We found that in healthy elderly men, preservation of whole brain volume was associated with larger total neck muscle cross-sectional area. Therefore in an elderly cohort, those that have a smaller muscle bulk have undergone more brain atrophy. This finding supports the common cause hypothesis, by demonstrating that the rate of sarcopenia and ARCD may occur in parallel within individuals, driven by core underlying biological processes. However, we found no significant association between total neck muscle CSA and ventricular volume (a different measure of brain atrophy), or hippocampal or cerebellar volumes.
We unexpectedly found that total neck muscle CSA was significantly negatively associated with estimated prior cognitive ability (NART) after adjustment for ICA and age, but we found no significant association between total neck muscle CSA and current cognitive abilities. This suggests that those with lower prior cognitive ability may have larger muscles in old age. Muscle mass in old age is determined by 2 factors. Firstly, peak muscle bulk obtained in young adulthood, and secondly, rate of muscle atrophy with ageing. Therefore we hypothesise that those with lower cognitive abilities may have undertaken more manual work
[49, 50] and therefore achieved a greater peak muscle bulk and a larger muscle mass in old age. We can find no plausible explanation as to why a lower prior cognitive ability would favour a slower rate of muscle atrophy. We unfortunately do not have sufficiently detailed previous occupational history or socio-economic class data for the participants to be able to test this theory further at this point.
We found only one previous study which investigated the relationship between muscle size and brain size, and this study also found a positive relationship between muscle bulk and whole brain volume. Burns et al. studied elderly people with early Alzheimer’s disease (AD) (n = 70) or normal cognition (n = 70) and found that whole brain volume, normalized for head size, was predictive of lean mass as measured by DEXA (Beta .20, p < .001) in both groups
. White matter volume was the primary driving factor for the relationship (Beta .19, p < .001) while grey matter volume showed no association with lean mass. This indicates that the cause of loss of lean muscle mass in AD may be different to normal ageing as it is primarily grey matter that is lost in AD.
In the above study Burns et al. also investigated the relationship between MMSE and a measure of global cognitive performance (a composite score made up of the results of a battery of tests, including the DSST and verbal fluency) with muscle mass
. They found a significant positive association between both the global cognitive performance score (Beta .12, p = .007) and MMSE (Beta .11, p = .009) and muscle mass, controlling for age and sex but not for prior cognition which we have shown to correlate with both brain and muscle size (Table
3). Our study was able to investigate the relationship between cognitive decline, by adjusting for prior cognition using the NART score, and current cognition, whereas this study only looked at cross-sectional data from current cognition. This may explain why they found an association between current cognition and muscle mass and we did not.
Several large studies have also investigated the links between muscle size and cognition. In a large cross-sectional study of community dwelling women aged 75 or over (n = 7105), Nourhashemi et al. found that low cognitive function was associated with low fat free mass
. However the cognitive test used was the Short Portable Mental Status Questionnaire (SPMSQ), which consists of only 10 questions and is mainly used as a screening test for cognitive impairment.
Conversely, Wirth et al. studied 4095 consecutive geriatric hospital patients and found that fat-free mass was not associated with cognitive dysfunction, measured using MMSE, after adjusting for age, sex and Barthel index
. Also, Auyeung et al. studied 2737 cognitively normal older people and found that appendicular skeletal muscle mass (ASM) was significantly predictive of MMSE 4 years later in men but not women
. However, after adjustment for age, years of education and baseline MMSE score, the relationship in men was not significant either.
Our study has the benefit of including tests of both prior and current cognitive function. This allows us to look at cognitive decline rather than purely at current cognitive ability, and is the only study we could find that specifically tested the relationship between prior cognition and muscle size. Also, the three large studies mentioned above used cognitive tests which are primarily designed to screen for cognitive impairment (ie SPMSQ and MMSE) rather than to detect the subtleties of change in cognition with age
[40–42], for which our cognitive tests were specifically chosen. Burns et al. used more detailed cognitive tests; however the numbers involved in their study are much smaller compared to the other three studies. Our study is the first to measure muscle cross-sectional area and cognition or brain size; the above mentioned studies used either bioimpedence analysis or DEXA as the measure of muscle bulk.
The main limitations of our study are the lack of longitudinal data and the small sample size. In ageing studies longitudinal data are crucial as it is the rate of loss of muscle size or brain size that is of interest rather than measurements at a cross-sectional time point. With brain size we can partially correct for this using intracranial area, but with muscle size we are unsure if someone has lost 10% of their lean body mass in the previous decade or 50%, as clearly the peak muscle bulk obtained will affect the final outcome greatly. The study also contained mainly white males and this will affect the generalisability of our results.