Ten trials were included in the review (n=1,609 participants): five RCTs in children and adolescents (n=1,160), one trial in adults (n=80) and four trials in older adults (n=369). Trial quality was variable. Most trials had a low risk of bias for sequence generation, allocation concealment and selective outcome reporting. Some trials had a high risk of bias for blinding, incomplete outcome data and other biases.
Children and adolescents: Compared with control, there was no significant difference in bone strength with exercise in young girls (mean difference -0.01, 95% CI -0.18 to 0.17, I2=0%; two RCTs). Subgroup analysis revealed no significant differences at distal tibia, tibia midshaft or femoral neck in young girls.
Compared with control, there was a statistically significant difference in bone strength with exercise in young boys (mean difference 0.17, 95% CI 0.02 to 0.32, I2=0%; three RCTs). Subgroup analysis revealed no significant differences at distal tibia, tibia midshaft or femoral neck in young boys, but there was a statistically significant difference with femoral shaft (mean difference 0.52, 95% CI 0.03 to 1.02; one trial).
Adults: Compared with control, there was no significant difference in bone strength with exercise in postmenopausal women (mean difference 0.00, 95% CI -0.15 to 0.15, I2=0%; four RCTs). Subgroup analysis revealed no significant differences at distal tibia, tibia midshaft, proximal tibia, femoral midshaft and femoral neck in postmenopausal women. One trial in premenopausal women also showed no difference.