Scientific Programme
Physiology & Nutrition
IS-PN08 - Variability of individual responses to exercise: from statistical considerations to practical recommendations
Date: 05.07.2024, Time: 08:00 - 09:15, Lecture room: Lomond Auditorium
Description
There is a growing interest in the variability of individual responses to exercise, whereby individuals experience idiosyncratic responses to a given exercise intervention. By better understanding this phenomenon, referred to as trainability, and its putative factors, sport and exercise scientists may be able to tailor interventions which maximise the health- and performance-related benefits derived from exercise. In this symposium, Dr Swinton will first address key statistical and methodological matters affecting the quantification of individual response variability to exercise, including the need to account for uncertainty and providing tools to differentiate between measurement error and within-subject variability, and true response variability. In the second talk, Dr Thomas will discuss the contribution of genetic and environmental factors to the observed response variability to exercise, a key question of the STRUETH study where 42 twin pairs performed 3 months of aerobic and resistance training, interspersed by a 3-month washout period. The results suggest that ‘non-responders’ to an exercise intervention can be rescued by changing the mode of exercise. In the last talk, Dr Muniz-Pumares will discuss factors affecting the response variability in maximum oxygen uptake, with a particular focus on how the method used to prescribe the intensity of exercise may affect metabolic stress during an acute bout of exercise, and thus, ultimately, affect response rates.
Chair
Daniel Muniz Pumares
University of Hertfordshire, Institute of Sport
United Kingdom
ECSS Glasgow 2024: IS-PN08
Speaker A
Paul Swinton
Robert Gordon University, School of Health Sciences
United Kingdom
Read CVECSS Glasgow 2024: IS-PN08
An overview of quantifying trainability in sport and exercise science
It has long been presumed that individuals respond differently to chronic exercise and maximising improvements requires an individualised approach. Whilst most exercise prescription adheres to some aspects of scaling (e.g., % 1 repetition maximum or % of maximum heart rate), individualised approaches call for additional and tailored specification in terms of the exercise content and/or dosing. It is in this context that researchers are increasingly exploring if systematic and meaningful differences occur both within and between individuals (e.g., trainability), and as a result, whether individualised approaches are indeed required. To address these questions researchers must employ statistical analyses distinct from those typically used to quantify mean differences. Importantly, researchers must account for variability in measured outcomes caused by measurement error and biological processes independent of the intervention. So far, research findings have been contradictory, with many studies reporting no convincing evidence of trainability, and others reporting clear evidence of meaningful differences between individuals that can be linked to genetic background. The statistical approach used to explore trainability is likely to account for some of the divergent findings reported in the literature. In this talk, I will summarise the terminology and important concepts associated with this emerging area of research. The talk will identify the evolution of terms such as response, responder, and trainability; common sources of variability in measurements made throughout an intervention; and different research designs that can be considered. The talk will then discuss some of the methodological limitations with responder analyses including the use of arbitrary criteria, and that these analyses do not capture true individual treatment effects. Following a discussion of statistical approaches to avoid, the talk will then summarise recommended practices. Firstly, a distinction will be made between heterogeneity of a treatment effect and effect modelling, with the former an important first step to assess whether trainability exists and is likely to be meaningful. A focus will be placed on the standard deviation of individual response statistic and its calculation from individual studies and meta-analyses. Assumptions that influence the validity of the statistic will be discussed. The talk will conclude with a discussion of the need for research to progress to assessment of effect modelling and the requirements for reliable calculation. The sample size required for individual studies to quantify effect modelling will be demonstrated based on reasonable assumptions, and an argument made for the need for individual participant data meta-analyses to ultimately address the question of trainability.
Speaker B
Hannah Thomas
Victoria University, School of Exercise and Nutrition Sciences
Australia
Read CVECSS Glasgow 2024: IS-PN08
Influence of genetic and environmental factors on exercise response variability
Physical inactivity is responsible for 6–10% of annual global mortality. Exercise interventions have been shown to improve traditional risk factors for lifestyle-related diseases and premature death. However, it is also suggested that up to 20–30% of individuals may fail to exhibit beneficial physiological responses to exercise interventions that accord with guideline recommendations for health. These individuals are often termed exercise ‘non-responders’ for particular health parameters. Recent literature has explored individual response variability and how we can improve exercise prescription to achieve optimal adaptations for every individual. It is generally acknowledged that distinct modalities of exercise induce different physiological adaptations, for example, aerobic exercise typically modifies cardiovascular capacity and oxygen uptake (V̇O2max), whereas resistance training principally modifies skeletal muscle function and strength. However, few studies have been designed to directly address inter-individual variation in responsiveness to different exercise modalities. This has practical implications for exercise prescription; when an individual is a non-responder for one modality of training, perhaps they may be ‘rescued’ by converting to an alternate mode. Furthermore, response variability in different health variables or multiple health variables to the same exercise intervention is an equally important question. Although an individual may not see improvements in fitness with aerobic training, it is possible that they have improved several other variables that may improve overall health (e.g., blood pressure, weight, and cholesterol). Another important question is whether changes in health variables with exercise training are dictated by heredity, that is, whether individuals are genetically pre-disposed to respond. A classical twin study design, utilizing both monozygotic and dizygotic twin pairs, has advantages over other heritability estimation methods as extensive variance (i.e., difference between twins) and covariance (i.e., common characteristics between twins) information are deduced. From this, we can determine what amount and proportion of the variation is a result of genes and environmental factors. In this talk, I will discuss the findings and implications of the STRUETH study on improving individualised exercise prescription for health. The STRUETH study employed a randomized cross-over design, where each participant completed two forms of training, to investigate whether responses to training are modality-dependent. Furthermore, monozygotic and dizygotic twins were used to understand exercise trainability. This design can be used to analyse concordance/discordance of responses, to determine whether individuals are universal responders or recalcitrant non-responders and the heritability of this response.
Speaker C
Daniel Muniz Pumares
University of Hertfordshire, Institute of Sport
United Kingdom
Read CVECSS Glasgow 2024: IS-PN08
Exercise prescription as a novel approach to increase response rates to V̇O2max
The most effective means to increase cardiorespiratory fitness, which is quantified as the maximum oxygen uptake (V̇O2max), is endurance training. However, response rates to endurance training are heterogenous, and typically 20-30% of those entering an endurance training programme may not realise an increase in V̇O2max. This is important because relatively small increases in V̇O2max (e.g., 3.5 mL·kg-1·min-1) translate to an 8-35% reduction in premature mortality. This talk will discuss strategies to increase response rates to endurance training, which can be attained by reducing response variability, increasing mean response, or both. Several factors affect response rates to V̇O2max, including non-modifiable (e.g., genetics, sex, age), and modifiable factors (e.g. those related to training characteristics). Among modifiable factors, the method of exercise intensity prescription has been identified as an under-investigated and underexploited, yet promising avenue to improve V̇O2max response rates. The intensity of exercise is normally prescribed relative to a maximal physiological capacity during endurance training programmes, such as relative to V̇O2max or maximum heart rate. Such approaches, however, do not account for the relative positioning of boundaries demarcating exercise domains. This is remarkable, because exercise domains represent a range of intensities where similar homeostatic perturbations are elicited. Indeed, more homogeneous responses to a single bout of exercise are elicited by prescribing the intensity of exercise relative to physiological thresholds, such as the gas exchange threshold as the boundary between moderate and heavy domain and, particularly, critical power as the boundary between the heavy and severe domain. Moreover, data from an individual participant data meta-analysis containing 42 studies and 1587 participants suggests that prescribing the intensity of exercise endurance training programmes relative to physiological thresholds results in greater response rates in V̇O2max, compared to that observed when the intensity of exercise is prescribed relative to traditional anchors of intensity. The greater response rates in V̇O2max appear to derive from higher mean response, without reducing response variability. Overall, this talk will present evidence from a range of studies (responses to an acute bout of exercise, chronic adaptations to training, and meta-analysis of individual participant data) demonstrating that prescribing the intensity of exercise relative to physiological thresholds can reduce response variability to an acute bout exercise, and increase the likelihood of an individual realising a meaningful increase in V̇O2max with endurance training.