Journal DOI 10.68956/journal.human-performance-futures

Creative Commons source reading · Load, readiness and fatigue

Predicting Next Day Heart Rate Variability Based on Training Load in Cyclists Using Machine Learning

Barsumyan A; Saukkonen A; Soost C; Graw JA; Burchard R

Originally published in Sports (Basel) · 2026-06-30

From the article

Selected source text. JournalOS removes markup and shortens excerpts where indicated. The full article provides the complete context.

Abstract

AbstractIntroduction: Day-to-day fluctuations in heart rate variability (HRV) are widely used to infer autonomic recovery in endurance athletes. However, the extent to which HRV can be forecast one day ahead from readily available external and internal training-load metrics remains unclear. In this study, we evaluated whether machine learning models can predict next-day HRV in competitive cyclists using the two load descriptors most commonly collected in practice: external load quantified as total mechanical work in kilojoules (kJ) and internal load quantified as session rating of perceived exertion (RPE). Methods: Seven male competitive endurance cyclists were monitored daily for sixteen weeks, yielding 590 athlete-days of longitudinal data (seven independent time series). Two machine learning approaches—support vector regression…

3. Results

A total of 590 athlete-days from seven cyclists were available. Descriptive characteristics and RMSE values for all models, scenarios and lag orders are summarized as best-lag performance per athlete in Table 1. Full per-lag results are provided in Supplementary Table S1.

4. Discussion

To our knowledge, this is the first study of competitive cyclists to compare two ML approaches with a classical autoregressive time series benchmark for forecasting next-day heart rate variability (HRV), under two carefully matched predictor scenarios and across multiple lag orders. Three key findings emerge: the modest incremental value of daily load predictors; the comparatively small gap between ML and the traditional benchmark in terms of physiological significance; and substantial between-athlete heterogeneity in predictive accuracy. Our first and most significant finding is that adding kJ and RPE to a model with access to the athlete’s recent HRV history only yields small improvements in forecasting accuracy. This contradicts the implicit assumption—prevalent in applied monitoring—that quantifying external and internal load in…

Editorial reading guide

  • How did the authors define training exposure and performance response?
  • What alternative explanations, confounders or individual differences limit the conclusion?