Repeatability and Accuracy of CHAMPS as a Measure of Physical Activity in a Community Sample of Older Australian Adults

Background: One- to two-week test-retest reliability and construct validity (against pedometer step counts) of the CHAMPS physical activity questionnaire were evaluated in older Australian adults. Methods: Participants (n = 100, age >65 years) were invited to complete CHAMPS by mail. Spearman correlation coefficients are reported for physical activity constructs time (min/wk) and sessions per week for walking, moderate-, and vigorous-intensity activity and total physical activity. Correct classification of participants as meeting physical activity recommendations was assessed using percent agreement and kappa statistics. Results: Seventy-three participants completed CHAMPS at T1; 54 provided repeat data (T2). Sixty percent of the participants provided complete data. Good to excellent test retest reliability was observed for all the physical activity constructs (r(s) =.70 to .89 for sessions/wk and r(s) = .65 to .75 for min/wk). Agreement between proportions classified as meeting recommendations at T1 and T2 was good (79%; kappa = 0.55). Fair to low validity coefficients were observed between steps and T1 CHAMPS walking and total activity sessions/wk (r(s) -= .57 and r(s) = .52), and min/wk (r(s) = .40 and r(s) = .21). Conclusions: Mailed self-complete CHAMPS data provided reliable and valid estimates of physical activity in older Australian adults. Observed measurement coefficients were comparable to those reported in previous evaluations of CHAMPS. Further work is required to identify strategies to prevent data loss.