Background: In previous cross-sectional (n=70) and retrospective (n=260) studies of older people (>65 yr) it was shown that correcting dual x-ray absorptiometry (DXA)-derived appendicular lean mass and mid-thigh lean mass for Body Mass Index (BMI) better associates with outcomes such as grip strength (GSt), gait speed (GSp) and timed-up-and-go (TUG) tests. In this study, we investigated whether the same is observed in a larger population of community-dwelling men and women of a wider age range.
Methods: Using whole-body DXA scans at 15 yr. follow up of the Geelong Osteoporosis Study, a longitudinal study of men and women (n=1322), lean mass in five regions of interest (ROIs) was calculated: 2.6 cm mid-thigh, 13 cm mid-thigh, whole thigh, whole calf and forearm. The lean masses in these ROIs as well as the appendicular lean mass (ALM) were corrected for BMI, (as recommended by the American consensus [FNIH]) and height squared (h2, recommended by the European consensus [EWGSOP2]). The correlation among these variables was evaluated using Pearson’s correlation coefficient and their association with available clinical outcomes (TUG, 1-year retrospective falls history and 5-year retrospective fracture history) was compared using linear and logistic regression.
Results: After adjusting for age and sex, muscle indices of all ROIs adjusted for BMI showed significant associations with TUG (p<0.01) and falls (p<0.027; except whole-calf and whole-forearm muscles [p=0.067 and 0.112]). None of the indices corrected for height squared was associated with TUG or falls. Fractures were only associated with ALM/BMI (p=0.016).
Conclusions: Adjusting DXA-derived lean mass for BMI better associates with physical performance and clinical outcomes than adjustment for h2. Considering the adoption of the European definition of sarcopenia by Australia, future Australian consensuses should consider the above findings. This may improve sarcopenia as well as osteosarcopenia diagnostic algorithms.