Reproducibility and Validity of the Shanghai Men's Health Study Physical Activity Questionnaire

Abstract

Reproducibility and validity of the physical activity questionnaire (PAQ) used in the Shanghai Men's Health Study (2003–2006, People's Republic of China) was evaluated in a random sample of 196 participants aged 40–74 years. Participants completed a PAQ at baseline and again 1 year later, 12 monthly 7-day physical activity recalls, and four quarterly 1-week physical activity logs. Reproducibility was evaluated by using the two PAQs and validity by comparing the PAQs with 1-year averages of the two criterion measures: 7-day physical activity recall and physical activity log. The PAQ had moderate to high reproducibility for measuring adult exercise participation (κ = 0.60) and energy expenditure (r_s = 0.68), nonexercise activities (correlation coefficients = 0.42–0.68), and total daily energy expenditure (r_s = 0.68, κ_quartiles = 0.47). Correlations between the PAQ and criterion measures of adult exercise were 0.45 (7-day physical activity recall) and 0.51 (physical activity log) for the first PAQ and 0.62 (7-day physical activity recall) and 0.71 (physical activity log) for the second PAQ. Correlations between PAQ nonexercise activities and the physical activity log and 7-day physical activity recall were 0.31–0.86. Correlations for total energy expenditure were high (0.62–0.77). Results indicate that the Shanghai Men's Health Study PAQ has reasonable reproducibility and validity for classifying men by their level of exercise and nonexercise activities in this cohort.