Reliability And Validity Of Global Physical Activity Questionnaire In Taizhou Population

[Background] With the economical development and improvement of life quality, the physical inactivity or sedentary lifestyle is increasingly becoming a improtant healthy problem. A large body of epidemiologlcal studies indicated that physical inactivity or sedentary lifestyle have been significantly associated with some noncommunicable chronic diseases, such as cancers and cardiovascular diseases. Physical inactivity increases the risk of cardiovascular diseases, diabetes, breast cancer and colon cancer. The world health organization (WHO) repeatedly reported that the physical inactivity or sedentary lifestyle was identified as the one of leading risk factors for global mortality and disability. However, the relationship between the physical activity (PA) exposure and chronic diseases remains elusive. Precisely measuring the levels of physical activities (occupational activity, walking activities), albeit challenging, is important for etiological study of chronic diseases. The methods to access the physical activity level are the Prerequisites of the studies related bewteen physical activity and diseases. A varity of measurements, such as doubly labelled water, indirect calorimetry, motion sensors and questionnaires, were developed in order to access physical activity level. Physical Activity Questionnaires(PAQ) are the most popular way for its convenience and containing plentiful information,especially in a large-seal epidemiological studies. Various factors such as culture background, population, environment and so on,may lead to significant variation in results. Hence researchers should evaluate the reproducibility and validity of PAQ before implementation of this measurement in population-based investigation. Some PAQs have been developed and assessed the reliability and validity in a few countries. Physical activity questionnaires of self-designed and some PAQs from foreign countries have been started to put into use in China, recently. However, there were only few studies to evaluate the reliability and validity of PAQs.[Objective] In this study, our aim is to examin the reproducibility and validity of the Global Physical Activity Questionnaire Chinese version (GPAQ-C) in Taizhou city, which lies in Jiangsu province in Southern China. The results of this study may not only provide a reliable and valid PAQ for the Taizhou Longitudinal Study (TZLS) and some valuable information for the future studies on physical activity in China, but also for comparable information cross countries.[Methods and Materials] Stratified random sampling was put it into use and three hundred local dewellers (age from30to75)distributed in the5different villages and communities were randomly selected from Taizhou city. These face-to-face interviews of GPAQ-C were finished by6trianed interviewers from Oct2010to Jun2011after all subjects provided informed consent. To examine the reliablity, the participants must complete the investigation of GPAQ-C twice during a7-days interval and intraclass correlation coefficients (ICCs) were used to evaluate the reliability. A last7-days diary or log about physical activity was conducted by all participants in the7-days interval, and then the criterion validity of GPAQ-C was eavluated with7-days PAL by Spearman correlation coefficients.[Results] Two hundred and five participants (79men,126women, mean age51.36±10.24yr)were included in the present analyses. The results indicated that good test-retest reliabilities and criterion validity were observed between the first GPAQ-C and the second GPAQ-C。intraclass correlation coefficients (ICC) for each domain ranged from0.67to0.85, espically total physicla activity domain (ICC=0.84) and moderate physical activity domain(ICC=0.85). The ICC was lowest for transportation domain(0.67,95%CI:0.57-0.75) and greatest for moderate activity(except transportation) domain(0.85,95%CI:0.80,0.88). The total physical activity domain of GPAQs-C was strongly associated with that domain of7-days PAL (rs=0.57,p<0.01). Significant correlations between GPAQs and7-days PALwere noted from0.21to0.57. The spearman correlations bewteen the GPAQ-C and the7-days Pal was lowest for transportation domain(rs=0.21, p<0.01) and greatest for combination between moderate activity domain and vigorous activity domain(rs=0.57, p<0.01). Results of reliability of sub-group showed that ICC values for vigorous activity in female group or in elderly group was not significant and ICC values for other sub-category activity in different group were above d.62(p<0.01). the spearman coefficients for vigorous activity in female group between GPAQ-C and7-days PAL was nonsignificantly0.11. likewise, the spearman coefficients for vigorous activity and transportation related activity in unschooled group or in elderly group were not significant. Other spearman coefficirnts for sub-category activity in sub-groups were ranged from0.25to0.69(p<0.01).[conclusion] Our study indicated that the GPAQ-C had acceptable test-retest reliability and criterion validity for various domains of physical activity, except for lower ICC and rs in transportation domain of GPAQ-C than two coefficients of other domain; The nonsignificant correlation coefficients for vigorous activity and transport related activity in unschooled group,femal group and elderly group suggested that GPAQ-C could be used in female pepole,unschooled people or elderly people with caution; The questionnaire could lost the function of domains if GPAQ-C was used in special population that rarely participated in leisure activities; GPAQ-C overestimated the kinds of activity above3Mets and underestimated sedentary time; overall, GPAQ-C was a reliable and valid physical activity questionnaire to measure physical activity level in etiological studies between physical activity and common diseases..