Comparison of subjective and objective measures of sedentary behavior using the Yale Physical Activity Survey and accelerometry in patients with rheumatoid arthritis

Background: Sedentary behavior is associated with increased risk of functional decline and disability. Individuals with rheumatoid arthritis (RA) spend more time sedentary than healthy adults. Self-reported assessment of sedentary behavior has not been well-developed in this patient population. Methods: 172 adults with RA wore an accelerometer for 7 days and completed a modified version of the Yale Physical Activity Survey (YPAS). YPAS-derived sedentary estimates included 1) daily sitting categories (<3, 3 to 6, 6 to 8, >8 hours/day), 2) continuous daily sedentary time calculated by subtracting hours spent sleeping or in physical activity from a 24-hour day, and 3) rank order of YPAS-derived continuous daily sedentary time. Each estimate was compared with objective accelerometer-derived sedentary time using linear regression and Bland-Altman analysis. Results: A significant relationship was observed between accelerometer-derived sedentary time and all 3 estimates. Bland-Altman plot demonstrated systematic bias, however Bland-Altman plot of rank-order demonstrated that the ranked YPAS-derived continuous estimate was an unbiased predictor of ranked accelerometer sedentary time though limits of agreement were wide. Conclusions: This patient-reported approach using the YPAS shows promise to be a useful tool to identify the most sedentary patients. Providing a practical and accurate tool may increase the frequency sedentary behavior is assessed by clinicians.