Elevated GDF-15 tracked with higher death risk in three human cohorts

Is raising GDF-15 a longevity strategy?

No. Higher endogenous GDF-15 tracked with worse outcomes in all three human cohorts that measured it, never better ones ^[5]⁠^[6]⁠^[7].

Three cohorts in three disease contexts point the same direction. In 940 older men followed for a decade, each one-standard-deviation increase in log GDF-15 was associated with a 48% higher risk of cardiovascular mortality and a similar increase in total mortality after adjustment for standard risk factors ^[5]. In 1,531 patients with cancer, a doubling of GDF-15 was associated with a 57% higher hazard of death, along with higher risk of blood clots ^[6]. In 883 people with chronic kidney disease, each one-standard-deviation increase in GDF-15 carried an 87% higher adjusted hazard of death and was also linked to heart failure hospitalization ^[7].

Those are association studies, not causal proof that GDF-15 does the killing. GDF-15 rises in response to the same stresses that independently predict death: renal dysfunction, cardiac strain, and tumor burden. The direction of the association is unanimous across all three cohorts.

What does GDF-15 do in the body?

GDF-15 is a stress-induced cytokine that reduces food intake by acting on the brainstem receptor GFRAL. Tissue damage, mitochondrial stress, and certain drugs push cells to secrete more of it into circulation.

GFRAL is expressed almost exclusively in the brainstem, in the area postrema, a region that also processes nausea, vomiting, and taste aversion signals from toxins. Activating that circuit reduces food intake.

The mechanism proposed in the metformin work is not a satiety signal layered on top of normal appetite regulation. It runs through the same brainstem circuitry that evolved to make animals stop eating something that poisoned them ^[3].

That circuit is the throughline connecting metformin's appetite effect, hyperemesis gravidarum, and cancer cachexia. All three involve elevated GDF-15 acting on a receptor built for illness behavior rather than a clean metabolic dial.

Does GDF-15 cause nausea in pregnancy?

GDF-15's role in hyperemesis gravidarum is emerging rather than established. A systematic review notes that evidence for GDF-15 and its receptor in nausea and vomiting during pregnancy is emerging, and that the literature specific to GDF-15 in that condition remains thin ^[1].

That review was built around polyunsaturated fatty acid metabolism, and it flags GDF-15 as a mechanism worth pursuing rather than a settled cause ^[1].

The hyperemesis connection is plausible and consistent with the brainstem-aversion model. The available source is a review calling for more direct study, not a causal trial.

What does blocking GDF-15 do in cancer cachexia?

Blocking GDF-15 helped patients gain weight. A Phase 2 randomized, double-blind trial of ponsegromab, an antibody that blocks GDF-15, enrolled 187 patients with cancer cachexia and serum GDF-15 at or above 1500 pg per milliliter ^[2].

At 12 weeks, patients receiving the antibody gained significantly more weight than those on placebo. The median between-group difference was 1.22 kg at the 100 mg dose and 1.92 kg at the 200 mg dose ^[2].

That trial is the clearest human evidence available: neutralizing GDF-15 improved weight and appetite in people whose GDF-15 was already pathologically high. It argues that in a clinical population GDF-15 drives unwanted weight loss, and that removing it helps. It does not argue that adding more of it would help anyone else.

Why does metformin raise GDF-15?

Metformin raises circulating GDF-15, and in mice that rise is required for the drug's weight and appetite effects ^[3]. Two independent randomized controlled trials found that oral metformin increases circulating GDF-15, and in mice the drug's weight and appetite effects depended on GDF-15 and GFRAL, disappearing in animals lacking either ^[3].

That is trial-supported mechanistic evidence, with the causal part coming from the animal knockouts. The human arm establishes that metformin raises GDF-15, not that GDF-15 alone explains the full human weight effect.

Metformin is the one common exposure that reliably raises GDF-15, and it does so as a side effect of a drug taken for glycemic and weight reasons, not as a wellness intervention aimed at the biomarker ^[3].

Do GLP-1 drugs work through GDF-15?

No. A randomized trial of liraglutide, a GLP-1 receptor agonist, found weight loss without any change in GDF-15. Over 26 weeks in 94 people with type 2 diabetes, liraglutide produced weight loss but did not change GDF-15 levels, and changes in GDF-15 did not correlate with weight change ^[4].

That is useful negative evidence. GDF-15 and GFRAL are not the universal appetite-suppression pathway behind every weight-loss drug, only the one metformin appears to activate. Liraglutide's own outcomes trial covers what that drug does achieve, and the guide to using GLP-1 medications well covers the class mechanism.

Can anything optimize GDF-15 in a favorable direction?

No dose, food, fasting protocol, or supplement in the studies cited below has been shown to push GDF-15 in a favorable direction. The one intervention that moved GDF-15 beneficially did so in already-sick, cachectic patients, and it was an antibody rather than a supplement ^[2].

The 1500 pg/mL threshold used to enroll cachexia patients into the ponsegromab trial is a clinical cutoff for illness, not a target range to aim for from the other direction ^[2]. The lifespan evidence behind popular longevity peptides covers how other longevity claims in the category hold up.

Can GDF-15 suppress appetite without causing nausea?

No study cited below shows that separation. Whether GDF-15's appetite-suppressing effect can be pharmacologically separated from the nausea and aversion response is the central open question, and none of those studies answers it.

The animal knockout data show that metformin's weight effect and the GFRAL receptor are inseparable ^[3]. None of those studies demonstrates a drug that hits GFRAL and produces satiety without malaise.

Until that separation is shown, GDF-15's appetite effect and its aversive, illness-signaling effect look like one mechanism observed from two angles, not two effects waiting to be split apart.