HMB AFTER 50: FOUR EVIDENCE SYNTHESES, THREE DIFFERENT ANSWERS

HMB is sold as the muscle-preservation supplement for people over 50. Four recent evidence syntheses have looked at it — and the most rigorous design among them, the one that put resistance training in both arms, is also the one where almost everything washed out.

HMB is β-hydroxy-β-methylbutyrate, a metabolite of the amino acid leucine. Your body makes a small amount of it from the leucine in dietary protein. The supplement industry's pitch is that taking it directly does something the leucine in food doesn't, specifically for people losing muscle with age. Here is what the trials actually measured, one synthesis at a time.

The Broadest Look: 21 Trials, Everyone Over 50

Li and colleagues published a meta-analysis in Frontiers in Nutrition in April 2025 covering 21 randomized controlled trials and 1,935 participants, all over 50 years old.

It found positive effects across every category it examined:

• Appendicular skeletal muscle mass — weighted mean difference 1.56 kg (95% CI 0.03 to 3.09)
• Lean mass — 0.28 kg (95% CI 0.16 to 0.41)
• Handgrip strength — 0.54 kg (95% CI 0.04 to 1.04)
• Five-time chair stand test — 0.73 seconds faster (95% CI −1.35 to −0.11)
• Gait speed — 0.06 m/s (95% CI 0.01 to 0.10)

Subgroup analysis pointed to 3 g/day for longer than 12 weeks as the dose and duration where effects showed up. Note how close several of those confidence intervals run to zero — the muscle mass interval's lower bound is 0.03 kg, which is to say roughly nothing.

The Narrower Look: Clinically Diagnosed Sarcopenia

Gu and colleagues, in Maturitas in April 2025, restricted the question to patients diagnosed with sarcopenia by accepted clinical consensus. Five RCTs met that bar — 154 participants for muscle mass, 359 for strength and physical performance. Risk of bias was low in four of the five.

They found a benefit on skeletal muscle mass index (SMD 0.32, 95% CI 0.00 to 0.64, P = 0.048) and on handgrip strength (SMD 0.65, 95% CI 0.05 to 1.25, P = 0.034). They found nothing on gait speed (SMD 0.19, 95% CI −0.14 to 0.53, P = 0.255).

That muscle mass result has a lower confidence bound of exactly 0.00 and a P value of 0.048. It cleared the line and not much more. The authors' own framing: the evidence is "limited and requiring interpretation with utmost caution."

The Design That Matters Most: Training in Both Arms

Here is where it gets interesting, and where anyone already lifting should pay attention.

Feng and colleagues, in Frontiers in Nutrition in September 2024, asked a different and harder question: not "does HMB help?" but "does HMB add anything on top of exercise?" Five RCTs, 257 older patients with sarcopenia, comparing exercise plus HMB against exercise plus placebo.

Nearly everything came back null:

• Skeletal muscle mass index — SMD 0.06 (95% CI −0.20 to 0.32), P = 0.66
• Grip strength — SMD 0.23 (95% CI −0.05 to 0.52), P = 0.11
• Five-time chair stand — SMD −0.83 (95% CI −1.88 to 0.21), P = 0.12
• Fat-free mass — SMD 0.04 (95% CI −0.26 to 0.35), P = 0.78
• Fat mass — SMD 0.01 (95% CI −0.25 to 0.27), P = 0.94
• BMI — SMD −0.09 (95% CI −0.43 to 0.25), P = 0.60

One outcome did move: gait speed, SMD 0.48 (95% CI 0.15 to 0.82), P = 0.005. Which is the exact outcome the other two syntheses found nothing on. The authors' conclusion is that effects on muscle mass, strength, and body composition "are likely minimal," and they flag that the finding rests on five studies and needs further validation.

Set the three side by side and the pattern is hard to miss. When HMB is compared against a control group that isn't training, it looks like it builds muscle and strength. When the control group is training, that advantage largely disappears and a different outcome moves instead. Three syntheses, three different answers about which measure responds.

The View From Above

Bideshki and colleagues published an umbrella review in the Journal of Cachexia, Sarcopenia and Muscle in February 2025 — a meta-analysis of the meta-analyses. Eleven studies, 41 datasets, participants aged 23 to 79.

Pooled effect sizes: muscle mass 0.21 (95% CI 0.06 to 0.35), muscle strength index 0.27 (95% CI 0.19 to 0.35), fat-free mass 0.22 (95% CI 0.11 to 0.34). Fat mass and body mass showed no significant change.

Effect sizes of 0.21 to 0.27 are small by the usual conventions. They are statistically real and they are not large. The authors' own quality assessment is worth quoting alongside the numbers: of the eleven included studies, five rated high quality on AMSTAR2, three low, and three critically low. Their conclusion scopes the benefit to people "experiencing muscular atrophy due to physiological conditions" — not to healthy trained adults.

What I Could Not Find

I could not find a randomized trial testing HMB in trained masters athletes. The literature I searched sits in sarcopenic, frail, clinical, and untrained older populations. That is a meaningfully different person from a 58-year-old who has been lifting three days a week for fifteen years, and the Feng result — where an already-exercising control group erased most of the effect — is the reason that gap matters rather than being a technicality.

None of these syntheses measured falls, fractures, disability, or mortality either. The outcomes are muscle mass, strength, and walking speed over weeks to months.

The Protocol

What the trials that found effects actually used:

• 3 g/day, the dose the subgroup analysis in Li 2025 identified
• Longer than 12 weeks — shorter durations did not show the effect
• Calcium-HMB was the form in the trials specifying one
• Resistance training alongside it in most protocols, which is also the variable that shrank HMB's apparent advantage when the control group had it too

The honest summary of four syntheses: the effect sizes are small, they are largest in people who are frail or clinically sarcopenic and not training, and the single design that controlled for training found almost nothing except a walking speed change the other reviews didn't replicate. Whether that is worth the cost of a daily supplement is a judgment about your own situation, not something these papers answer.

Resistance training is in every one of these protocols, including the placebo arms. It is doing the heavy lifting in the literature, in both senses.

Battle Hard. — Will Power