Longitudinal study investigating associations of sarcopenic obesity with bone mineral density and incident falls and fractures among community-dwelling older men: The concord health and ageing in men project (CHAMP).

Objective: To investigate associations of sarcopenic obesity with five-year changes in BMD, and six-year incident falls and fractures, among Australian community-dwelling older men. Material(s) and Method(s): 1529 CHAMP participants with complete baseline data for hip, spine and total body DXA (BMD, appendicular lean mass (ALM) and body fat), hand grip strength (HGS) and gait speed were included. 1367 and 958 had DXA at two and five years, respectively. Participants were contacted every 4 months for 6.0+/-2.2 years to ascertain incident falls and fractures (confirmed by radiographic reports). Sarcopenia was defined by European Working Group on Sarcopenia (EWGSOP) and Foundation for the National Institutes of Health (FNIH) algorithms. Obesity was defined as body fat >=30 %. Generalised estimating equations and logistic regression determined associations of sarcopenia and obesity with changes in BMD, falls and fracture adjusting for age, physical activity, smoking, living alone, comorbidity, vitamin D, albumin, haemoglobin, and psychotropic, bisphosphonate and corticosteroid use. Result(s): After multivariate adjustment, ALM was associated with BMD at all sites (beta=0.08; p=0.004). Body fat percentage was associated with hip BMD only (beta=0.01; p<0.001). During follow-up, 51 % of participants reported at least one fall, 26 % had multiple falls, and 10 % had at least one fracture. The likelihood of at least one (beta=-0.87; p=0.004) and multiple (beta=-1.42; p<0.001) falls was reduced with higher gait speed but multiple falls increased with higher body fat percentage (beta=0.03; p=0.03). Higher hip BMD (beta=-4.07; p<0.001), but not body fat or sarcopenia components, was associated with lower likelihood of incident fracture. According to EWGSOP and FNIH definitions, 106 (7 %) and 63 (4 %) participants were sarcopenic obese, respectively. Compared to non-sarcopenic non-obese, EWGSOP-defined sarcopenic obese men had increased likelihood of at least one (odds ratio: 2.0