Relationship Between Accelerometer-Measured Activity and Self-Reported or Performance-Based Function in Older Adults with Severe Aortic Stenosis.

Abstract

In older adults with aortic stenosis, we evaluated whether accelerometer-measured physical activity provides distinct clinical information apart from self-reported surveys or performance-based function tests. We employed wrist-mounted accelerometry in 52 subjects with severe aortic stenosis prior to transcatheter aortic valve replacement (TAVR). Daily daytime activity was estimated using the maximum 10 h of daily accelerometer-measured activity (M10) reported in activity counts. Subjects completed baseline surveys (New York Heart Association (NYHA), Short Form 12 (SF12), Kansas City Cardiomyopathy Questionnaire (KCCQ), EuroQol-5D (EQ-5D), Revised Life Orientation Test (LOT-R), Life Space, Detailed Activity Form) and performance-based function tests (Short Physical Performance Battery, 6-min walk test distance, grip strength) to estimate functional status. Simple and multiple linear regression models were used to evaluate the relationship between accelerometer-measured activity and survey data and performance-based function tests. Among all baseline surveys and performance-based function tests, the only statistically significant univariable relationships identified were weak, negative associations between M10 and SF-12 Mental Composite Score (R2=0.1970, P=0.04) and between M10 and grip strength (R2=0.1568, P=0.004). Neither multiple linear regression of overall survey data (R2=0.6159, P=0.23) nor performance-based function tests (R2=0.1743, P=0.10) correlated with M10. Self-reported surveys and performance-based function tests are not meaningfully correlated with daytime accelerometer-measured activity. The results of our study suggest that accelerometer-measured physical activity provides distinct clinical information apart from self-reported surveys or performance-based function tests.