Comparing Physical Activity Measures in a Diverse Group of Midlife and Older Adults

PURPOSE: To compare self-report and objective measures of moderate- and vigorous-intensity physical activity (MVPA min·d(-1)) in midlife and older adults. METHODS: Seventy-one participants (69% female, 74.6% Caucasian, 25.4% African American) completed the Behavioral Risk Factor Surveillance System physical activity (PA) questions, the Aerobic Center Longitudinal Study PA short survey (PASS), and the Aerobic Center Longitudinal Study PA long survey (PALS) and wore an accelerometer for seven consecutive days. Accelerometer MVPA minutes per day were determined using 1- and 10-min MVPA bout methods. RESULTS: Participants were older (mean ± SD; age = 57.4 ± 9.9 yr) and overweight (body mass index = 27.9 ± 4.9 kg·m(-2)) but otherwise healthy. Median (interquartile range) MVPA minutes per day were 42.9 (51.4) from the Behavioral Risk Factor Surveillance System PA questions, 51.4 (68.6) from the PASS, 25.7 (48.6) from the PALS, 32.4 (33.5) from the 1-min MVPA bout accelerometer data, and 4.6 (16.8) from the 10-min MVPA bout accelerometer data. Pearson correlations adjusted for participant demographics revealed low to moderate correlations between self-report and 1-min MVPA bout accelerometer-determined MVPA minutes per day (r = 0.11-0.31), with the PASS (P < 0.05) and PALS (P < 0.01) having significant correlations with accelerometry. Cohen κ coefficients showed poor agreement between all three questionnaires and 1-min MVPA bout accelerometry for having ≥150 MVPA min·wk(-1) (κ = 0.26-0.38, all P < 0.05). CONCLUSIONS: Our results indicate that there was poor agreement between self-report and accelerometer-based assessments of PA in midlife and older adults.