Socio-environmental correlates of physical activity and sedentary behaviors in primary schoolchildren in Hong Kong

The first study (Chapter 3) examined the reliability and validity of the modified Children's Leisure Activities Study Survey (CLASS) questionnaire in assessing physical activity and sedentary behavior of the Hong Kong Chinese children. The modified Chinese questionnaire (CLASS-C) consisted of a checklist of 31 physical activities and 13 sedentary pursuits. The test-retest reliability of the questionnaire was examined in 84 boys and 136 girls aged 9 to 14 years by the comparison of data from the two survey administrations carried out one week apart. Intraclass correlation coefficients (ICCs) were analyzed to determine consistency in results. A sub-sample of 139 children completed the 7-day accelerometer protocol. The validity of the questionnaire was determined by comparing data from the second administration with the accelerometer estimates of moderate- and/or vigorous-intensity physical activity. Bland-Altman plots were performed to further examine the agreement of the two methods. Acceptable reliability was found in self-reported time in moderate-to-vigorous physical activity (MVPA) (ICC = 0.71, 95%CI: 0.61-0.77) and in the sedentary behaviors (ICC = 0.69, 95%CI: 0.59-0.77). Self-reported estimate in daily MVPA was moderately correlated with minutes assessed by the accelerometer (girls: r = 0.48, P<0.05; boys: r = 0.27, P>0.05). Bland-Altman plots showed that self-reports of vigorous physical activity were higher than those assessed by the accelerometer.;The second study (Chapter 4) developed an instrument based on the social ecological model to evaluate the socio-environmental correlates of the children's physical activity and sedentary behavior. The instrument consisted of segments that incorporated reports from both the children and their parents. Children were asked about intrapersonal factors and perceived environments at home and in the neighborhood. Corresponding parents reported demographic information (age, education, occupation, income, and weight status), their physical activity and sedentary behavior habits, the control on their child's sedentary behavior, and perception of physical environment in the neighborhood. The questionnaires were administrated in a group of 160 primary schoolchildren aged 10 to 14 years and their parents on two occasions with 8 to 10 days apart. The test-retest reliability of each individual item was determined using percent agreement and kappa statistics for the categorical variables and intraclass correlation coefficients (ICCs) for the continuous variables. The overall missing rates were 2.5% to 5.6% for children- and parent-reported items. The ICC values for continuous variables in children-reported individual and familial factors ranged from 0.66 to 0.79. For self-reported categorical variables, kappa values ranged from 0.27 to 0.76, and an percent agreement statistics were greater than 80%. Meanwhile, reliability statistics for parent-reported home and environment variables showed acceptable consistency with most of the ICC values greater than 0.70.;The third study (Chapter 5) aimed to investigate the physical activity and sedentary behavior patterns among a sample of 1013 primary 4-6 schoolchildren using the CLASS-C questionnaire. The individual, social and physical environmental correlates of the activity behaviors were investigated through a sub-sample of 303 children, whose habitual activity was objectively assessed by ActiGraph accelerometer, in addition to subjective measures by the CLASS-C questionnaire. The multiple correlates of activity behavior for the children in the sub-sample group were assessed using the environmental correlates questionnaire which was developed in Study II. Cluster analyses were used to group children with similar activity patterns. Hierarchical regression analyses were performed to assess the associations of the multiple variables with activity behaviors. Approximately half of the children did not reach the recommended physical activity level and one fifth of them engaged in excessive TV viewing of 2 hr every day. Cluster analysis identifies 5 clusters of children who shared similar patterns of physical activity and sedentary behaviors. Hierarchical regression analyses showed that participation in school sports team and having higher self-efficacy was positively associated with self-reports in MVPA in boys after controlling age, body mass index (BMI), and parental education. Girls who reported participation in school sports team and perceived the availability of more sports facilities at home self-reported more time spent in MVPA after controlling socio-demographic variables. Perception of parental support on physical activity was positively correlated with accelerometer-assessed MVPA among girls, while it was inversely associated with sedentary behaviors in boys. Residence density was positively associated with boys' sedentary behaviors, while girls whose parents reported more safety concerns on neighborhood environment were more likely to be sedentary. (Abstract shortened by UMI.).