Relationship of pase to moderate/vigorous physical activity measured by accelerometry in the osteoarthritis initiative

Abstract

Purpose: Physical activity conveys health benefits for people with osteoarthritis (OA). Epidemiological studies often use self-reported instruments to explore this relationship, but these measures do not effectively capture actual time spent in physical activity in varying intensities. Public health physical activity guidelines are tied to time spent in physical activity of moderate/vigorous intensity. This study utilizes accelerometers to objectively assess time spent in moderate/vigorous physical activity (MVPA) in adults with knee OA.We investigated the correlation of objectivemeasures with less burdensome subjective measures from Physical Activity Scale in the Elderly (PASE) scores. We also evaluated the association of subjective and objective measures to physical function. Methods: Cross-sectional accelerometer data from 969 adults aged 55 and above with radiographic knee OA (Kellgren-Lawrence 2 in one or both knees) participating in the Osteoarthritis Initiative accelerometer monitoring ancillary study was assessed for physical activity and function. Participants’ response to the PASE questionnaire was followed by 7 days of accelerometer monitoring. Accelerometer measures included average daily minutes in MVPA and average daily minutes in MVPA sessions of 10 minutes or more (MVPA bouts). Other subjective measures including WOMAC (Western OntarioMacMaster) function, SF12 (12-Item Short Form Health Survey) physical function, objective measure gait speed averaged from two 20 meter walks, and covariables including demographics (age, gender), and heath factors (BMI and waist circumference) were assessed. Results: Mean PASE score was 149.80 (SD1⁄478.4) and mean accelerometer MVPA was 15.55 minutes/day (SD1⁄417.18). PASE scores were modestly correlated with average daily minutes in MVPA (r1⁄40.31) and MVPA bouts (r1⁄40.21). PASE scores correlated better with accelerometer measures of MVPA and MVPA bouts in participants who were particularly inactive (characterized by older age, higher BMI, and greater waist circumference). There were no significant gender effects. Accelerometer measures of MVPA and MVPA bouts compared to PASE scores had stronger correlation in WOMAC function (r:-0.14, -0.12 versus -0.02), SF12 physical function (r: 0.22, 0.19, versus 0.07), and gait speed (r: 0.35, 0.29, versus 0.16). Conclusions: In this population with radiographic knee OA, modest correlations of PASE with accelerometer MVPA assessments limit the ability to use this self-reported instrument to explore relationships of activity intensity and OA. Accelerometry data are complementary and help to better characterize the effects of physical activity on health outcomes. Accelerometer MVPA and MVPA bouts were more strongly associated with physical function than PASE, although that correlation was modest at best. The choice of PASE versus objectivemeasurement for future epidemiologic studies must take into account the purpose for which physical activity is being measured.