[Study on the Reliability and Validity of International Physical Activity Questionnaire (Chinese Vision, IPAQ)]

Objective: To study the reliability and validity of Chinese version of International Physical Activity Questionnaire (IPAQ) and to provide an instrument for physical activity measurement in Chinese-spoken population.

Methods: Test-retest reliability was systemically assessed in 94 participants sampled from college students. Questionnaires were completed twice with a three-day interval. The validity was established in 39 volunteers by Caltrac accelerometer monitoring and 24-hour activity recording for seven consecutive days.

Results: Both long vision (LV) and short vision (SV) had intraclass correlation coefficients above 0.7 for physical activity. The total energy expenditure measured by LV, SV and PA records were 264.5 +/- 260.9, 185.4 +/- 128.9 (compared with activity records, P < 0.05) and 250.5 +/- 141.2 MET-min/d respectively. Energy expenditure of moderate physical activity were 81.7 +/- 165.4, 32.0 +/- 42.5 (compared with activity record, P < 0.05) and 61.3 +/- 72.0 MET-min/d. Caltrac accelerometer was moderately correlated with LV (r = 0.50) and SV (r = 0.63) while SV measured total daily energy expenditure was lower than activity records. When participants were categorized into two groups according to their time spent in physical activity above or below the target level, proportions of agreement of questionnaires and 24-hour activity records were high, including vigorous physical activity above 90% and moderate physical activity above 70%. LV, SV and activity records were measured during sedentary condition at an approximate level.

Conclusions: Both LV and SV of IPAQ appeared to have acceptable reliability and validity, compared to other physical activity instruments that were used in various large epidemiological studies. The total or physical energy expenditures were similar between LV and activity records. For activity levels, the proportion of agreement were similar between activity records and LV or SV. However, SV underestimated the energy expenditure of total and moderate physical activity.