Abstract
Introduction
Although prior studies suggest that employment status influences sleep duration, findings reported in the literature remain inconsistent. Little is known about how employment status relates to specific components of sleep architecture, particularly among older Black adults. This study aimed to assess whether sleep architecture and total sleep is differed by employment status (employed versus unemployed) among older Black individuals.
Methods
In a seven-day observational sleep study (NIH-funded MOSAIC), 100 older Black volunteers (72.7% females, ages 64.5±5.6 years) provided objective sleep data using a SleepImage ring (n=100) and Fitbit Inspire 2 (n=97). The SleepImage Ring captured total sleep time (TST), rapid eye movement (REM), and non-REM (NREM) duration, while Fitbit captured TST, light NREM (N1/N2), deep NREM (slow-wave sleep [SWS]), and REM duration. A baseline questionnaire captured demographic data, including employment status (yes/no). A linear regression analysis was conducted to examine the relationship between sleep architecture and employment status, controlling for age, sex, smoking history (yes/no), diagnosis of Obstructive Sleep Apnea (yes/no), marital status (single/married), and sleep medications (yes/no).
Results
Analysis of SleepImage ring data showed individuals who were employed had worse sleep parameters, namely duration of REM (F[7,92]=2.41, R2=0.16, p=0.03) and TST (F[7,92]=2.85, R2=0.18, p=0.01), compared with those who were unemployed. Specifically, being employed was linked to lower weekly REM (β=-0.27, p=0.01) and total (β=-0.24, p=0.02) sleep duration, whereas total NREM sleep duration showed no significant associations (β=-0.19, p=0.07). In contrast, analysis based on Fitbit data revealed significant differences in the proportion of SWS duration (F[7,89]=2.32, R2=0.15, p=0.03). More specifically, older employed Black adults were more likely to have higher proportions of SWS, compared with their counterparts (β=0.22, p< 0.05), but we found no significant associations with TST (β=0.01, p=0.90) or the proportions of light NREM sleep (β=-0.15, p=0.16) and REM sleep (β=0.05, p=0.64).
Conclusion
Employment status influenced the amount of TST, REM sleep, and SWS among older Black adults. These findings align with prior evidence linking social and occupational factors to sleep outcomes. Further research is needed to explore all factors that adversely affect sleep amounts in this population.
Support (if any)
The study is supported by R01-AG20210569.