Associations Between Circulating Levels of Myostatin and Plasma β-Amyloid 42/40 in a Biracial Cohort of Older Adults

Abstract

Background: Myostatin, a cytokine produced by skeletal muscle, may influence Alzheimer's disease (AD) pathogenesis, but sparse evidence exists in humans. We assessed the association between circulating levels of myostatin at Year 1 and plasma levels of β-amyloid 42/40 at Year 2, a marker of AD pathology, in a biracial cohort of older adults.

Methods: We studied 403 community-dwelling older adults enrolled in the Health, Aging and Body Composition Study from Memphis, Tennessee, and Pittsburgh, PA. Mean age was 73.8 ± 3 years; 54% were female; and 52% were Black. Serum myostatin levels were measured at Year 1, plasma β-amyloid 42/40 levels in Year 2 (higher ratio indicating lower amyloid load). Multivariable linear regression analyses tested the association of serum myostatin with plasma levels of β-amyloid 42/40 adjusted for computed-tomography-derived thigh muscle cross-sectional area, demographics, APOe4 allele, and risk factors for dementia. We tested for 2-way.interactions between myostatin and race or sex; results were stratified by race and sex.

Results: In multivariable models, myostatin was positively associated with plasma levels of β-amyloid 42/40 (standardized regression coefficient: 0.145, p = .004). Results were significant for white men and women (0.279, p = .009, and 0.221, p = .035, respectively) but not for Black men or women; interactions by race and gender were not statistically significant.

Conclusions: Higher serum myostatin was associated with lower amyloid burden, independently of APOe4 alleles, muscle area and other established risk factors for dementia. The role of myostatin in AD pathogenesis and the influence of race should be further investigated.

Keywords: Alzheimer’s; Biomarkers; Cognitive aging; Muscle.

Figure 1.

Results of multivariable linear regression models of myostatin as a determinant of plasma β-amyloid 42/40, in the total cohort and stratified by race and gender. All models are adjusted for age, APOE4, muscle area, education, physical activity, and hypertension. Models in the full cohort are also adjusted for race and gender. *p < .05.