Kristin M Davis, Margeaux M Schade, Sofia H Carfagno, Katherine Knauft, Orfeu M Buxton, Samuele Zilioli
No mediating effects of stress on the association between SES and sleep duration or sleep quality were found. Results suggest that SES affects within-person variability in sleep duration and sleep quality, but not night to night sleep.
BACKGROUND: Stress purportedly mediates the association between socioeconomic status (SES) and sleep health. However, this mediation pathway remains poorly delineated, especially among African Americans.
METHODS: Data from the Health among Older Adults Living in Detroit study included African American adults aged 50-89 years. Participants wore an actigraphy watch and completed daily diaries, including a modified 4-item Perceived Stress Scale, for five days. Measures of sleep duration (nocturnal total sleep time (TST)) and sleep quality (wake after sleep onset (WASO)) were derived from actigraphy. Participants with ≥3 valid days of actigraphy and stress data were included (n = 150). Regression was used to test the effect of SES on variability in TST and WASO. Multilevel modeling was used to test whether day-level stress mediated the effect of SES on nightly sleep measures.
RESULTS: Lower SES was associated with greater TST variability (B = -8.06, p = 0.03), WASO variability (B = -2.83, p = 0.03), and mean stress (r = -0.32, p < 0.001). No effects of SES or stress on nightly TST or nightly WASO were found. Mediation of SES by day-level stress was not supported for nightly TST (indirect Effect [95% CI] = 1.52 [-0.25, 3.23]) or nightly WASO (indirect Effect [95% CI] = 0.40 [-0.03, 1.00]).
CONCLUSIONS: No mediating effects of stress on the association between SES and sleep duration or sleep quality were found. Results suggest that SES affects within-person variability in sleep duration and sleep quality, but not night to night sleep.