Criterion Validity and Reliability of 2 Brief Physical Activity Questionnaires in Ethnically Diverse Adults

Abstract

Purpose: The study compares moderate- to vigorous-intensity physical activity (MVPA) as evaluated by the Exercise Vital Signs (EVS) and Physical Activity Vital Signs (PAVS) questionnaires to accelerometry, and evaluates the reliability of the questionnaires in ethnically diverse adults. Methods: Ninety-nine participants (mean age 38.1 y; 49.5% women; Hispanics 43.8%; European American 18.8%; African American 14.6%) were included in the analyses. Participants wore an accelerometer at the hip for at least 7 days and completed the EVS and PAVS questionnaires at the beginning (T1) and at the end (T2) of the 7 days. Associations between the questionnaires and accelerometry were examined using Spearman rho. The reliability of the questionnaires was evaluated using intraclass correlation coefficient. Sensitivity and specificity were also calculated. Results: Weak positive correlations were observed between the accelerometer MVPA and the EVS MVPA at T2 (ρ = .263, P = .013), and the PAVS MVPA at T2 (ρ = .327, P = .003). The sensitivity of the EVS and PAVS was 73.2% and 82.6%, respectively. The specificity for each questionnaire was 35.3%. The reliability for the EVS questionnaire (intraclass correlation coefficient = .855; 95% CI, .791–.901; P < .001) was good, while the reliability of the PAVS questionnaire (intraclass correlation coefficient = .652; 95% CI, .511–.758; P < .001) was moderate. Conclusion: Caution should be used when utilizing the EVS and PAVS questionnaires in ethnically diverse adults.