Association between objectively determined physical activity levels and body fat percentage in 6-8 year-old children from a black South African population: BC-IT study
Background: Physical activity (PA) levels in children are declining, a widespread predicament affecting children from different backgrounds. Decreasing levels of PA coexist with overweight / obesity and stunting in children of sub-Saharan Africa. PA has been shown to decrease adiposity; however, misclassification of PA level and body composition (BC) has led to inconclusive findings. Studies measuring objective PA and direct BC are needed in South African children. Methods: A total of 93 children aged 6–8 years were sampled from a larger study measuring BCusing isotope technique. Birth weight was obtained from Road-to-Health cards. Height and weight, measured according to standard procedures, were used to calculate body mass index (BMI) and BMI z-score. Body fatness was determined using the deuterium dilution method. Accelerometry (ActiGraph: Model GT3X-BT) and a PA questionnaire (PAQ-C) for the older children (aged 8 years only) were used to measure PA levels. Participant characteristics were reported as means and standard deviations (SD), and frequency analysis performed to determine the prevalence of variables. Independent sample t-test and Mann–Whitney U test were performed to report sex differences; level of significance was set at p<0.05. Multiple regression analysis was performed to determine the relationship between birth weight and BC, and between PA and BC. Crude and adjusted models were reported, with adjustments made for age, gender and socio-economic variables. Bland–Altman plots were used to assess the level of agreement between ActiGraph and PAQ-C. Results: Thirteen percent of the children had low birth weight and almost a third of the children had high adiposity, with girls more affected by high adiposity than boys. Approximately 23% of the children did not meet the recommended PA guidelines. Girls spent significantly more time in sedentary behaviour than boys, who reported higher PA levels than girls (p<0.05). In the adjusted regression model, birth weight was associated with fat mass (FM) (β=0.21, p=0.02, 95%CI: 0.001; 0.412) and fat free mass (FFM) (β=0.24, p=0.013; 95%CI: 0.032; 0.441) in childhood. Time spent in vigorous PA was significantly associated with lower FM (β=-0.25, p=0.01) in the age-adjusted model and with lower FM% (β=-0.20, p=0.04) in the gender-adjusted model. BMI z-score was not associated with any PA indicators (p>0.05). There was a poor level of agreement between PA levels by ActiGraph and PAQ-C. PAQ-C was found to underestimate the PA of the sample population. Conclusion: Obesity in childhood has an early onset, with birth weight significantly associated with higher FM and FFM. High levels of PA, especially of high intensity, is associated with reduced adiposity. This is only notable when BC is directly assessed, with proxies for BC such as BMI z-score not showing any significant association. Girls show higher adiposity and report lower levels of PA than boys. Interventions to promote healthy birth weight and increase high intensity PA are needed, especially in girls. Use of PAQ-C for the assessment of PA levels in 8-year-old South African children should be approached with caution; other methods such as ActiGraphs are needed.
- Health Sciences