Reliability and relative validity of three physical activity questionnaires in Taizhou population of China: the Taizhou Longitudinal Study

B Hu,L F Lin,M Q Zhuang,Z Y Yuan,S Y Li,Y J Yang,M Lu,S Z Yu,L Jin,W M Ye,X F Wang
DOI: https://doi.org/10.1016/j.puhe.2015.03.024
IF: 4.984
Public Health
Abstract:Objective: To examine the test-retest reliabilities and relative validities of the Chinese version of short International Physical Activity Questionnaire (IPAQ-S-C), the Global Physical Activity Questionnaire (GPAQ-C), and the Total Energy Expenditure Questionnaire (TEEQ-C) in a population-based prospective study, the Taizhou Longitudinal Study (TZLS). Study design: A longitudinal comparative study. Methods: A total of 205 participants (male: 38.54%) aged 30-70 years completed three questionnaires twice (day one and day nine) and physical activity log (PA-log) over seven consecutive days. The test-retest reliabilities were evaluated using intra-class correlation coefficients (ICCs) and the relative validities were estimated by comparing the data from physical activity questionnaires (PAQs) and PA-log. Results: Good reliabilities were observed between the repeated PAQs. The ICCs ranged from 0.51 to 0.80 for IPAQ-C, 0.67 to 0.85 for GPAQ-C, and 0.74 to 0.94 for TEEQ-C, respectively. Energy expenditure of most PA domains estimated by the three PAQs correlated moderately with the results recorded by PA-log except the walking domain of IPAQ-S-C. The partial correlation coefficients between the PAQs and PA-log ranged from 0.44 to 0.58 for IPAQ-S-C, 0.26 to 0.52 for GPAQ-C, and 0.41 to 0.72 for TEEQ-C, respectively. Bland-Altman plots showed acceptable agreement between the three PAQs and PA-log. Conclusion: The three PAQs, especially TEEQ-C, were relatively reliable and valid for assessment of physical activity and could be used in TZLS.
What problem does this paper attempt to address?