Wrist-worn accelerometer measured sleep fragmentation and obesity risk: Socioeconomic gradients and mediation by allostatic load.

Abstract

Objectives: Sleep fragmentation is a recognized risk factor for obesity, yet whether socioeconomic status moderates this association and whether chronic stress responses mediate it remain unclear.

Methods: We analyzed 3007 U.S. adults (≥18 years) from NHANES 2013-2014 with valid wrist-worn accelerometer data, representing approximately 132.7 million adults nationwide. Sleep fragmentation was quantified as wake minutes per hour (WASO/h) and awakenings per hour (AWAKE/h) using a Hidden Markov Model. Outcomes were general obesity (Body Mass Index, BMI) and central obesity (Body Roundness Index, BRI). Moderation by a composite socioeconomic status index (poverty income ratio, education, insurance) and mediation by Allostatic Load Index were evaluated using survey-weighted multinomial logistic regression and moderated mediation models.

Results: Higher sleep fragmentation was associated with greater odds of obesity (WASO/h: OR, 1.32; 95% confidence interval (CI), 1.15-1.50; AWAKE/h: OR, 1.52; 95% CI, 1.27-1.83; both P < .001), with similar patterns for Body Roundness Index-defined obesity. These associations were stronger among higher-socioeconomic status adults and less pronounced in middle-socioeconomic status groups. Moderated mediation analyses indicated partial mediation by Allostatic Load Index, with significant indirect effects in high-socioeconomic status groups accounting for 13%-16% of the total association. Results were consistent across sensitivity analyses including multiple imputation, alternative Allostatic Load Index definitions, and additional adjustment for sleep duration and physical activity.

Conclusions: Sleep fragmentation was significantly associated with increased risk of both general and central obesity, with socioeconomic heterogeneity in effect magnitude. Chronic stress pathways partially explained this relationship, particularly among higher-socioeconomic status adults, suggesting that incorporating sleep quality assessment may benefit obesity prevention across diverse socioeconomic status contexts.