Clustering of Multilevel Factors Among Children and Adolescents: Associations With Health-Related Physical Fitness.

Abstract

To identify the clustering characteristics of individual-, family-, and school-level factors, and examine their associations with health-related physical fitness. A total of 145,893 Chinese children and adolescents aged 9-18years participated in this cross-sectional study. The 2-step cluster analysis was conducted to identify clusters among individual-, family-, and school-level factors. Physical fitness indicator was calculated through sex- and age-specific z scores of forced vital capacity, standing long jump, sit-and-reach flexibility, body muscle strength, endurance running, and body mass index. Three, 3, and 5 clusters were automatically identified at individual, family, and school levels, respectively. Students with low physical fitness indicator were more likely to be in the "longest sedentary time and skipping breakfast" cluster (odds ratio [OR] = 1.18; 95% confidence interval [CI], 1.12-1.24), and "physical inactivity and insufficient protein consumption" cluster (OR = 1.07; 95% CI, 1.02-1.12) at individual level, the "single children and high parental education level" cluster (OR = 1.15; 95% CI, 1.10-1.21), and "no physical activity support and preference" cluster (OR = 1.30; 95% CI, 1.25-1.36) at family level, and the "physical education occupied" cluster (OR = 1.06; 95% CI, 1.01-1.11), and "insufficient physical education frequency" cluster (OR = 1.16; 95% CI, 1.08-1.24) at school level. Girls were more vulnerable to individual- and school-level clusters, while boys were more susceptible to family clusters; the younger students were more sensitive to school clusters, and the older students were more susceptible to family clusters (P-interaction < .05). This study confirmed different clusters at multilevel factors and proved their associations with health-related physical fitness, thus providing new perspective for developing targeted interventions.