Exercise and bone mineral density in men: A meta-analysis of randomized controlled trials

Objective: Use the meta-analytic approach to examine the effects of ground and/or joint reaction force exercise on femoral neck (FN) and lumbar spine (LS) bone mineral density (BMD) in men. Methods: Randomized controlled exercise trials >= 24 weeks were included. Standardized effect sizes (g) were calculated and pooled using random-effects models, z-score alpha values and 95% confidence intervals (CI). Heterogeneity was examined using Q and I-2. Statistical significance was set at a two-tailed alpha value (p) of <= 0.05 and a trend at >0.05 to <= 0.10. Results: A moderate and statistically significant improvement was found at the FN (3 g's, 187 participants, g=0.583, 95% CI=0.031, 1.135, p=0.04, Q=5.6, p=0.06, I-2=64%) while a small trend was observed at the LS (5 g's, 275 participants, g=0.190, 95% CI=-0.036, 0.416, p=0.10, Q=3.0, p= 0.55, I-2=0%) Results were sensitive to influence analysis as well as collapsing multiple groups from the same studies so that only one g represented each study. Conclusions: There is currently insufficient evidence to recommend ground and/or joint reaction force exercise for improving and/or maintaining FN and LS BMD in men. Additional well-designed randomized controlled trials are needed before any final recommendations can be formulated. (C) 2012 Elsevier Inc. All rights reserved.