Agreement and Predictive Validity Using Less-Conservative Foundation for the National Institutes of Health Sarcopenia Project Weakness Cutpoints

Objectives: To derive lean mass cutpoints based on a less-conservative Foundation for the National Institutes of Health (FNIH) Sarcopenia Project Weakness cutpoint for grip strength (Weak_I) and to assess their agreement with European Working Group on Sarcopenia in Older People (EWGSOP) and prediction of incident slow walking and mortality. Design: Longitudinal analysis. Setting: Baltimore Longitudinal Study of Aging. Participants: Individuals aged 65 and older (287 men, 258 women) with 2 to 10 years of follow-up. Measurements: Weakness was determined according to handgrip strength using a hand dynamometer, appendicular lean mass (ALM) using dual-energy X-ray absorptiometry, and walking speed according to 6-m usual pace walk speed. Analyses were performed using classification and regression tree analysis, Cohen's kappa, and Cox models. Results: Cutpoints derived from Weak_I for ALM (ALM_I) were less than 21.4 kg in men and less than 14.1 kg in women and for ALM adjusted for body mass index (ALM/BMI_I) were less than 0.725 in men and less than 0.591 in women. Kappas with EWGSOP were 0.65 for men and 0.75 for women for ALM_I and 0.34 for men and 0.47 for women for ALM/BMI_I. Men with Weak_I + ALM_I were twice as likely to develop slow walking as those not weak with normal ALM_I (Hazard ratio (HR) = 2.44, 95% confidence interval (CI) = 1.02-5.82). Under EWGSOP, men with weakness and low RALM were almost 3 times as likely to develop slow walking as those not weak with normal RALM (HR = 2.91, 95% CI = 1.11-7.62). Neither approach predicted incident slow walking in women. Conclusion: The ALM_I cutpoints agree with EWGSOP and predict slow walking in men. Future studies should explore sex differences in the relationship between body composition and physical function and the effect of change in muscle mass on muscle strength and physical function.