Results of this research present evidence that gender affected the effect of MVPA on FM gain, but not BMI or FFM gain. A high baseline MVPA predicted a lower FM gain in girls. Findings also support that even though gender did not influence the effect of MVPA on gains in BMI or FFM, boys increasing MVPA demonstrated higher gains in these variables.
The size of the effect of MVPA on FM gain was biologically relevant: for each additional 10 min spent in MVPA at baseline FM gain was reduced by −0.12 kg at endline. So that theoretically, if the recommendation of 60 min/d of MVPA was met [12
], gains in FM could be reduced by −0.72 kg. This number is very close to the figures obtained in this analysis for the most active girls, who gained −0.92 kg less FM at endline. Other studies reported a similar association. In a cohort of American children 10 minutes of MVPA at age 5 y resulted in 0.2 kg less FM at age 8 and 11 y [41
]. In a different American cohort, children with low preschool physical activity levels gained substantially more FM during follow-up than their more active counterparts [42
In the present study, the lagged models do not evaluate any baseline/endline change in MVPA, therefore, the observed reduction in FM gain could be attributed to either initial MVPA or changes in MVPA level during follow-up. The same type of association was confirmed in a dynamic regression model but it was significant only for girls. The effects of persistently high (β
−1.06 kg), increasing (β
−1.23 kg) or decreasing (β
−1.36 kg) MVPA category (dynamic model) were similar to the effect seen for baseline high MVPA (β
−0.92 kg, lagged model) on FM gain. These results suggest that having a high MVPA at any time has a protective role against FM gain. Several studies have evaluated the associations between changes in physical activity with changes in adiposity. In an 8-month follow-up of Chinese school-aged children, girls with the highest MVPA had less increases in body fatness (−0.5%) [13
]. Another longitudinal study in French children found no associations between baseline physical activity and changes in adiposity; however girls decreasing their physical activity level demonstrated higher adiposity gains. [43
] An 8 y follow-up study in children reported that higher accumulated physical activity was associated with less body fat at later age [44
]; however predictions of initial physical activity level on adiposity changes were not investigated.
The gender difference in the effect of MVPA on FM gain was also found in previous studies, but results were more consistent during pubertal age [43
] than in childhood [41
]. The endline age of children in this study varied from 8–9 y, at which an undetermined proportion of girls may have already started their pubertal spurt, making them more susceptible to FM accumulation and therefore to a stronger protective role of MVPA. Though, sexual development was not evaluated. The difference in the effect of MVPA could be also explained by a reported threshold of 115–120 min/d of MVPA to detect impact of physical activity on FM [10
]. At endline, boys in the low/medium MVPA category of the study barely reached this threshold (MVPA
113.0 min/d, 95%CI
104.4, 121.5), and girls were far below it (MVPA
75 min/d, 95%CI
66.0, 77.5, data not shown). Finally, other studies have explained similar gender differences by a stronger effect of dietary energy intake on FM, than the influence of physical activity; [48
] in this analysis, regressions were controlled for dietary intake which did not show any positive association.
In our sample, boys increasing MVPA category gained more BMI and FFM compared to their peers with a persistently low MVPA. These findings are consistent with the results of a physical activity intervention in 6–8 y boys which found an increase of total body and regional lean mass in exercised boys [49
]. This unexpected association suggests that the effect of MVPA on body components depends upon gender differences in baseline body composition. In our sample, boys had a significantly larger FFM and a smaller FM than girls. Muscle remodelling occurs rapidly in response to physical activity and protein intake [50
], thus, it can be speculated that the larger increases in BMI and FFM seen in boys in higher MVPA categories are due to enhancement of muscle mass following remodelling induced by a more intense physical activity.
The main strengths of this study include its longitudinal design, providing robustness to causality implications. The study is based on robust and objective measurement of adiposity and physical activity and the statistical models were adjusted for known potentially confounding variables. The MVPA categories were stratified by age and gender to account for gender differences and age-related change in activity over time [31
]. Finally we demonstrated that losses-to-follow-up did not bias the results. However, some limitations are recognized. Residual confounding may remain. Accelerometers are unable to measure water sports such as swimming and underestimate weight bearing activities such as cycling or climbing stairs. There is a possibility of misclassification in the four categories of change in MVPA. A single 5-d accelerometry measurement might not accurately represent variations in MVPA along follow-up. Therefore a non-systematic misclassification could have occurred among the four categories. In that case, the expected differences between categories could be larger. Another limitation is the fact that sexual development was not evaluated. Since girls are more susceptible to FM accumulation [51
], results in girls of the effect of MVPA in FM gain herein presented could be lower.