Sleep. 2025 Oct 24:zsaf337. doi: 10.1093/sleep/zsaf337. Online ahead of print.
ABSTRACT
STUDY OBJECTIVES: We examined sex differences in within-person bidirectional and between-person associations between dimensions of Fitbit-measured sleep and multiple self-reported emotional health measures among a large sample of U.S. adults across ~1 year.
METHODS: Data were from American Life in Realtime (ALiR), a probability-based cohort representative of U.S. adults. Participants wore Fitbits to measure sleep and answered monthly questionnaires on depression, anxiety, and stress symptoms. Linear mixed models examined bidirectional associations of nighttime Fitbit sleep measures (mean and SD of total sleep time [TST], clock time of sleep onset [CTSO], and sleep maintenance efficiency [SMEff]) across 14-day intervals that immediately preceded or succeeded each monthly report of emotional health and between-person associations. Analyses adjusted for sociodemographic variables and examined sex differences.
RESULTS: The sample (n=733-754) had a mean age of 47.1±14.6, 48.7%F. Generally, within-person findings indicated that shorter and longer TST, poorer SMEff, and greater variability in CTSO and SMEff than one’s average predicted higher depression, anxiety, and stress scores in the subsequent monthly report. In reverse, worse than average emotional health predicted poorer sleep health in the succeeding 14 days with fewer significant associations than models with sleep predicting emotional health. Between-person results consistently showed that on average, adults with later CTSO and more sleep variability had poorer emotional health. Some findings were stronger in females.
CONCLUSIONS: Poor sleep predicts worse emotional health and vice versa in a large representative sample, particularly among females, highlighting the need for tailored intervention strategies that consider sex-specific patterns in sleep-emotional health dynamics.
PMID:41134282 | DOI:10.1093/sleep/zsaf337
Recent Comments