GDF11 and GDF15 — the controversial aging factors discovered in young blood

In 2013 and 2014, work from Wagers's lab put a candidate on the table. GDF11 — growth differentiation factor 11, a member of the TGF-beta superfamily of signaling proteins — appeared to decline with age in mice and to reverse aspects of cardiac aging when administered to old animals. In one widely cited paper, GDF11 supplementation reduced cardiac hypertrophy in old mice, a hallmark of aging hearts. Subsequent papers from the same group reported effects on skeletal muscle regeneration and brain vasculature. GDF11 was presented as a candidate youth factor — a circulating protein whose decline with age contributed to age-related tissue deterioration and whose restoration might reverse some of that deterioration.

The scientific community's response was initially enthusiastic and then complicated.

In 2015, a group at Novartis published what became the central challenge to the GDF11 story. Using what they argued were more specific and better-validated immunoassays, they found not that GDF11 declined with age in mice and humans, but that it increased with age or remained stable. They also found that GDF11 treatment appeared to impair rather than improve skeletal muscle regeneration in at least one experimental model. The implication was that the Wagers lab's findings were artifacts of assay cross-reactivity: the antibodies used to measure GDF11 might have been detecting a closely related protein, GDF8 (also known as myostatin), which is structurally very similar to GDF11 and for which commercially available antibodies have known cross-reactivity problems. If GDF11 was not actually declining with age, the entire rationale for it as a youth factor that could be supplemented back collapsed.

The dispute that followed was unusually public and technically detailed. Wagers's lab responded that the Novartis assay had its own limitations and that the different findings might reflect genuine biological differences in the animals and tissues studied rather than purely methodological artifacts. They and collaborators published subsequent work with additional methods supporting some aspects of the original story. The Novartis group and other independent labs have continued to find divergent results in different contexts. The current state of the GDF11 literature is genuinely unsettled: it appears that GDF11's aging trajectory may differ across tissues and species, that its effects on regeneration may differ across tissue types (beneficial in cardiac and neural contexts, potentially inhibitory in skeletal muscle), and that the antibody specificity problem has made it very difficult to draw clean conclusions from any single study that does not use orthogonal methods to confirm its measurements.

This is not unusual territory in biomedical science. New discoveries in high-profile areas attract rapid follow-up, follow-up reveals unexpected complexity, and the field spends years reconciling data generated under different assumptions with different tools. What made the GDF11 story different was the speed at which it left the scientific literature and entered consumer culture. Startup companies were discussing parabiosis-adjacent approaches to human aging before the antibody controversy had even fully developed. The concept of a young blood rejuvenation factor was easy to communicate, resonated with something people intuitely believed, and had dramatic animal data behind it — even data that turned out to be contested. The gap between preliminary findings and popular narrative was unusually short and unusually wide.

GDF15 is a different and somewhat cleaner story, though it intersects with GDF11's neighborhood in ways worth understanding. GDF15 is also a TGF-beta superfamily member, but its biology is quite distinct. Rather than a potential youth factor, GDF15 is, in most contexts, a stress response cytokine. It is produced in response to cellular stress — mitochondrial dysfunction, DNA damage, inflammation, proteotoxic stress — and it rises with age in humans in a consistent and reproducible pattern across populations. It also rises in response to metformin treatment, sometimes dramatically; the GDF15 elevation produced by metformin may actually account for some of metformin's appetite-suppressing effects, because GDF15 signals through the GFRAL receptor in the area postrema of the brainstem — a brain region with a weak blood-brain barrier — to induce nausea and reduce food intake. This makes GDF15 a molecular mediator of a known clinical side effect of a widely used drug, which is a useful kind of discovery.

As a biomarker, GDF15 has attracted interest as a potential indicator of mitochondrial stress and biological aging trajectory. Its levels correlate with biological age estimates derived from epigenetic clocks and other aging biomarkers, and its elevation in the context of various diseases — heart failure, cancer, CKD, metabolic syndrome — tracks with adverse outcomes. Whether measuring GDF15 in healthy individuals tells you something actionable about their aging rate, and whether interventions that lower GDF15 produce better outcomes or simply reflect the fact that a less-stressed metabolism generates less of it, are questions that ongoing research is addressing but hasn't fully resolved.

The parabiosis work underlying both stories has continued to develop, and the picture has become more complex than the early results suggested. Subsequent work identified other candidate rejuvenating and aging-promoting circulating factors beyond GDF11. The fraction of plasma that appeared to carry the regenerative effects shifted in different experiments — sometimes attributed to proteins, sometimes to extracellular vesicles, sometimes to other small molecules. Work from Tony Wyss-Coray's lab at Stanford using plasma transfer approaches has suggested that young plasma may contain factors broadly beneficial to aged brains, and that old plasma may contain factors that actively suppress hippocampal function, with specific proteins identified as candidates in each direction. The multi-factor nature of the blood-borne aging signal makes the original single-protein GDF11 story look like an underestimate of complexity rather than simply an error.

The clinical translation space in this area has ranged from serious scientific programs to approaches that have drawn regulatory and ethical concern. The FDA issued a warning in 2019 against commercial young plasma infusion businesses, noting the lack of evidence that such infusions produced the benefits being marketed and the potential for harm from plasma products. The warning was not directed at the scientific research but at direct-to-consumer applications that had outpaced the evidence. The underlying science of circulating aging factors is legitimate and ongoing; the consumer applications that preceded the science are a different matter.

There is a navigational lesson in the GDF11 controversy that applies broadly to preliminary aging research, and it is worth making explicit. The structure of the problem is as follows: an animal study produces striking results in a short-lived model organism; the effect is mechanistically interesting; early papers are published in high-impact journals; a narrative forms around the finding; consumer products or services are developed in advance of replication; subsequent work reveals that the picture is more complicated than the original finding suggested or that the original finding may be partly artifactual; the consumer narrative is slow to update because the original story was compelling and the technical rebuttal is dense. This cycle has played out in longevity biology multiple times — not only with GDF11 but with aspects of the resveratrol story, with telomere-targeting approaches, and with certain stem cell applications.

The appropriate response to this pattern is not cynicism about aging science, which would miss the genuine and robust biology underlying many findings. It is calibration: distinguishing between what has been well-replicated across independent laboratories using orthogonal methods, what has been shown in a limited number of studies with known methodological limitations, and what has been communicated in popular venues in ways that have outrun the evidence. GDF11 sits firmly in the middle category — interesting, biologically plausible in at least some contexts, methodologically contested, not ready for clinical translation as a standalone rejuvenating factor. GDF15 sits somewhere between middle and well-characterized, more reproducible as a biomarker than GDF11 and with cleaner mechanistic links to specific biology, but with the same distance from actionable therapeutic use.

The broader question the parabiosis work poses is one of the genuinely interesting open questions in aging biology: to what extent is the aging of individual tissues driven by intrinsic cellular programs versus by circulating systemic factors? If old blood actively inhibits tissue regeneration — if there are factors that rise with age and suppress the maintenance programs that younger organisms use — then addressing those factors could theoretically rejuvenate aged tissues even without cell replacement or genetic reprogramming. The dilution of aging inhibitory factors, rather than addition of youth factors, may turn out to be the more tractable version of the young blood hypothesis. Some of the research direction has moved in this direction, asking what to remove from old plasma rather than what to add.

The thread from a parabiosis experiment to a mechanism to a therapeutic is longer than it looks from the outside. What has been established is that blood-borne factors influence aging trajectories in mouse models in reproducible and striking ways. What specific factors drive those effects, in which direction, in which tissues, and across what species has turned out to require a lot more work than the first papers implied. The science is real and ongoing. The translation is early. The distance between those two facts is exactly where it is most important for anyone engaging with this area — as a patient, a clinician, or a consumer — to hold their position carefully.

What the GDF controversy ultimately teaches is that aging biology is operating in a domain where the mechanisms are genuinely complex, the tools are still being validated, and the first story is rarely the whole story. That is not a reason to dismiss the field. It is a reason to read the second and third papers.