BMC Public Health (Apr 2017)

Age, period and cohort effects and the predictors of physical activity and sedentary behaviour among Chinese children, from 2004 to 2011

  • Xinping Wei,
  • Yu Zang,
  • Xiaodong Jia,
  • Xiangui He,
  • Shurong Zou,
  • Hui Wang,
  • Meihua Shen,
  • Jiajie Zang

DOI
https://doi.org/10.1186/s12889-017-4215-x
Journal volume & issue
Vol. 17, no. 1
pp. 1 – 11

Abstract

Read online

Abstract Background Very few studies have explored the effects of age, time period, and cohort in association with biological, behavioral, economic, and environmental factors predictors on physical activity (PA) and sedentary behaviour (SB) among Chinese children. Methods We used data from a cohort study of the China Health and Nutrition Survey (CHNS) between 2004 and 2011 (2004, 2006, 2009 and 2011). The outcomes of interest were metabolic equivalent of task (MET) hours per week from both active and sedentary activities. Age, gender, individual characteristics, household size, asset ownership, and urbanisation were included as covariates. Age, period and cohort effects analyses for PA and SB of children (6–17 y, n = 3528) was conducted to explicitly assess differences in PA and SB due to age vs. period effects, and implicitly assess differences by cohorts due to the period-specific experiences across individuals of varying ages. Results The mean age of the sample in each time point fluctuated from 12.6 to 11.3 years and PA slightly decreased from 50.0 ± 63.2 MET hours per week (MET-hr./wk) in 2004 to 47.1 ± 54.9 MET-hr./wk. in 2011. However, SB increased from 31.8 ± 22.0 MET-hr./wk. to 37.6 ± 22.2 MET-hr./wk. Girls had lower PA and higher SB levels than boys. Controlling for age effects, marginal period effects on PA were observed in some survey years. Higher levels of urbanisation and number of household computers served as negative and positive predictors for PA and SB, respectively. Higher household income was a positive predictor of SB. Surprisingly, bigger household size was the only negative predictor of SB (P < 0.05). Conclusions This longitudinal study followed a large cohort of children over a significant period of their childhood. We observed potential age and secular trends in PA levels. Higher community urbanisation and number of home computers were associated with both PA and SB levels. Larger household size was the only factor that was negatively associated with SB. These findings shed light on health policy and preventative health strategies for China and other countries that are now facing similar public health challenges.

Keywords