Changes in Cardiorespiratory Fitness Following Exercise Training Prescribed Relative to Traditional Intensity Anchors and Physiological Thresholds: A Systematic Review with Meta-analysis of Individual Participant Data

Background: It is unknown whether there are differences in maximal oxygen uptake ( $V$ O_2max) response when prescribing intensity relative to traditional (TRAD) anchors or to physiological thresholds (THR).

Objectives: The present meta-analysis sought to compare: (a) mean change in $V$ O_2max, (b) proportion of individuals increasing $V$ O_2max beyond a minimum important difference (MID) and (c) response variability in $V$ O_2max between TRAD and THR.

Methods: Electronic databases were searched, yielding data for 1544 individuals from 42 studies. Two datasets were created, comprising studies with a control group ('controlled' studies), and without a control group ('non-controlled' studies). A Bayesian approach with multi-level distributional models was used to separately analyse $V$ O_2max change scores from the two datasets and inferences were made using Bayes factors (BF). The MID was predefined as one metabolic equivalent (MET; 3.5 mL kg^-1 min^-1).

Results: In controlled studies, mean $V$ O_2max change was greater in the THR group compared with TRAD (4.1 versus 1.8 mL kg^-1 min^-1, BF > 100), with 64% of individuals in the THR group experiencing an increase in $V$ O_2max > MID, compared with 16% of individuals taking part in TRAD. Evidence indicated no difference in standard deviation of change between THR and TRAD (1.5 versus 1.7 mL kg^-1 min^-1, BF = 0.55), and greater variation in exercise groups relative to non-exercising controls (1.9 versus 1.3 mL kg^-1 min^-1, BF = 12.4). In non-controlled studies, mean $V$ O_2max change was greater in the THR group versus the TRAD group (4.4 versus 3.4 mL kg^-1 min^-1, BF = 35.1), with no difference in standard deviation of change (3.0 versus 3.2 mL kg^-1 min^-1, BF = 0.41).

Conclusion: Prescribing exercise intensity using THR approaches elicited superior mean changes in $V$ O_2max and increased the likelihood of increasing $V$ O_2max beyond the MID compared with TRAD. Researchers designing future exercise training studies should thus consider the use of THR approaches to prescribe exercise intensity where possible. Analysis comparing interventions with controls suggested the existence of intervention response heterogeneity; however, evidence was not obtained for a difference in response variability between THR and TRAD. Future primary research should be conducted with adequate power to investigate the scope of inter-individual differences in $V$ O_2max trainability, and if meaningful, the causative factors.