Deng et al. 2025, supplement network meta-analysis (power, VO₂max, endurance)

A network meta-analysis of 30 RCTs (693 athletes) in Food Science & Nutrition comparing dietary supplements taken during conditioning training across anaerobic power, VO₂max and endurance performance. It excluded caffeine, sodium bicarbonate and branched-chain amino acids as outside its inclusion criteria.

For HMB specifically, it found no significant effect on VO₂max (SMD 0.28, 95% CI −0.05 to 0.61) or on endurance performance (SMD 0.29, 95% CI −0.44 to 1.03), while noting HMB ranked relatively high in the probabilistic ordering for both. No supplement in the analysis reliably improved VO₂max. HMB did improve anaerobic power: peak power (SMD 0.60, 95% CI 0.28 to 0.94, moderate certainty) and mean power (SMD 0.45, very low certainty), though it ranked behind protein, creatine and beta-alanine on those outcomes. The authors read HMB as most useful under heavy, muscle-damaging training loads, and hardest to show a benefit for in already well-trained athletes under normal load. This is the main counterweight to the more favourable endurance-specific meta-analysis of Fernández-Landa et al. 2024. Moderate evidence.