Osteoarthritis (OA) affects an estimated 32.5 million US adults and up to 50% of those aged over 65; over 80%
of this burden is attributable to knee OA. Pain is consistently identified as the most troublesome symptom for
people living with knee OA, with higher pain intensity associated with reduced muscle strength, activity
limitations, and reduced quality of life. Exercise is central in clinical practice guidelines for knee OA pain
management; however, evidence synthesis has demonstrated variable and, overall, modest effects.
Observational studies have indicated that the pain-relieving effects of exercise may be modulated by sleep
disturbances, including symptoms of insomnia, thus providing one possible explanation for the variable effect
of exercise on OA pain reduction. This information is especially critical given that an estimated 70% of older
adults with knee OA report having problems with sleep. This study is based on the premise that improving
sleep may improve the pain-relieving benefits of exercise for knee OA. Informed by existing evidence and
focus groups conducted with people living with OA and sleep disturbances, we developed and subsequently
feasibility-tested a hybrid exercise and sleep improvement intervention for OA pain management (Move and
Snooze). The intervention combines a personalized exercise program with an automated and digitally
delivered evidence-based course of Cognitive Behavioral Therapy for Insomnia (CBT-I). The primary aim of
this clinical trial is to determine the effectiveness of the eight-week hybrid program on knee OA pain reduction.
Participants will be randomized to receive the hybrid intervention (personalized exercise program + CBT-I) or
the personalized exercise program alone. In addition to determining effectiveness, we will investigate whether
specific characteristics identify individuals who derive particular benefit from the hybrid intervention.
Specifically, we will determine whether insomnia symptoms need to exceed a specific threshold for provision of
the sleep-enhanced exercise program to be warranted and will examine whether the intervention is more
effective for people presenting with indication of centralized pain (as opposed to predominantly peripheral
pain). We will also determine whether pre-specified psychological factors (e.g., symptoms of depression and
anxiety) alter the effectiveness of the intervention. Finally, as we will collect data on hypothesized possible
mediators of the intervention throughout the course of the trial (e.g., pain vigilance and awareness; positive
and negative affect; fatigue intensity), we will apply contemporary causal inference methods to determine
pathways by which the hybrid intervention, and exercise alone, exert their effects on changes in pain. This
research will generate evidence to advance osteoarthritis pain management towards greater efficacy and
personalization, serving our long-term goal of improving pain outcomes for the vast and growing number of
people living with OA by identifying the most appropriate behavior change interventions for specific patient
phenotypes.
