Dataset Information

Training load responses modelling and model generalisation in elite sports.

ABSTRACT: This study aims to provide a transferable methodology in the context of sport performance modelling, with a special focus to the generalisation of models. Data were collected from seven elite Short track speed skaters over a three months training period. In order to account for training load accumulation over sessions, cumulative responses to training were modelled by impulse, serial and bi-exponential responses functions. The variable dose-response (DR) model was compared to elastic net (ENET), principal component regression (PCR) and random forest (RF) models, while using cross-validation within a time-series framework. ENET, PCR and RF models were fitted either individually ([Formula: see text]) or on the whole group of athletes ([Formula: see text]). Root mean square error criterion was used to assess performances of models. ENET and PCR models provided a significant greater generalisation ability than the DR model ([Formula: see text], [Formula: see text], [Formula: see text] and [Formula: see text] for [Formula: see text], [Formula: see text], [Formula: see text] and [Formula: see text], respectively). Only [Formula: see text] and [Formula: see text] were significantly more accurate in prediction than DR ([Formula: see text] and [Formula: see text]). In conclusion, ENET achieved greater generalisation and predictive accuracy performances. Thus, building and evaluating models within a generalisation enhancing procedure is a prerequisite for any predictive modelling.

SUBMITTER: Imbach F

PROVIDER: S-EPMC8799698 | biostudies-literature | 2022 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Training load responses modelling and model generalisation in elite sports.

Imbach Frank F Perrey Stephane S Chailan Romain R Meline Thibaut T Candau Robin R

Scientific reports 20220128 1

This study aims to provide a transferable methodology in the context of sport performance modelling, with a special focus to the generalisation of models. Data were collected from seven elite Short track speed skaters over a three months training period. In order to account for training load accumulation over sessions, cumulative responses to training were modelled by impulse, serial and bi-exponential responses functions. The variable dose-response (DR) model was compared to elastic net (ENET), ...[more]

PMID: 35091649

Similar Datasets

Project description:The hematological module of the Athlete Biological Passport (ABP) is used in elite sport for antidoping purposes. Its aim is to better target athletes for testing and to indirectly detect blood doping. The ABP allows to monitor hematological variations in athletes using selected primary blood biomarkers [hemoglobin concentration (Hb) and reticulocyte percentage (Ret%)] with an adaptive Bayesian model to set individual upper and lower limits. If values fall outside the individual limits, an athlete may be further targeted and ultimately sanctioned. Since (Hb) varies with plasma volume (PV) fluctuations, possibly caused by training load changes, we investigated the putative influence of acute and chronic training load changes on the ABP variables. Monthly blood samples were collected over one year in 10 male elite cyclists (25.6 ± 3.4 years, 181 ± 4 cm, 71.3 ± 4.9 kg, 6.7 ± 0.8 W.kg-1 5-min maximal power output) to calculate individual ABP profiles and monitor hematological variables. Total hemoglobin mass (Hbmass) and PV were additionally measured by carbon monoxide rebreathing. Acute and chronic training loads-respectively 5 and 42 days before sampling-were calculated considering duration and intensity (training stress score, TSSTM). (Hb) averaged 14.2 ± 0.0 (mean ± SD) g.dL-1 (range: 13.3-15.5 g·dl-1) over the study with significant changes over time (P = 0.004). Hbmass was 1030 ± 87 g (range: 842-1116 g) with no significant variations over time (P = 0.118), whereas PV was 4309 ± 350 mL (range: 3,688-4,751 mL) with a time-effect observed over the study time (P = 0.014). Higher acute-but not chronic-training loads were associated with significantly decreased (Hb) (P <0.001). Although individual hematological variations were observed, all ABP variables remained within the individually calculated limits. Our results support that acute training load variations significantly affect (Hb), likely due to short-term PV fluctuations, underlining the importance of considering training load when interpreting individual ABP variations for anti-doping purposes.

Project description:Background:Pediatric sports specialization, defined as intense year-round training in a single sport as a result of excluding other sports for more than 8 months per year, is common in the United States. There are demonstrated physical and social risks to early pediatric sports specialization (defined as before age 12 years). While thought to be needed to acquire appropriate experience and excel in a given sport, there remains little information on when athletes at the highest levels of their sport specialized. This study aimed to define when professional and collegiate ice hockey players specialized.Hypothesis:Early sports specialization before age 12 years will not be common among elite-level (professional and collegiate) ice hockey players.Study design:Retrospective cross-sectional survey study.Level of evidence:Level 3.Methods:Male professional and collegiate ice hockey players within 1 National Hockey League organization and 2 National Collegiate Athletic Association (NCAA) organizations who were 18 years of age or older completed a survey at training camp detailing their history of sports participation and specialization.Results:A total of 91 athletes participated in the study (mean age, 22.8 years; range, 18-39 years). The mean age at the start of any sports participation was 4.5 years, and the mean age of sports specialization was 14.3 years. The mean age of specialization in the professional group, the NCAA Division I group, and the NCAA Division III group was 14.1, 14.5, and 14.6 years, respectively.Conclusion:Early pediatric sports specialization is not common in elite-level (professional and collegiate) ice hockey players.Clinical relevance:Early pediatric sports specialization before age 12 years is not necessary for athletic success in professional and collegiate ice hockey. This study provides further evidence supporting the recommendations of the American Medical Society for Sports Medicine, American Academy of Pediatrics, and American Orthopaedic Society for Sports Medicine against early sports specialization.

Dataset Information

Training load responses modelling and model generalisation in elite sports.

Publications

Training load responses modelling and model generalisation in elite sports.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets