Background/Purpose: Physical activity amongst individuals with knee osteoarthritis (KOA) has been linked to improved pain, functional status, and less disability. It is important to understand possible barriers to physical activity within this population in order to minimize less favorable health outcomes associated with low levels of physical activity. One possible risk factor for lower physical activity is taking a large number of prescription medications, the side effects of which may reduce physical activity. This study evaluated the relationship between the number of prescription medications taken and moderate-vigorous physical activity.

Methods: Adults with or at high risk for KOA from an Osteoarthritis Initiative (OAI) substudy had physical activity assessed by accelerometer monitoring. We used a hierarchical median quantile regression analysis to determine the relationship between the number of prescription medications taken in the 30-days prior to the visit and the median daily minutes of moderate-vigorous (MV) physical activity. Model 1 of the analysis controlled for demographic factors known to affect physical activity (age, gender, race, BMI), model 2 controlled for model 1 variables and disease confounders (Kellgren and Lawrence grade, WOMAC pain), and model 3 additionally controlled for comorbidity confounders (Charlson comorbidity index and high depressive symptoms defined as Center for Epidemiological Studies Depression Scale ³16).

Results: Valid physical activity data and prescription medication data at the 4-year visit were available for 1889 men and women; mean age 65, 55.5% female, 83.9% white, non-Hispanic. A total of 278 (14.7%) participants reported not taking any prescription medications in the 30 days prior to the visit, 1265 (67%) reported 1-5 prescription medications, and 346 (18.3%) reported 6 or more prescription medications. Controlling for demographic factors, we found a decrease of 0.58 minutes in the median daily moderate-vigorous physical activity per additional prescription medication (95% CI -.87, -.29; p<0.001). This decrease persisted when additionally controlling for disease confounders (0.53 minutes/prescription medication, 95% CI -.83, -.23; p=0.001) and comorbidity confounders (0.52 minutes/prescription medication, 95% CI -.85, -.19; p=0.002).

Conclusion: Participants taking more prescription medications had lower levels of moderate-vigorous physical activity independent of OA severity and comorbidities. These results show that polypharmacy might be a risk factor for lower physical activity. Future research should explore the relationship between certain medication classes common to KOA patients (e.g. narcotic analgesics, psychotropics, beta-blockers, statins, antihistamines) and physical activity.