Background Previous research suggests that sarcopenia is associated with lower cognitive functioning. Evidence on the longitudinal relationship between cognition and sarcopenia, according to the revised criteria of the European Working Group on Sarcopenia in Older People (EWGSOP2), is scarce. This study aimed to investigate both cross-sectional and longitudinal associations between sarcopenia and its defining parameters (muscle strength, muscle mass and physical performance) and cognitive performance in middle-aged and older men.

MethodsThis was a secondary analysis of data from the European Male Ageing Study (EMAS), a multicentre cohort study of men aged 40-79 years, recruited from population registers in eight European centres. Cognitive functioning was assessed by using a battery of three neuropsychological tests, measuring fluid intelligence: Rey-Osterrieth Complex Figure (ROCF-Copy and ROCF-Recall), Camden Topographical Recognition Memory (CTRM) and Digit Symbol Substitution Test (DSST). Sarcopenia-defining parameters appendicular lean mass (aLM), gait speed (GS), chair stand test (CST) and handgrip strength (HGS) were measured. Sarcopenia was diagnosed according to the criteria of the EWGSOP2. All measurements were performed at baseline and after a follow-up of 4.3 years. Cross-sectional associations between cognition, sarcopenia-defining parameters and prevalent sarcopenia (EWGSOP2) were analysed. Longitudinally, the predictive value of baseline cognition on decline in sarcopenia-defining parameters, onset of new sarcopenia and vice versa was examined. Linear and logistic regression were used and adjusted for putative confounders.

Results In the whole cohort (n = 3233), ROCF-Copy (beta = 0.016; P < 0.05), ROCF-Recall (beta = 0.010; P < 0.05), CTRM (beta = 0.015; P < 0.05), DSST score (beta = 0.032; P < 0.05) and fluid cognition (beta = 0.036; P < 0.05) were significantly and independently associated with GS at baseline. In the Leuven + Manchester subcohorts (n = 456), ROCF-Copy (beta = 1.008; P < 0.05), ROCF-Recall (beta = 0.908; P < 0.05) and fluid cognition (beta = 1.482; P < 0.05) were associated with HGS. ROCF-Copy (beta = 0.394; P < 0.05), ROCF-Recall (beta = 0.316; P < 0.05), DSST (beta = 0.393; P < 0.05) and fluid cognition (beta = 0.765; P < 0.05) were associated with aLM. The prevalence of sarcopenia in this population was 17.8%. No associations were detected between cognition and prevalent or incident sarcopenia. Longitudinal analysis showed that low ROCF-Copy score at baseline was associated with an increase in CST in men >= 70 years (beta = -0.599; P < 0.05). In addition, a decrease in ROCF-Recall was associated with a decrease in GS, and a decrease in DSST was associated with an increase in CST (beta = 0.155; P < 0.0001, beta = -0.595; P < 0.001, respectively) in persons with the highest change in both cognition and muscle function.

Conclusions Sarcopenia was not associated with cognitive performance in this population, whereas several components of sarcopenia were associated with domain-specific cognitive performance. Longitudinally, baseline and change in subdomains of cognition predicted change in muscle function in specific subgroups.