Changes in spinal bone density, back muscle size, and visceral adipose tissue and their interaction following a multi-component exercise program in older men: secondary analysis of an 18-month randomized controlled trial

In middle-aged and older men, an 18-month multi-component exercise program improved spinal trabecular BMD, paraspinal, and psoas muscle cross-sectional area (CSA) but not visceral adipose tissue (VAT). However, changes in both muscle and VAT CSA were associated with changes in spinal BMD, independent of the exercise intervention. Introduction In older men, we previously reported that a multi-component exercise program improved lumbar spine (LS) trabecular volumetric BMD (Tb.vBMD) compared with no exercise. This study aimed to investigate the following: (1) the effect of the exercise program on paraspinal and psoas (back) muscle CSA and VAT, and 2) if any exercise-related changes in muscle CSA and/or VAT were associated with changes in spinal BMD. Methods Men (n = 180) aged 50-79 years were randomized to an exercise or no-exercise group. Exercise involved high-intensity progressive resistance training (60-85% max) with weight-bearing impact exercise (3 days/week) for 18 months. Quantitative computed tomography was used to assess L1-L3 Tb.vBMD, paraspinal, and psoas muscle CSA and VAT. Results Exercise resulted in a 2.6% ((95% CI, 1.1, 4.1), P < 0.01) net gain in back muscle CSA, but no effect on VAT (-1.6% (95% CI, -7.3, 4.2)) relative to no exercise. Robust regression indicated that percentage changes in Tb.vBMD were positively associated with changes (expressed as z-scores) in back muscle CSA in both the exercise (beta (beta)-coefficient = 1.9, 95% CI 0.5, 3.2, P = 0.007) and no-exercise (beta = 2.6, 95% CI, 1.1, 4.1, P = 0.001) group, and negatively with the changes in VAT (beta = -2.0, 95% CI -3.3, -0.7, P = 0.003) in the exercise only group. There were no group differences in the slopes for the muscle-bone or VAT-bone relationships. Regression analysis (pooled data) revealed that back muscle CSA and VAT were independent predictors of the change in Tb.vBMD, explaining 14% of the variance. Conclusion A multi-component exercise program in middle-aged and older men improved spinal BMD and back muscle size but not visceral fat. However, changes in back muscle size and VAT were associated with the changes in spinal BMD, independent of exercise.