Related Experiment Video
Updated: Aug 7, 2026

Assessing the Accuracy of Fitness Smartwatch Data for Cardiovascular and Physical Activity Monitoring: A Validation Study in Digital Health
Published on: February 21, 2025
The relative age effect among Malaysian university athletes: a cross-sectional survivorship analysis with exploratory
Haashwein Moganan1, Mohansundar Sankaravel2, Gunathevan Elumalai2
1Department of Coaching Science, Faculty of Sport Science and Coaching, Sultan Idris Education University, Tanjong Malim, Perak Darul Ridzuan, Malaysia.
Abstract:
The relative age effect (RAE) is defined as the overrepresentation of athletes born early in the selection year and is well documented in talent identification. However, whether the selection pressures that produce the RAE leave a detectable biomechanical signature in the athletes who reach elite level remains unresolved. This cross-sectional study examined the relative age effect among national-level Malaysian university athletes and, as a secondary exploratory question, whether any biomechanical loading signature accompanied it. Specifically, it tested whether (i) selection asymmetry consistent with the RAE was present, (ii) the RAE differed by gender, and (iii) AI-derived loading-risk indices differed across birth quartiles. A total of 170 athletes (116 males, 54 females; M = 18.6, SD = 0.7 years) were assessed using the Holomotion markerless motion capture system, generating four pre-specified outcomes: a composite total movement score, an exercise risk score, a joint pain risk index, and a ligament strain risk index (the latter two reflecting modelled tissue-level loading rather than clinical injury events). The birth quartiles were derived from national identification numbers using a 1st January cut-off. A significant RAE was observed overall [χ 2(3) = 11.74, p = .008, w = 0.26] and in males [χ 2(3) = 8.55, p = .036, w = 0.27; Q1:Q4 = 2.06], but not in females [χ 2(3) = 3.78, p = .286]. No loading-risk index differed across quartiles, either as continuous scores (Kruskal-Wallis, all p ≥ .720; η 2 ≈ 0) or as binary high-risk classifications (chi-square, all p ≥ .660; Cramér's V ≤ 0.10). The selection-related birthdate asymmetry was therefore not accompanied by detectable differences in AI-derived biomechanical loading estimates, a pattern consistent with a survivorship interpretation.
