.0%); EPDS≥9: 27 (14.0%); EPDS≥12: 11 (5.7%). - Pre vs during restrictions (all subjects): Pre (1/2–3/12/2020; n=264): mean 4.50; EPDS≥1: 217 (82.2%); EPDS≥9: 39 (14.8%); EPDS≥12: 19 (7.2%). During (3/13–6/30/2020; n=252): mean 3.04; EPDS≥1: 201 (79.8%); EPDS≥9: 31 (12.3%); EPDS≥12: 16 (6.3%). Wilcoxon–Mann–Whitney: U=25,084.5, z=-4.90, p<0.001 indicating a significant decrease in symptomatology during restrictions. - SES comparisons: • Before restrictions: No difference between low and high SES (U=7956.0, z=-1.05, p=0.293). • During restrictions: Significant difference between SES groups (U=4895.0, z=-3.48, p<0.001), driven by improvement in the low SES group. - Within-SES pre vs during: • Low SES: Significant decrease in symptomatology during restrictions (U=9209.0, z=-4.56, p<0.001); means: pre 4.44 vs during 2.55. • High SES: No significant change (U=4045.5, z=-1.06, p=0.288); means: pre 4.57 vs during 4.20.

The study found that postpartum mood symptomatology decreased during COVID-19 social restrictions overall, with a differential effect by socioeconomic status. Patients in low SES exhibited a significant reduction in EPDS scores during restrictions, while high SES patients showed no significant change. The authors suggest that pandemic-related policies (stay-at-home orders, closure of non-essential businesses, widespread remote work, school closures) may have temporarily alleviated key stressors that disproportionately affect low SES mothers—such as limited childcare availability, constrained partner/family support, and inflexible work schedules—thereby improving postpartum mood. These findings challenge the exclusive focus on negative impacts of restrictions and highlight potential benefits of policies that ease early maternal burdens. Given the established link between maternal mental health and long-term offspring outcomes, the results underscore the need for social and health policies that support work-family-childcare balance, particularly for postpartum women in low SES. The universal screening approach enhances generalizability compared to prior self-selected samples and reveals SES-specific dynamics in postpartum mood during the pandemic.

Postpartum depression is a prevalent condition with significant morbidity. This study demonstrates that, in New York City, COVID-19 social restrictions were associated with improved postpartum mood among women in low SES, while high SES patients experienced no significant change. These findings suggest that policies reducing early maternal burdens (e.g., enhanced childcare support, flexible work arrangements, parental leave) may improve postpartum mental health in populations at higher risk and potentially interrupt the cycle linking poor maternal mental health and poverty. Future research should investigate mechanisms underlying SES-related differences, examine longer-term outcomes as pandemic policies evolve, and evaluate targeted policy interventions that support postpartum women in low SES.

- Generalizability: The sample was a treatment-seeking clinical cohort from New York City care-based centers and may not represent the broader population. Universal screening still captures only those who attended in-person or virtual visits, which may be influenced by access barriers, especially in low SES. - Limited covariate data: Demographic and obstetrical variables (e.g., age, race/ethnicity, parity, delivery route) were not available for analysis; within-SES income variability (especially in the high SES group) could not be examined. - Potential seasonal/cyclical confounding: Although a simulated historical comparison (2015 dates) showed no such trend, unmeasured temporal factors cannot be fully excluded. - Evolving context: Longer-term socioeconomic consequences of the pandemic are unknown; benefits observed during temporary restrictions may reverse as supports (expanded access to care, housing protections, stipends, unemployment benefits) are lifted. Post-pandemic employment and childcare challenges may disproportionately harm low SES postpartum women.