Non-restorative Sleep and Functional Limitation in Adults Aged 25–74 Years: A Propensity Score–matched Analysis

Objectives: This study examined the association between non-restorative sleep (NRS) and functional limitation among US adults aged 25-74 years, assessed dose-response patterns, and estimated the population attributable fraction (PAF). Methods: This cross-sectional study used nationally representative Midlife in the United States Refresher 2 data from 2022-2024 (n=1,699). NRS was categorized as none or rare (never/rarely), intermediate (sometimes), or frequent (often/almost always). A secondary exposure quantified cumulative sleep burden as the count of 3 problems-trouble falling asleep, night waking, and feeling unrested-reported as often or almost always. Functional limitation comprised difficulty with basic or instrumental activities of daily living. Propensity score matching balanced 15 covariates. The PAF was estimated using a generalized Miettinen approach. Results: Frequent NRS was associated with higher odds of functional limitation than no or rare NRS (odds ratio [OR], 2.26; 95% confidence interval [CI], 1.61 to 3.18). Each 1-category increase was associated with 45% higher odds (OR, 1.45; 95% CI, 1.24 to 1.70). Odds also increased with cumulative sleep problems: 1 problem (OR, 1.66; 95% CI, 1.14 to 2.41), 2 problems (OR, 2.50; 95% CI, 1.63 to 3.81), and 3 problems (OR, 2.50; 95% CI, 1.51 to 4.11), suggesting a plateau at higher burden levels. Under causal assumptions, the PAF was 23.9% (95% CI, 14.3 to 34.3), corresponding to approximately 6.9 million affected US adults. Conclusions: RS and cumulative sleep problems were associated with functional limitation in graded patterns that plateaued at higher levels. Sleep disturbance may represent a scalable target for intervention when addressed in middle adulthood.