Back

Developmental trajectories of aerobic endurance performance in highly trained youth soccer players aged 14 to 19 years.

Ruf, L.; Haertel, S.; Altmann, S.

2025-12-04 sports medicine
10.64898/2025.12.04.25341632 medRxiv
Show abstract

PurposeWe aimed to identify distinct clusters of developmental trajectories in aerobic endurance performance among highly trained youth soccer players aged 14 to 19 years, moving beyond traditional group-level analyses to characterize potential heterogeneity of individual developmental pathways. MethodsMixed-longitudinal data from 149 male youth soccer players from a professional German academy were analyzed. Aerobic endurance was assessed using velocity at individual anaerobic threshold (vIANT) and velocity at 4 mmol.L-1 blood lactate (v4) from incremental treadmill tests. A hybrid clustering framework combining Generalized Additive Models (GAM) with model-based clustering (Mclust) was employed. Individual trajectories were estimated using GAMs with factor smooth basis functions, followed by comprehensive feature extraction (15 trajectory characteristics). Growth mixture models were applied across 3-7 clusters with multiple covariance structures, with model selection based on weighted consensus scoring emphasizing BIC, silhouette coefficient, and AIC. ResultsFour distinct developmental trajectory clusters were identified with consistent patterns across both vIANT and v4. The overall population mean trajectory demonstrated minimal systematic change across the age range. However, substantial between-player heterogeneity was evident: approximately one-fifth of players showed substantial improvement followed by stabilization, one-quarter exhibited progressive decline, while the majority displayed minimal change patterns with either slight improvement or slight decline. Endurance performance converged with increasing age across all clusters. ConclusionAerobic endurance development in highly trained youth soccer players follows distinct individual trajectories rather than uniform patterns during adolescence. These findings highlight that group-level analyses mask meaningful individual differences, emphasizing the importance of individualized talent evaluation and training prescription.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.