THE EFFECTIVENESS OF COLD BEVERAGES VERSUS ICE-SLURRY DRINKS ON THE ATHLETIC PERFORMANCE OF THAI FUTSAL PLAYERS USING K-MEANS CLUSTERING

Abstract

Cooling interventions during futsal halftime breaks show substantial individual variability in physiological responses, yet standardized protocols fail to account for athlete-specific thermal stress susceptibility. This study employed This study used K-means clustering to compare the effectiveness of cold beverages versus ice slurry and to identify distinct physiological response phenotypes for personalized cooling strategy optimization. Ten competitive male futsal players (22.4 ± 2.1 years; 68.5 ± 8.2 kg) completed a randomized crossover design. Following the Futsal Intermittent Shuttle-Run Protocol (FIRP), participants consumed either ice slurry (-1°C) or cold sports beverages (4°C) at 7.5 g/kg body mass during 10-minute recovery. Futsal-specific reactive agility tests (RAG-D, RAG-T), blood lactate, heart rate, urine specific gravity, and perceived exertion were measured. K-means clustering analysis with silhouette validation identified response patterns. Three distinct physiological phenotypes emerged (silhouette coefficient = 0.67). Cluster 1 (High-Response, n=4): elevated blood lactate (>8.0 mmol/L), highest cardiovascular stress, superior ice-slurry response. Cluster 2 (Moderate-Response, n=3): balanced responses to both modalities. Cluster 3 (Low-Response, n=3): conservative responses with maintained performance, preferential ice-slurry benefits. Strong correlations existed between body mass and response magnitude (r = 0.78, p < 0.01). Unsupervised machine learning effectively discerned unique cooling response phenotypes, facilitating evidence-based customization of cooling therapies. This signifies a substantial progression in the accuracy of sports performance enhancement.