Background: We investigated whether there is a synergistic association of bone and muscle quality on mortality risk.
Methods: Data from the Osteoporotic Fractures in Men study (MrOS) was used to build Cox proportional hazards models. Predictors included HR-pQCT total volumetric bone mineral density (BMD) (mg/cm3) and bone strength (failure load, Newtons) at distal radius and tibia; DXA BMD at total hip (g/cm2); HR-pQCT muscle volume (mm3) and density (mg/cm3) at diaphyseal tibia (calf); D3-creatine dilution (D3Cr) muscle mass (whole-body, kg); and grip strength (kg) and leg force (newtons/kg). Deaths were ascertained via death certificates. Covariates including age, race, clinical center, alcohol, smoking, comorbidities, limb length, weight, % fat, physical activity and cognitive function were included statistical analysis.
Results: 1,353 men with a mean age of 84.1 ± 4.0 years (77-101 years, 91.7% white) were followed for 6.20 ± 2.24 years. A total of 558 (41.2%) men died. In unadjusted models using continuous predictors, interaction terms were significant in all models including muscle mass or muscle volume with bone variables (p<0.05); interaction terms were not significant for other muscle variables (muscle density, grip strength or leg force). In multivariable-adjusted models, the association of hip BMD on mortality risk was significant when muscle mass was low (<50th percentile; HR= 0.72 [per SD increment in hip BMD]; 95% CI 0.61-0.85, p<0.001) vs. high (>50th percentile; HR= 0.97; 95% CI 0.82-1.14, p=0.680). Likewise, in multivariable-adjusted models, the association of hip BMD on mortality risk was significant when muscle volume was low (<50th percentile; HR= 0.74; 95% CI 0.63-0.87, p<0.001) vs. high (>50th percentile; HR=1.13; 95% CI 0.95-1.35, p=0.167). Interchanging hip BMD for bone strength at distal limbs resulted in similar findings on mortality risk (p<0.001 to 0.056 when muscle mass or muscle volume was low vs. high [p>0.05] in multivariable-adjusted models).
Conclusions: In this prospective cohort study of older men, we found strong evidence for a synergistic association of bone and muscle quality on mortality risk. Findings were consistent across all bone anatomical sites and were not materially altered by interchanging D3Cr muscle mass for HR-pQCT muscle volume.