The effect of multimorbidity on physical performance in community-dwelling older adults in China: a retrospective, population-based cohort study

Abstract

BackgroundFew studies have been done on the associations of multimorbidity on physical performance in China. We investigated the associations of multimorbidity and its patterns with physical performance in older adults (60–100 years of age). MethodsIn this retrospective, population-based cohort study, we used data from participants recruited in the China Health and Retirement Longitudinal Study in 2011, 2013, and 2015, and excluded those participants with missing data on physical performance (n=1893), multimorbidity (n=2), and weight (n=46). Multimorbidity was assessed via questionnaires, and physical performance, as the primary outcome, was measured by maximum grip strength (kg) and gait speed (m/s). Grip strength was assessed using a hydraulic handgrip dynamometer, and gait speed by asking participants to walk along a 2·5 m walkway at twice their normal walking speed. These measures were done at baseline and at each follow-up assessment. Patterns of multimorbidity were assessed using exploratory factor analysis, and a standardised factor score for each pattern was calculated for each participant. Generalised estimating equations were used to evaluate the associations between multimorbidity, multimorbidity pattern scores, and physical performance after adjusting for potential confounding factors. FindingsWe included data from 10 067 participants whose average age at baseline was 67·3 years (SD 6·7); 5004 (50·4%) were women and 5063 (49·6%) were men. Four multimorbidity patterns were identified at baseline: respiratory, cardio-metabolic, visceral, and vascular-visual. The baseline average grip strength was 29·8 kg and gait speed was 0·8 m/s for the whole cohort; men had a higher grip strength than women (34·3 kg vs 22·5 kg, p<0·0001) but gait speed was similar (0·75 m/s vs 0·79 m/s, p=0·50). An increasing number of comorbidities was significantly associated with poorer grip strength, with the greatest estimate for participants with three or more chronic conditions (β=–1·42, 95% CI −1·87 to −0·98; p<0·0001). Respiratory (β=–0·47, 95% CI −0·69 to −0·25; p<0·0001), visceral (–0·41, −0·64 to −0·18; p=0·0006), and vascular-visual (–0·82, −0·82 to −0·31; p<0·0001) patterns were also associated with reduced grip strength. No significant relationship was identified between multimorbidity, as well as its four patterns, and gait speed. However, participants with two conditions (odds ratio [OR] 1·57, 95%CI 1·32 to 1·88; p<0·0001) and with three or more conditions (1·60, 1·34 to 1·92; p<0·0001) both had an increased risk of poor gait speed (≤0·8 m/s). Respiratory (OR 1·14, 1·07 to 1·22; p=0·0001), visceral (1·11, 1·04 to 1·18; p=0·0027), and vascular-visual (1·16, 1·08 to 1·24; p<0·0001) patterns were also associated with increased risk of poor gait speed. InterpretationThe number of chronic conditions in older people has a significant negative relationship, and multimorbidity patterns are differentially associated, with physical performance. Maximum grip strength and poor gait speed could therefore be useful performance measures to incorporate into the care of older patients with multimorbidities. FundingNational Natural Science Foundation of China (grant 81703304).