FGF21 analogs show modest benefits for MASH in early trials

Is FGF21 a proven longevity hormone?

No. The human evidence on FGF21 analogs in the pooled analyses comes from Phase 2b and earlier trials in liver disease (MASH), not from longevity studies.

The "exercise mimetic, body-composition hormone" narrative traces back to transgenic mice that overexpress FGF21. Those mice do lose weight and show improved insulin sensitivity.

None of the cited studies measured whether the circulating FGF21 concentrations reached by transgenic overexpression, or by pharmacologic dosing of an analog, resemble what fasting or exercise raises endogenous FGF21 to in humans. That gap between rodent overexpression and human physiology is exactly what the biohacker framing skips over. Any claimed equivalence between fasting or exercise and FGF21 analog dosing is unproven.

The same gap between hormone biology and longevity marketing runs through a stress signal sold as a longevity lever, and the wider review of longevity peptides covers how other lifespan claims hold up.

How does FGF21 signal in the body?

FGF21 is produced mainly in the liver and signals through a receptor complex built from FGFR1c and an obligate co-receptor, beta-klotho (KLB). Genetic association work confirms that beta-klotho is necessary for FGF21 to act in the brain, where FGF21 influences macronutrient and alcohol preference in addition to its metabolic effects in liver and adipose tissue ^[8].

That multi-tissue receptor requirement is the mechanistic basis for calling FGF21 a liver-adipose-brain signal rather than a single appetite-suppression pathway like GLP-1.

How much do FGF21 analogs reduce liver fat?

FGF21 analogs reduced hepatic fat fraction on MRI by a mean of roughly 47% relative to placebo (95% CI -58.8 to -35.3) in pooled randomized, placebo-controlled trials ^[5]. That result ranked FGF21 analogs second among five compared drug classes, behind GLP-1-based polyagonists and ahead of THR-beta agonists and GLP-1 receptor agonists alone ^[5].

The same analysis found FGF21 analogs were the most effective class tested for reducing liver stiffness on elastography ^[5].

Liver enzymes moved in parallel. In efruxifermin trials, ALT fell by a mean of about 14 U/L and AST by about 13 U/L versus placebo, alongside a drop in the ELF fibrosis biomarker score ^[2].

Do FGF21 analogs reverse liver fibrosis?

In pooled trials, FGF21 analogs improved fibrosis more than placebo, but the signal is uneven: efruxifermin shows no significant benefit at the stricter ≥2-stage threshold ^[2]⁠^[3]. MRI fat fraction is a surrogate marker. Biopsy-confirmed fibrosis improvement without disease worsening is the harder endpoint.

Pooling eight trials (963 patients) of efruxifermin, pegbelfermin, and pegozafermin, the class produced significantly more ≥1-stage fibrosis improvement (RR 1.83, 95% CI 1.27-2.62) and more ≥2-point NAFLD activity score improvement (RR 2.85, 95% CI 2.06-3.95) than placebo ^[3].

Efruxifermin alone, across four RCTs (419 patients), improved fibrosis in 39.7% of treated patients versus 17.1% on placebo (RR 1.83, 95% CI 1.17-2.84) ^[2]. MASH resolution occurred in 45.1% versus 13.5% (RR 2.42, 95% CI 1.48-3.94) ^[2]. A separate efruxifermin-only pooling found a comparable fibrosis-improvement risk ratio of 1.97 (95% CI 1.21-3.19) ^[9].

The caveat rarely reaches consumer coverage. Efruxifermin showed no significant benefit for ≥2-stage fibrosis improvement (p = 0.25), and a trial sequential analysis of the same pooled data concluded that the accumulated evidence is not yet statistically sufficient to call the fibrosis signal settled ^[2].

In compensated MASH cirrhosis, efruxifermin was the only agent among nine trial arms to significantly beat placebo on fibrosis regression, ranking highest by SUCRA (77.44) ^[4]. That is one compound at one disease stage, not proof that the whole class works equally well in advanced fibrosis.

Which FGF21 analog has the strongest evidence?

Pegozafermin ranked first for fibrosis regression in a 29-trial network meta-analysis, and efruxifermin ranked first in compensated cirrhosis ^[1]⁠^[4].

Pooling 29 RCTs (9,324 patients) spanning the full MASH drug landscape, pegozafermin ranked as the single most effective agent for fibrosis regression without worsening MASH (SUCRA 79.92) ^[1].

A separate 48-trial network focused on fibrosis ranked pegbelfermin (SUCRA 77.11%) and pegozafermin (SUCRA 74.91%) highest at the F1-3 fibrosis stage. Pegozafermin also significantly beat placebo for reducing liver stiffness on elastography in that analysis ^[7].

Efruxifermin's clearest standout result is the cirrhosis-stage ranking ^[4]. Pegbelfermin has fibrosis-ranking data in the 48-trial analysis ^[7] but a thinner trial record than pegozafermin or efruxifermin across these analyses.

Do FGF21 analogs cause weight loss?

FGF21 analogs cause modest but real weight loss: pooled across eight randomized trials, body weight fell significantly versus placebo ^[6]. "No weight loss" overstates the null.

The same pooled analysis found reductions in fasting insulin and total cholesterol, and no significant effect on fasting glucose, HbA1c, HOMA-IR, free fatty acids, or systolic blood pressure ^[6]. Evidence quality behind those pooled outcomes ranged from moderate to very low, limited mainly by imprecision ^[6].

No reported weight-loss magnitude for FGF21 analogs approaches the double-digit percentage reductions associated with GLP-1 receptor agonists or dual agonists tested in comparable MASH populations. The network analyses rank FGF21 analogs on liver fat and fibrosis outcomes, not weight ^[1]⁠^[5].

That absence is informative. The pooled data support FGF21 analogs as a liver and lipid-remodeling drug class, with weight change as a modest secondary effect rather than the driver of benefit. The guide to using GLP-1 medications well covers the drug class behind those double-digit reductions.

What side effects do FGF21 analogs cause?

FGF21 analogs cause more adverse events than placebo, and efruxifermin roughly triples discontinuations for adverse events ^[2]⁠^[3]. Across the class, treatment-emergent adverse events run higher than placebo (RR 1.17, 95% CI 1.08-1.27), and treatment-related adverse events are roughly 75% more common (RR 1.75) ^[3].

In efruxifermin-specific pooling, drug-related adverse events occurred in 71.2% of treated patients versus 43.7% on placebo (RR 1.35) ^[2]. Discontinuations for adverse events were about three times more common on drug (10.3% vs 2.2%, RR 3.13), though serious adverse events were not significantly increased ^[2].

None of the cited trials reports what happens to liver fat or fibrosis after FGF21 analog treatment stops.

What questions do the FGF21 analog trials leave open?

The FGF21 analog trials leave four questions open: surrogate validity, durability after stopping, the fasting comparison, and long-term bone and cardiovascular safety.

• Surrogate versus biopsy. No trial establishes whether MRI-measured fat reduction reliably predicts biopsy fibrosis improvement across every candidate and disease stage.
• Durability. How long any benefit lasts after treatment stops is unreported.
• Fasting and exercise. Whether fasting- or exercise-induced FGF21 elevation reaches concentrations physiologically comparable to pharmacologic dosing is unmeasured in these trials.
• Bone and cardiovascular safety. Long-term safety is unestablished, because none of the meta-analyses reports bone turnover markers or extended outcome data.

No FGF21 analog has regulatory approval. The entire evidence base cited below is Phase 2b and earlier.