The relation of strength training participation to changes in physical function over 7 years: the most study

Purpose: Several professional rheumatology, osteoarthritis, and health organizations recognize exercise as an effective approach to manage knee osteoarthritis (OA). Strength training exercises are often recommended, given that lower extremity weakness is a modifiable risk factor for the incidence and progression of knee OA. While the proliferation of wearable technology has facilitated our understanding of physical activity (e.g., unstructured energy expenditure above resting levels), less is known about the long-term impact of exercise (structured and repeated energy expenditure above resting levels), and strength training in particular, on physical function. Quantifying longitudinal patterns of physical function associated with strength training participation may facilitate our understanding of the broader impacts of this particular exercise mode and better inform tailored exercise prescription for adults with knee OA. Therefore, the aim of this study was to determine the relation of self-reported strength training, accounting for moderate-vigorous aerobic activity, to trajectories of objectively measured and self-reported physical function over 7 years. Methods: Participants were included from the Multicenter Osteoarthritis (MOST) study. MOST is a NIH-funded, longitudinal cohort study of individuals with or at risk of developing knee OA. Strength training participation status: Participants completed the Physical Activity Scale for the Elderly (PASE) questionnaire at baseline from which level of strength training participation was obtained. Participants were asked “Over the past 7 days, how often did you do any exercise specifically to increase muscle strength and endurance, such as lifting weights or pushups, etc.?”. Strength training participation was then categorized into ‘no participation’ or ‘any participation,’ with ‘any’ participation defined by reporting at least 1-2 days/week of strength exercise. Physical function: 20-meter walk test (seconds), repeated chair-stand test (seconds to complete five repeated chair stands), and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) function subscale (higher scores represent low function) were used to assess physical function at baseline, 30-, 60-, and 84-months. Statistical analysis: Trajectories of physical function were identified using group-based trajectory modeling. Trajectory group sizes of at least 5% of the study population, and Bayesian Information Criterion, were used to select the optimal number of trajectory groups. We calculated posterior probabilities to determine the proportion of individual trajectories that fit within the group trajectories. Logistic regression was used to determine the association between strength training participation status at baseline and trajectories of changes in physical function. Multinomial logistic regression was used when more than two trajectories were identified. Analyses were adjusted for age, sex, BMI, radiographic knee OA, self-reported moderate-vigorous aerobic activity (never versus any), and knee pain severity. Odds ratios with 95% confidence intervals were determined for each analysis. Results: 3,026 MOST participants reported strength training participation status at baseline. The mean age and BMI were 61.9 (8.1) years and 30.3 (5.8) kg/m2, respectively; 58.3% female. The proportion of participants in the “Never” versus “Any” strength training participation categories were 57.7% and 42.3%, respectively. Three stable trajectories, neither improving or declining over 84 months, were identified for self-report physical function (higher WOMAC scores reflect low function): high function [mean (SD) = 3.3 (3.4)], intermediate function [mean (SD) = 14.4 (8.3)], and low function [mean (SD) = 31.7 (9.6)] (Fig. 1). Compared to participants who reported strength training participation, those with did not strength train had 0.9 (0.7, 1.1) and 1.3 (1.0, 1.8) times the odds of being in the intermediate and low function trajectories, respectively, compared to being in the high function trajectory. We identified two trajectories for the objective measures of physical function over 84 months: stable and worsening (Fig. 1). The stable trajectory for gait speed and chair stands included 90% and 86.5% of participants, while the worsening trajectories included 10% and 14.5%, respectively. Compared to participants who reported strength training participation, those who did not strength train had 1.6 (1.1, 2.3), and 1.8 (1.3, 2.4) times the odds of being in the worsening gait speed and chair stands trajectories, respectively. The mean posterior probabilities of classifying participants to each self-report and objective physical function trajectory ranged from 90.4% to 98.9%, indicating excellent model fit. Conclusions: Participants who engage in strength training exercise are more likely to have and maintain higher levels of physical function over 7 years independent of self-reported level of moderate-vigorous aerobic activity.