Transition to sarcopenia and determinants of transitions in older adults: a population-based study

Abstract

Background: Diagnostic criteria for sarcopenia from appendicular lean mass (ALM), strength, and performance have been proposed, but little is known regarding the progression of sarcopenia. We examined the time course of sarcopenia and determinants of transitioning toward and away from sarcopenia.

Methods: ALM, gait speed, and grip strength were assessed seven times over 9 years in 2,928 initially well-functioning adults aged 70-79. Low ALM was defined as less than 7.95 kg/m(2) (men) or less than 6.24 kg/m(2) (women), low performance as gait speed less than 1.0 m/s, low strength as grip strength less than 30 kg (men) or less than 20 kg (women). Presarcopenia was defined as low ALM and sarcopenia as low ALM with low performance or low strength. Hidden Markov modeling was used to characterize states of ALM, strength, and performance and model transitions leading to sarcopenia and death. Determinants of transitioning toward and away from sarcopenia were examined with logistic regression.

Results: Initially, 54% of participants had normal ALM, strength, and performance; 21% had presarcopenia; 5% had sarcopenia; and 20% had intermediate characteristics. Of participants with normal ALM, strength, and performance, 1% transitioned to presarcopenia and none transitioned to sarcopenia. The greatest transition to sarcopenia (7%) was in presarcopenic individuals. Low-functioning and sarcopenia states were more likely to lead to death (12% and 13%). Higher body mass index (p < .001) and pain (p = .05) predicted transition toward sarcopenia, whereas moderate activity predicted transition from presarcopenia to more normal states (p = .02).

Conclusions: Pain, physical activity, and body mass index, potentially modifiable factors, are determinants of transitions. Promotion of health approaching old age is important as few individuals transition away from their initial state.

Keywords: Aging; Epidemiology.; Muscle; Physical function; Sarcopenia.

Figure 1.

Bayesian information criteria results for Hidden Markov models with between three and eight states. Lower Bayesian information criteria values indicate better fitting models, in this case the eight latent state model.

Figure 2.

Model of eight states of appendicular lean mass (ALM), muscle performance and strength. Bars represent the probability of observing a participant having gait speed, ALM or grip strength above their respective cut-points thus longer bars indicate higher probability of having a normal measure (ie, in state 1, the most normal state nearly all participants have gait speed >1 m/s, and all participants have ALM and grip strength above cut-points). States 2 and 3 are intermediate states, state 4 is a state characterized by low function, state 5 represents presarcopenia, states 6 and 7 represent sarcopenia, and state 8 is death. Participants can be classified as belonging to any of these eight states and may transition to any of these states during the 9-year follow-up period. ALM = appendicular lean mass.

Figure 3.

State prevalence across the 9-year study period. States of appendicular lean mass, gait speed, and grip strength are ordered from the most normal: state 1 (bottom in dark blue) to least: state 8 (top in brown). Each color represents a state.