Rest-activity rhythm fragmentation, functional limitation, and mortality in adults with arthritis: A nationally representative cohort study

To characterize rest-activity rhythms in adults with arthritis, examine the role of functional limitation, and evaluate whether rhythm fragmentation is associated with mortality.

We analyzed 7920 adults from the National Health and Nutrition Examination Survey (NHANES) with at least 4 valid days of wrist accelerometry; arthritis status was available for 7904, including 2182 with arthritis. Nonparametric rest-activity rhythm features were evaluated using weighted linear and Cox models with primary-sampling-unit-clustered robust standard errors. Sensitivity analyses examined cohort selection and quadratic self-reported and device-estimated sleep duration.

After demographic and clinical adjustment, arthritis was associated with higher intradaily variability (β = 0.063 SD, 95% CI 0.016-0.110) and lower most-active 10-hour (M10) activity (β = -0.030, 95% CI -0.051 to -0.009; both q = 0.028). Functional-limitation adjustment attenuated the intradaily variability difference (β = 0.005). Each 1-SD higher intradaily variability was associated with all-cause mortality after full adjustment (HR 1.18, 95% CI 1.10-1.26). Estimates were similar across alternative sleep-duration specifications and within participants with arthritis (HR 1.19, 95% CI 1.01-1.39). The estimated indirect effect through intradaily variability was near zero.

Adults with arthritis had more fragmented rest-activity rhythms and lower peak daytime activity, with functional limitation accounting for most of the fragmentation difference. Higher fragmentation was associated with mortality, including within the arthritis subgroup, supporting further evaluation of this wearable-derived sleep-health and behavioral-rhythm phenotype in arthritis research.