Bone Abstracts

Periods of growth are thought to be the best time to improve bone health through increased loading due to the high rates of bone modeling and remodeling. Although numerous observational studies find higher aBMD and greater bone size in physically active children than sedentary children, this may be a result of confounding and publication bias. Randomized trials are the gold standard in determining causality. A meta-analysis of 22 pediatric exercise trials was conducted using Metafor in R. The majority of trials measured bone by DXA; few trials provided information on lean mass changes with exercise. There was a greater overall percent increase in femoral neck (FN), hip, and spine (LS) BMC and FN and LS aBMD among prepubertal children randomized to exercise compared to controls, with a mean percent difference ranging from 0.6% for FN aBMD to 3.7% for LS BMC. BMC results were not significant among postpubertal children and too few trials in postpubertal children reported aBMD. Including calcium intake and length of intervention in the statistical analysis reduced heterogeneity among studies in LS aBMD results, while all BMC analyses had significant heterogeneity. Results related to periosteal circumference and cortical thickness were inconsistent, perhaps due to the variety of measurement methods used. There was evidence that other factors may modify the bone response to exercise, including baseline BMI, baseline activity levels and calcium intake. In summary, only prepubertal children randomized to exercise had greater increases in BMC and aBMD than control children, but few studies have been done in postpubertal children. Other factors may modify the bone response to exercise.

Disclosure: The authors declared no competing interests.