Weight loss is supposed to be the success metric for GLP-1 therapy. A new meta-analysis suggests we may be measuring the wrong thing — or at least not measuring enough.

Researchers synthesized 19 randomized controlled trials from the STEP, SURMOUNT, SCALE, and OASIS programs — the pivotal trials behind semaglutide, tirzepatide, and liraglutide — and focused specifically on nutritional outcomes during high-potency incretin therapy.

What they found deserves more attention than it's getting.

The Gap Between What Was Logged and What the Labs Showed

Investigator-reported malnutrition during treatment: 0.12%

That's a very low number. If you stopped there, you'd conclude nutritional risk is negligible on GLP-1 therapy.

But the labs told a different story.

Low total lymphocyte counts (below 910 per microliter — a marker of nutritional status): 2.90% of treated participants vs. 1.77% on placebo

That gap — between what clinicians documented as an adverse event and what blood work actually detected — is the central finding of this paper.

These were trials run by the world's largest pharmaceutical companies, with extensive safety monitoring. If laboratory-defined nutritional deficits showed up at meaningful rates even in those trials, the question for real-world clinical practice is: what are we not catching between quarterly appointments?

Source: Ampofo E, Apprey C, Amoako M, Turkson FD. "A Systematic Review and Meta-Analysis of Malnutrition and Metabolic Failure in High-Potency Incretin Therapy." Obesity Science & Practice. 2026;12(5):e70188. PMID 42707648. DOI 10.1002/osp4.70188.

The Lean Mass Problem

The malnutrition finding sits alongside data we've covered extensively: lean mass loss during GLP-1 therapy is real, clinically meaningful, and often under-discussed.

From the SURMOUNT-1 DXA substudy (tirzepatide, 72 weeks, 160 participants):

Total weight lost: 21.3%
Of that weight: 74% fat mass, 26% lean mass
Lean mass fell 10.9% vs. 2.6% on placebo — an estimated treatment difference of 8.3 percentage points (the substudy measured lean mass by DXA, not fat-free mass)
In the meta-analysis, tirzepatide 15 mg was associated with a mean fat-free mass reduction of 1.60 kg — 2.80% of body weight

From STEP 1 (semaglutide, 68 weeks, DEXA substudy):

Lean mass loss: approximately 6.9 kg over 68 weeks on semaglutide 2.4 mg — about 40% of the weight lost (DXA substudy, 95 semaglutide participants)
Total lean body mass −9.7% from baseline in the exploratory DXA analysis

The critical point: 26% (SURMOUNT-1) to roughly 40% (STEP 1) of all weight lost being lean tissue is not a rounding error. In a person who loses 40 lbs, that's approximately 10 lbs of muscle and organ tissue — not fat.

Why Trials Undercount This

The meta-analysis makes an important methodological observation: the underlying trials were not designed to diagnose malnutrition. Malnutrition was not a primary or secondary endpoint. Nutritional screening was not standardized across sites.

This means investigator-reported malnutrition rates of 0.12% likely reflect the limit of detection in an outcome framework that wasn't looking for it — not the true prevalence.

The lymphocyte signal (2.9% vs. 1.77%) came from standard lab panels already being run for safety monitoring. It showed up anyway. That's a passive detection of something the trials weren't actively hunting.

The implication: if you design a trial specifically to find nutritional deterioration in GLP-1 users, you'll probably find more of it than 0.12%.

Who Is Most at Risk

The meta-analysis authors specifically flag older adults as warranting structured nutritional screening.

The reasoning is straightforward: older adults have lower baseline lean mass reserves, higher risk of sarcopenia, reduced protein synthesis efficiency, and often lower baseline dietary protein intake. When GLP-1 therapy suppresses appetite by 20-30%, the reduction in total caloric intake hits protein adequacy harder in someone already eating suboptimally.

Additional higher-risk groups based on the evidence:

People with the most aggressive caloric restriction (larger deficits = more lean mass at risk)
People who are not resistance training (muscle protein synthesis requires mechanical stimulus, not just protein)
People with baseline nutritional deficits (vitamin D, B12, iron — common before therapy starts)
People using the highest-potency agents (retatrutide produces the largest deficits and therefore the most total tissue movement)

The Right Way to Think About GLP-1 Success

The old outcome model for GLP-1 therapy:

Starting weight: 220 lbs → End weight: 170 lbs = Success.

That framework misses everything downstream of the scale.

The more complete outcome picture requires asking:

1. How much of the weight lost was fat vs. lean mass? A DXA or InBody scan at baseline and at 3-6 month intervals answers this. A scale doesn't.

2. Is protein intake adequate? GLP-1 therapy suppresses total caloric intake, but protein needs don't decrease proportionally. In a large caloric deficit, protein adequacy requires active monitoring — not just telling someone to "eat more protein."

3. Is strength improving or declining? Grip strength, squat capacity, or any validated functional strength measure tracks what the lean mass data can't fully capture — whether the tissue being preserved is functional.

4. What happened to visceral fat specifically? Total weight loss and visceral fat reduction are related but not identical. SURMOUNT-1 showed -40.1% visceral fat on tirzepatide. That's the metabolically meaningful number — not total pounds.

5. What are the downstream biomarkers doing? Inflammatory markers, liver enzymes, insulin sensitivity, lipid panels — these are the actual cardiovascular and metabolic outcomes that GLP-1 therapy should improve. Weight is a proxy for them.

6. Can the patient maintain the outcome? Without resistance training and adequate protein, lean mass lost during GLP-1 therapy doesn't automatically return when the drug is stopped. The maintenance question requires a plan, not just a prescription.

What This Means for GLP-1 Programs

A program that prescribes semaglutide or tirzepatide, schedules quarterly check-ins, and tracks weight is doing the minimum. The gap between the minimum and adequate monitoring is where nutritional deterioration occurs invisibly.

The standard of care needs to include:

Body composition measurement at baseline and follow-up (not just BMI)
Nutritional screening with standardized tools (not just "how are you eating?")
Resistance training as a protocol requirement, not an optional recommendation
Lab monitoring that specifically looks for nutritional markers (lymphocytes, albumin, micronutrients)
Functional strength assessment alongside weight

That's the GLP-1 2.0 framework. Not as a marketing position — as a response to what the data is showing.

→ Track your body composition at PeptidesGPT.com

Sources:

Ampofo E, Apprey C, Amoako M, Turkson FD. "A Systematic Review and Meta-Analysis of Malnutrition and Metabolic Failure in High-Potency Incretin Therapy." Obesity Science & Practice. 2026;12(5):e70188. PMID 42707648. DOI 10.1002/osp4.70188.
Look M, Dunn JP, Kushner RF, et al. "Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study of adults with obesity or overweight." Diabetes Obes Metab. 2025;27(5):2720–2729. PMID 39996356. DOI 10.1111/dom.16275.
Wilding JPH, et al. "Once-Weekly Semaglutide in Adults with Overweight or Obesity" (STEP 1), N Engl J Med. 2021;384:989–1002, supplementary appendix (DXA substudy). DOI 10.1056/NEJMoa2032183. Exploratory DXA analysis: J Endocr Soc. 2021;5(Suppl 1). DOI 10.1210/jendso/bvab048.030.
Tinsley GM, et al. Case series and review of lean soft tissue loss with incretin therapy. SAGE Open Med Case Rep. 2025. PMID 41122508. DOI 10.1177/2050313X251388724.

PeptidesGPT is an educational platform. This content is for informational purposes only and does not constitute medical advice. Nutritional decisions during GLP-1 therapy should be made in consultation with a licensed healthcare provider.

Are GLP-1 Users Becoming Undernourished? What a 19-Trial Meta-Analysis Found