A Case Study using Physiological and Wellness Indicators for Performance Optimization in Basketball

Purpose This study investigated relationships between recovery-related biomarkers and game performance in professional basketball players to identify actionable indicators for optimizing athletic performance. Methods Thirteen professional male basketball players were monitored longitudinally over 15 weeks during games at 3-4 day intervals. Twelve independent variables—four salivary biomarkers (testosterone, cortisol, testosterone-to-cortisol ratio, salivary nitrates) and eight self-reported measures (anger, calmness, stress, energy, muscle soreness, tiredness, sleep duration, sleep quality)—were examined against 26 game performance metrics. Stratification analysis on four key outcomes (Plus/Minus, Efficiency, Player Impact Rating per minute, minutes played) identified non-linear relationships and predictor importance. Results Among the 69 significant correlations among recovery variables and game performance metrics, Tiredness and muscle soreness showed the strongest positive relationships, accounting for 39.1% of significant correlations. Sleep duration was the most consistent negative predictor (13 inverse correlations). Stratification revealed 20 significant associations with predominantly non-linear patterns. Sleep duration best predicted minutes played (H=18.47, p=0.0001, ε²=0.131); muscle soreness best predicted PIR/min (H=10.97, p=0.0042, ε²=0.081). Psychological variables primarily influenced playing time and Plus/Minus. Salivary biomarkers showed small but significant effects. Subjective recovery metrics demonstrated superior predictive value (43.8%) versus biomarkers (18.8%). Conclusions Recovery-performance relationships in elite basketball are complex, non-linear, and context-dependent. Subjective markers provide more informative readiness assessment than isolated biomarkers. Inverse sleep duration-performance associations suggest compensatory mechanisms, emphasizing individualized interpretation. Practical monitoring should prioritize daily self-reports, player-specific reference ranges, and integrate recovery data with training load contextual factors.