Linear and non-linear associations between physical activity, body composition and multimorbidity over 10 years among community-dwelling older adults.

BACKGROUND: This study aims to describe the relationships between physical activity (PA), body composition and multimorbidity over 10 years. METHOD(S): Participants (N=373; 49% women; average age 61.3+/-6.7 years) were followed for 10 years. Multimorbidity was defined by self-report as the presence of two or more of 12 listed chronic conditions. PA (steps-per-day) at baseline was assessed by pedometer, handgrip strength (HGS) by dynamometer and appendicular lean mass (ALM) and total body fat mass by dual x-ray absorptiometry . Relative HGS and ALM were calculated by dividing each body mass index (BMI). Regression cubic splines were used to assess evidence for a non-linear relationship. RESULT(S): After 10 years, 45% participants had multimorbidity. There was a non-linear relationship between PA and multimorbidity - PA was associated with lower multimorbidity risk among individuals who engaged in <10,000 steps-per-day (RR=0.91, 95% CI: 0.85, 0.97, per 1000 steps-per-day), but not among those who participated in >=10,000 steps-per-day (RR=1.04, 95% CI: 0.93, 1.09, per 1000 steps-per-day). Higher BMI (RR=1.05, 95% CI: 1.02, 1.08, per kg/m 2) and fat mass (RR=1.03, 95% CI: 1.01, 1.04, per kg), and lower relative HGS (RR=0.85, 95% CI: 0.77, 0.94, per 0.1 psi/kg/m 2) and ALM (RR=0.93, 95% CI: 0.88, 0.98, per 0.1 kg/kg/m 2) were linearly associated with a higher risk of multimorbidity. Absolute HGS and ALM were not significantly associated with multimorbidity. DISCUSSION: These findings highlight the potential clinical importance of maintaining adequate levels of PA and of reducing adiposity and maintaining muscle function for minimising the risk of multimorbidity in older adults.Copyright © The Author(s) 2021. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.