Reply from Ross D. Pollock, Norman R. Lazarus, Janet M. Lord and Stephen D. R. Harridge

In our paper on the physiology of human ageing (Pollock et al. 2015) we have utilised a cohort of highly active healthy individuals in order to examine the relationship between chronological age and physiological function. We did this in order to remove the confounding effects of negative lifestyle factors, principally physical inactivity, which have influenced many previous studies of human ageing. In their letter, Jones et al., unfortunately, seem to have somewhat missed the point of our study. They raise a number of issues in regard to the measurement of muscle mass, maximal oxygen uptake (), the use of controls, and our contention that ageing is likely to be individualistic. We are happy to address these.

The skeletal muscle index (SMI) is a metric derived from a dual-energy X-ray absorptiometry (DXA) scan and routinely used to diagnose sarcopenia, having been recommended in a European consensus statement (Cruz-Jentoft et al. 2010), and hence its use in our study. However, we agree with Jones et al. that magnetic resonance imaging (MRI) provides a superior measure of muscle mass and that there are other ways in which to normalise muscle mass data. We also agree that the SMI has limitations given that this measure identified seven individuals in our study as being sarcopenic, when clearly they were in no way functionally impaired.

With regards to , this is an integrated physiological index which is commonly used as a gold standard for physical fitness, and has also been shown to be a good marker of health and predictor of all-cause mortality. In their letter, Jones et al. cite the study by Loe et al. (2013) which reports a statistically weak association between and levels of physical activity in a large number of subjects aged 20–90 years. This weak correlation, in which participants of all ages are included, is used by them to support the contention that habitual physical activity might not be an important factor in studies of healthy ageing. This seems a strange argument given that this particular correlation analysis does not describe the interaction between physical activity, and age. Indeed, it is well documented that in people of similar ages, is related to levels of physical activity (e.g. Mcdonough et al. 1970; Ogawa et al. 1992; Dyrstad et al. 2015) and/or training volume or intensity (e.g. Tanaka et al. 1986; Midgley et al. 2007). That said, it is not our contention that represents the marker for, or should even predict, the biological ageing process. We did not find any function that would fit the criteria of a robust physiological marker for age (i.e. one that had an r² value greater than 0.8) even in this relatively homogeneous group of healthy individuals. Indeed, if anything it will not be a single marker, but possibly a suite of physiological and molecular markers, that may ultimately have the potential to do this and determine the degree of biological ageing present in any one individual. However, it is our contention that only in exercising individuals will optimal physiology be maintained throughout the lifespan. Moreover, in order to identify elements of intrinsic biological ageing, the confounding effect of physical inactivity must be removed, with lifelong exercisers representing the ideal group for such studies. Unfortunately, we don’t know the type, intensity or duration of exercise/physical activity required to deliver such a model. Indeed, it might be that much lower levels of exercise than those undertaken by our cyclists are adequate for optimal ageing to occur, but until this is shown it makes sense to start with those who are doing more than the likely requirement. In our study, as with others, we are compromised in our understanding of the relationship between function and age by cross-sectional designs, a point we shall return to. However, given the known deleterious effects of inactivity (exemplified by bed rest studies) on function and health, it seems perfectly reasonable to hypothesise that inactivity will interact with an intrinsic ageing process to determine physiological function. Furthermore, as with the adaption to exercise itself (e.g. Bouchard & Rankinen, 2001), there will be genetic factors involved which contribute to the variance in function at all ages and are likely to influence the rate of any change as a result of ageing.

This brings us to the third point raised and the use of a comparator group – we use this term rather than ‘control group’ deliberately. While non-exercisers may be deemed clinically healthy, there is no doubt that many physiological indices are much reduced when compared to an exercising population. Furthermore, the likelihood that young sedentary individuals will develop subclinical deleterious effects on their physiology before full blown disease processes manifest themselves is high. Our overarching hypothesis is that exercisers, such as our cyclists, are the control group for studies of human ageing. This is not because they have a high per se, but because their high levels of physical activity mean there is a high probability they will remain free from the contaminating factors of disuse (and other accompanying lifestyle issues such as poor nutrition) which negatively influence physiological function, compared to those who are sedentary. From an evolutionary perspective, it is clear that the body is designed to be physically active and not sedentary, the latter being a product of our modern day lifestyle (Booth & Lees, 2006). From this standpoint, a sedentary group would be considered as the intervention arm of any study of human ageing and not a control group.

We are likely to have come as far as we can using cross-sectional studies to improve our understanding of the ageing process and we acknowledge the limitations of our own study in this regard. Robust longitudinal studies, where individuals are followed over extended periods of time, are now needed. In these studies the physical activity status of the subjects needs to be clearly defined before appropriate interpretation can be given as to whether changes in function can be attributed to intrinsic physiological ageing or to an interaction between lifestyle and physiological ageing. For such studies exercising individuals who maintain high levels of physical activity throughout the lifespan are the logical controls for the biological (if not societally influenced) ageing process. At this point we will be studying biogerontology without any overtones of geriatrics.

Additional information

Competing interests

None declared.