Reliability and validity of the global physical activity questionnaire in Vietnam.

BACKGROUND The Global Physical Activity Questionnaire (GPAQ) was developed as an improvement of the International Physical Activity Questionnaire (IPAQ) for use in cross-cultural settings. This study compared the reliability and validity of GPAQ and IPAQ in Vietnam. METHODS 251 adults were randomly selected from a population-based survey (n = 1978) of noncommunicable disease risk factors. GPAQ and IPAQ were administered on 2 occasions. Participants wore pedometers and logged their physical activity (PA) for 7 consecutive days. RESULTS Test-retest correlations of GPAQ measurements differed for participants (n = 153) with stable work patterns (work PA r = .43, total PA r = .39) and those (n = 98) with unstable work patterns (work PA r = -0.02, total PA r = -0.05). IPAQ measurements did not differ in this way. GPAQ reliability was poorer for transport (GPAQ r = .25, IPAQ r = .60) and for leisure (GPAQ r = .21, IPAQ r = .45) PA. GPAQ estimates of total PA for participants with stable work patterns were moderately correlated with IPAQ total PA (r = .32), steps per day (r = .39), and PA log (r = .31). CONCLUSIONS The modifications made when designing GPAQ improved its reliability for persons with stable work patterns, but at the expense of poorer reliability for persons with more variable PA. GPAQ did not have superior validity to IPAQ.