Background: Although muscle strength and mass are highly correlated, the relationship between direct measures of low muscle mass (sarcopenia) and strength in association with mortality has not been examined.
Methods: Total mortality rates were examined in the Health, Aging and Body Composition (Health ABC) Study in 2292 participants (aged 70-79 years, 51.6% women, and 38.8% black). Knee extension strength was measured with isokinetic dynamometry, grip strength with isometric dynamometry. Thigh muscle area was measured by computed tomography (CT) scan, and leg and arm lean soft tissue mass were determined by dual energy x-ray absorptiometry (DXA). Both strength and muscle size were assessed as in gender-specific Cox proportional hazards models, with age, race, comorbidities, smoking status, level of physical activity, fat area by CT or fat mass by DXA, height, and markers of inflammation, including interleukin-6, C-reactive protein, and tumor necrosis factor-alpha considered as potential confounders.
Results: There were 286 deaths over an average of 4.9 (standard deviation = 0.9) years of follow-up. Both quadriceps and grip strength were strongly related to mortality. For quadriceps strength (per standard deviation of 38 Nm), the crude hazard ratio for men was 1.51 (95% confidence interval, 1.28-1.79) and 1.65 (95% confidence interval, 1.19-2.30) for women. Muscle size, determined by either CT area or DXA regional lean mass, was not strongly related to mortality. In the models of quadriceps strength and mortality, adjustment for muscle area or regional lean mass only slightly attenuated the associations. Further adjustment for other factors also had minimal effect on the association of quadriceps strength with mortality. Associations of grip strength with mortality were similar.
Conclusion: Low muscle mass did not explain the strong association of strength with mortality, demonstrating that muscle strength as a marker of muscle quality is more important than quantity in estimating mortality risk. Grip strength provided risk estimates similar to those of quadriceps strength.