The Influence of Dietary Carbohydrate on Perceived Recovery Status Differs at the Group and Individual Level—Evidence of Nonergodicity Among Endurance Athletes

Abstract

Purpose Research findings are typically reported at the group level but applied to individuals. However, an emerging issue in sports science concerns nonergodicity—whereby group-level data cannot be generalized to individuals. The purpose of this study was to determine if the relationship between daily carbohydrate intake and perceived recovery status displays nonergodicity. Methods Fifty-five endurance athletes recorded daily measures of self-selected dietary intake, training, sleep, and subjective wellbeing for 12 weeks. We constructed linear models to measure the influence of daily carbohydrate intake on perceived recovery status while accounting for training load, sleep duration, sleep quality, and muscle soreness. Using linear model coefficients for carbohydrate intake we tested whether the distributions (mean and SD) differed at the group and individual levels (indicating nonergodicity). Additionally, a decision tree was created to explore factors that could provide an indication of an individual athlete’s relationship between carbohydrate intake and perceived recovery status. Results Mean values were not different between group- and individual-level analyses, but SDs at the individual level were ~2.4 times larger than at the group level, indicating nonergodicity. Model coefficients for carbohydrate intake were negative for three participants, positive for four participants, and non-significant for 37 participants. The κ value measuring accuracy of the decision tree was 0.52, indicating moderate prediction accuracy. Conclusion For most individuals, carbohydrate intake did not influence recovery status. However, the influence of dietary carbohydrate intake on daily recovery differs at the group and individual level. Therefore, practical recommendations should be based on individual-level analysis.

Predicting daily recovery during long-term endurance training using machine learning analysis

“…All study protocols and materials were approved by the Auckland University of Technology Ethics Committee (22/7), and all participants provided informed consent prior to starting the study. Data collected from the same athletes relating to the influence of carbohydrate intake on daily recovery have been reported elsewhere (Rothschild et al 2023).…”

“…Details of dietary assessment have been described elsewhere (Rothschild et al 2023). Briefly, participants were instructed to maintain their typical dietary habits and record all caloriecontaining food and drink consumed for the duration of the 12-week study, using the MyFitnessPal application (www.…”