News

Nutrition

Supplements and jump performance: HMB stands out, most evidence is very low certainty

By Monica O'Reilly

Published · Updated

Plyometric box, training shoes, a bowl of white powder and a closed notebook on a wooden floor
Illustration; not the study materials or participants. Illustration: Optimize Protocols

What the analysis compared

A Bayesian network meta-analysis in the Journal of the International Society of Sports Nutrition pooled 119 randomised trials of dietary supplements. Participants were healthy people with exercise backgrounds, from recreationally active to world-class athletes. Four separate networks compared supplements with placebo on peak power (W), relative peak power (W per kg), vertical jump height and countermovement jump height (cm).

HMB had the most consistent result

Against placebo, HMB raised peak power by 64.70 W (95% credible interval 19.15 to 110.50, two direct studies) and relative peak power by 0.84 W per kilogram (0.42 to 1.28, four direct studies). Both comparisons were rated moderate certainty, and the relative peak power result held across several sensitivity analyses.

Prediction intervals, which describe the range expected in a future study, also excluded zero for both HMB results. The authors describe the number of direct studies as limited and call for validation in independent direct comparisons.

Other supplements

Beta-alanine raised peak power by 40.96 W (5.10 to 76.78), and creatine and vitamin D raised relative peak power by 0.36 and 0.66 W per kilogram. For countermovement jump, betaine added 3.72 cm (1.24 to 6.18) from two direct studies. Creatine, caffeine, palmitoylethanolamide, multi-ingredient products and caffeine with Rhodiola rosea also had intervals that excluded zero.

Why the results are fragile

For beta-alanine, creatine, vitamin D and most countermovement jump results, the prediction intervals included zero. Iron raised vertical jump by 6.81 cm (0.84 to 12.83), but one study with 11 supplemented participants supports it. Studies mixed doses, durations, training levels and jump test procedures within each supplement.

Rankings and certainty

Of 1,636 pairwise comparisons, 17 were rated moderate certainty, 55 low and 1,564 very low. The three highest-ranked peak power supplements each rested on a single study, with credible intervals that included zero. The authors state that rankings should not guide individual supplement plans.

Registration and limits

The review was registered retrospectively. Screening and data extraction began in 2025, the protocol was submitted to PROSPERO in January 2026 and registration completed in June 2026 (CRD420261283835). Only peer-reviewed English-language papers were included, and some reports lacked data for pooling. Differences between the four outcome networks cannot show that outcome type changes a supplement's effect.

Sources (1)

This is news reporting and not medical advice.