Ecological Momentary Assessment of Variables Associated with Postpartum Depression in Women in Hawai‘i

Abstract

This study used a multivariate time-series design to examine the associations among daily stressors, social support, and mood symptoms in postpartum women. Six women who gave birth in the past year and were 1) Native Hawaiian and/or 2) living in a primarily Native Hawaiian community were enrolled in this study. Ecological momentary assessment (EMA) was used to measure these variables via handheld computer (i.e., Palm Pilot). On average, participants completed three weeks (M=23.8 days, SD=6.1 days) and about 60 daily assessments (M=59.2, SD=14.5) during this study. Momentary positive affect and negative affect were negatively correlated for all six participants and this relationship was significant for five participants. Weekly retrospective assessments were compared to average weekly EMA assessments using z-score transformed scores. Results of percent agreement found participant ratings using retrospective measures of positive affect and social support, and to a lesser degree negative affect and positive mood, tended to be higher than momentary measures. In concurrent time series models, stress was a significant predictor of positive affect for four out of five participants and of negative affect for three out of five participants. Similarly, in time-lagged analyses, stress was a significant predictor of positive affect for four out of five participants and of negative affect for two out of five participants. Social support was an inconsistent predictor of positive and negative affect. This is the first known study to examine the relationship between mood, stress and social support in postpartum women using EMA methodology in a sample of Native Hawaiian women/women living in a Native Hawaiian community. Several methodological limitations of this study were noted.