Moderators of metabolic threshold adaptations: a systematic review and meta-analysis

Optimal training intensity relative to individual metabolic thresholds remains ambiguous despite its fundamental importance in endurance sports. The present pre-registered systematic review and meta-analysis (PROSPERO: CRD420251020702) evaluated the relative influence of potential moderators on adaptations in the aerobic (AeT) and anaerobic (AnT) thresholds. The primary outcome variable was the magnitude of change in physiological threshold values from pre- to post-intervention, quantified through power output (W), speed (km/h), or oxygen uptake metrics (VO_2max). Three-level meta-analytic models with cluster-robust variance estimation quantified Log Response Ratio effect sizes across 13 studies (24 intervention groups, 338 participants). Among the substantive moderators examined (i.e., training intensity, participant status, and intervention length), training intensity relative to the metabolic threshold showed the most consistent association with changes in threshold-related outcomes. For AeT (k = 9), a significant effect of training intensity was observed (p = 0.017), with supra-threshold training showing the largest improvement (+38.7%) compared with at-threshold (+14.7%) and sub-threshold (+18.4%) training. For AnT (k = 15), a significant categorical effect of training intensity was observed (p = 0.008), with post-hoc comparisons revealing greater adaptations for supra-threshold versus sub-threshold training (p = 0.010), alongside a graded descriptive pattern across categories (below: +2.6%, at: ∼+7%, above: +12.2%). However, substantial heterogeneity persisted (AeT: I² ≈ 89%; AnT: I² ≈ 53%), likely reflecting threshold identification bias (e.g., lactate vs. gas analysis) and varying background training distributions. These findings identify training intensity as a primary moderator associated with threshold adaptations within the current literature, with supra-threshold metabolic range consistently outperforming at- or below-threshold prescriptions within the limited supra-threshold ranges represented in the included studies. Therefore, individualized, supra-threshold training with regular reassessment seems to be recommended to optimize individual training responses. However, substantial heterogeneity and methodological variability warrant caution, and further well-controlled studies are needed before definitive training recommendations can be made.