Background/Purpose: Sleep disturbance has been recognized as a major public health issue. Evidence has shown that individuals with osteoarthritis (OA) are more likely to have disturbed sleep. However, it is unknown if distinctive patterns of sleep disturbance over time exist among adults with OA. The purpose of this study is to examine trajectories of restless sleep in adults with or at high risk for knee OA and to identify modifiable risk factors associated with these trajectories.

Methods: We analyzed baseline to 8-year data from Osteoarthritis Initiative. Restless sleep was identified annually as self-report of 3 or more nights of restless sleep during the past week. Group-based modeling (PROC TRAJ) was conducted to identify homogeneous clusters of restless sleep trajectories over 8 years.  Baseline descriptive (age, sex, race, education, BMI, Kellgren-Lawrence grade, chronic knee symptoms, prior knee injury) and potentially modifiable (cardiovascular disease, high depressive symptoms, pulmonary disease, bodily pain interference, smoking, WOMAC function, gait speed, PASE physical activity) factors were examined using multiple logistic regression to identify predictors for membership in restless sleep trajectories.

Results: Four distinct restless sleep trajectories (Figure) were identified from the 4290 participants (mean age 61.1 years [SD 9.1], 58% women, mean BMI 28.5 kg/m² [SD 4.8]): good – persistently low (11.4%), worsening – low to high (66.4%), poor – persistently high (9.5%), improving – high to low (12.7%) probabilities of restless sleep.  Compared to the good group, individuals with worsening probabilities for restless sleep were more likely to report at baseline high depressive symptoms (odds ratio [OR]=2.04, 95% confidence interval [CI]=1.37, 3.05), pain in non-knee joints (OR=1.63, 95% CI=1.17, 2.26), smoking (OR=1.61, 95% CI=1.05, 2.47), and pulmonary disease (OR=1.46, 95% CI=1.46, 95% CI=1.06, 2.02), and were less likely to report physical activity scores above the median (OR=0.77, 95% CI=0.61, 0.96), adjusting for descriptive and other modifiable risk factors. Compared to the poor group, those who improved by reducing restless sleep probabilities were less likely to have baseline high depressive symptoms (OR=0.56, 95% CI=0.39, 0.81).

Conclusion: We identified four distinctive restless sleep trajectories among people with or at high risk of knee OA. Baseline factors including less physical activity, high depressive symptoms, pain in other joints, smoking, and pulmonary disease independently predicted a worsening probability trajectory of restless sleep. Future interventions may want to target these factors to address sleep quality.