GLP-1 FFM Systematic Review ACP 2026

Core PMIDs: PMID 39481534 (Annals ACP 2026) | PMID 41772149 (Nat Med 2026) | PMID 40544433 (NEJM 2025) Evidence grades: Confirmed > Supported > Reported > Contested > Gap Parent hub: GLP-1 Muscle Preservation


TL;DR

A landmark Annals of Internal Medicine systematic review (ACP 2026, PMID 39481534) found that 34.9% median of GLP-1 weight loss is fat-free mass (IQR 19–48.2%), with 68% of 36 RCTs exceeding the ~25% caloric restriction benchmark. The only clinically proven pharmacologic muscle preservation strategy is bimagrumab + semaglutide (Phase 2 RCT, PMID 41772149). CagriSema is approaching regulatory approval with plausible FFM benefit (PMID 40544433). The critical caveat: zero of 36 RCTs measured physical function (grip strength, SPPB, falls) as a primary endpoint — the clinical significance of FFM loss is unknown.


The 34.9% Figure — ACP 2026 Systematic Review

Evidence grade: Supported Source: Annals of Internal Medicine, ACP 2026 presentation; PMID 39481534

MetricValue
Median FFM proportion of weight lost34.9%
Interquartile range19–48.2%
RCTs exceeding ~25% diet benchmark68% (25 of 36)
Median RCT sample sizen=71 per study

Context: In a 15 kg weight loss on semaglutide, 34.9% FFM ≈ ~5.2 kg of lean mass lost. This is comparable to bariatric surgery proportions (25–30% FFM).

Important caveats:

  • The 25% comparator benchmark may not be appropriate for rapid pharmaceutical weight loss
  • DXA/BIA during rapid weight change measures glycogen depletion and fluid shifts as FFM loss — may overstate true skeletal muscle protein loss
  • IQR of 19–48.2% indicates enormous spread; median is not representative of any specific agent
  • No RCT has demonstrated functional impairment, falls, or fractures attributable to FFM loss

Cross-trial comparison of FFM% across GLP-1 agents — all indirect, no head-to-head DXA RCT exists:

AgentFFM% EstimateEvidence GradeNotes
Semaglutide 2.4 mg~25–35%SupportedDXA substudy STEP 1
Tirzepatide 15 mg~33–39%Reported (post-hoc)SURPASS program; no primary DXA endpoint
Retatrutide~33–38%ReportedPhase 2; glucagon component theoretically catabolic
Bimagrumab + semaglutideLean preserved/increasedConfirmed (Phase 2 RCT)Only proven pharmacologic strategy
CagriSemaNumerically better than semaglutideSupported (Phase 3)DXA not primary endpoint; FDA decision 2026–2027
MariTideUnknownGapGIP antagonism — could worsen muscle partitioning
OrforglipronUnknownGapNo DXA body composition data
SurvodutideUnknownGapGLP-2 component theoretically gut-protective

⚠️ No head-to-head DXA body composition RCT exists for any pair of GLP-1 agents. All comparative claims are hypothesis-generating only.


Bimagrumab + Semaglutide: The Only Proven Strategy

Evidence grade: Confirmed Source: Nat Med 2026; PMID 41772149; BELIEVE Phase 2 RCT (n=507), 48 weeks

The BELIEVE trial tested bimagrumab (ActRII/myostatin blockade) + semaglutide 2.4 mg across 9 arms.

ArmTotal Weight ChangeLean Mass ChangeLean % of Weight Lost
Semaglutide 2.4 mg alone−14.2 kgLost (−3.9 kg)28.9%
Bimagrumab 30 mg/kg alone−9.3 kgGained (+0.4 kg)>100% (anabolic)
Bimagrumab + semaglutide 2.4 mg−17.8 kgPreserved/increased7.3%

Key findings:

  • 67% reduction in proportional lean mass loss vs. semaglutide alone
  • Bimagrumab monotherapy was the only obesity pharmacotherapy with net anabolic lean mass signal in humans
  • Combo arm achieved the most total weight loss (−17.8 kg) with the least lean loss (7.3% of total)
  • No clinically significant BMD changes at 48 weeks

Critical limitation: BELIEVE did not measure physical function (grip strength, 6-minute walk). The sIBM bimagrumab extension data (PMID 32690797) showed +4.5% thigh muscle volume but progressive functional decline — pharmacological lean mass gain does not automatically translate to strength or mobility improvement.

Status: Investigational only. Bimagrumab approved for Barth syndrome (Sep 2025) — NOT for obesity/GLP-1 combo. Phase 3 required. No regulatory timeline announced.

See Bimagrumab Semaglutide Combo Obesity for full detail.


CagriSema: Approaching Approval, Plausible FFM Benefit

Evidence grade: Supported Source: REDEFINE-1 Phase 3 RCT; NEJM 2025; PMID 40544433; n=3,417

  • −20.4% total body weight loss at 68 weeks (superior to semaglutide monotherapy)
  • Body composition substudies show numerically better FFM preservation vs. semaglutide alone
  • DXA was not the primary endpoint — exact FFM% is not definitively quantified
  • FDA decision expected 2026–2027

Mechanistic rationale (Supported, not proven): Amylin receptor agonism in the hypothalamus may exert anabolic CNS effects that preserve muscle during caloric deficit. Amylin co-agonism also allows weight loss at lower semaglutide doses, potentially reducing catabolic stimulus.

Verdict: CagriSema is the most accessible plausibly muscle-sparing GLP-1 combination if approved. It is not proven to preserve muscle — only numerically better than semaglutide monotherapy in Phase 3 substudies.


The Critical Evidence Gap: Physical Function

Evidence grade: Gap

Zero of the 36 RCTs in the Annals systematic review measured physical function as a primary endpoint — no grip strength, no SPPB, no falls. This is the most critical gap in the field.

Implications:

  • The 34.9% FFM figure cannot be labeled clinically significant without functional outcome data
  • Whether FFM loss translates to increased falls, fractures, or mobility impairment is unknown
  • The semaglutide BMD reduction signal (lumbar spine modest reduction; no fracture rate increase demonstrated) is a monitoring flag, not a proven harm signal

High-risk populations with no data:

  • Elderly (>65) — excluded from most GLP-1 obesity RCTs
  • Sarcopenic or frail patients — excluded
  • Post-bariatric surgery — unstudied

DXA and Wearable Monitoring Guidance

Primary tool: DXA

DXA is the gold standard for tracking FFM changes during GLP-1 therapy. DXA was used in all pivotal body composition substudies.

TimepointAction
Baseline (week 0–1)DXA before or during first week of therapy
6 monthsDXA to assess body composition trajectory
12 monthsDXA for comprehensive assessment

Consumer BIA smart scales are not accurate enough to detect the 2–5 kg FFM changes typical of GLP-1 therapy in the short term (±3–5 kg error). Use for directional trends over 3–6 months only, measured consistently (same time, same hydration conditions).

Functional proxy: Grip strength every 3 months

  • Validated, cheap, field-ready proxy for overall muscle strength
  • No GLP-1-specific data exists, but it is the best functional proxy available
  • Decline >10–15% from personal baseline warrants clinical evaluation

Activity proxy: Step count monitoring

  • GLP-1 reduces NEAT and spontaneous activity
  • Monitor for sustained step count decline >20% from baseline — an early warning sign for activity-driven muscle loss

RMR and serum creatinine: not recommended as primary FFM monitoring tools — both are confounded by GLP-1-induced metabolic adaptation and hydration changes.

See GLP-1 Body Composition for the full detection model and alarm thresholds.


Evidence Summary

ClaimGradeSource
Median 34.9% of GLP-1 weight loss is FFM; 68% of 36 RCTs exceed 25% benchmarkSupportedACP 2026 / PMID 39481534
Bimagrumab + semaglutide preserves/increases lean mass (Phase 2 RCT)ConfirmedPMID 41772149
Bimagrumab monotherapy gains lean mass in humansConfirmedPMID 41772149
CagriSema numerically better FFM preservation vs semaglutide aloneSupportedPMID 40544433
Semaglutide modestly reduces lumbar spine BMD (no fracture rate increase)ReportedSTEP 1 substudy
Zero RCTs measured physical function as primary endpointGapPMID 39481534
Resistance training + protein prevents GLP-1 FFM lossGapNo registered trial exists
MariTide muscle effect unknownGapPMID 40549887
FFM loss comparable to bariatric surgery (25–30%)SupportedPMID 39481534

What Not to Claim

  • Do NOT claim tirzepatide “spares muscle” compared to semaglutide — no head-to-head DXA RCT
  • Do NOT claim the 34.9% FFM loss is “clinically significant” — zero functional outcome data
  • Do NOT claim MariTide spares or worsens muscle — no data
  • Do NOT claim CagriSema “preserves muscle” — only numerically better than semaglutide; not a primary endpoint


Supplement note — Batch 67 / April 2026 — GLP-1 FFM Systematic Review ACP 2026 Core PMIDs: PMID 39481534 | PMID 41772149 | PMID 40544433