Young Blood Transfusions and Aging: What the Evidence Actually Says

A decade after parabiosis captured the public imagination, the science has moved on — and so has the debate. Here is what the data really shows, from stitched-together mice to the first randomized human trials.

Updated for 2026 · Originally framed by TIME Magazine's November 2016 coverage

The premise that started it all

The idea that young blood could reverse aging did not emerge from a wellness clinic or a venture capitalist's pitch deck. It came from a surgical technique first described in 1864 by French physiologist Paul Bert, fell out of favor for decades, and was resurrected in the modern era by a single, striking experiment.

In 2005, researchers Irina Conboy and Michael Conboy, then at Stanford and later at UC Berkeley, published a study in Nature titled "Rejuvenation of aged progenitor cells by exposure to a youthful systemic environment." They surgically sutured pairs of mice together — a young mouse and an old mouse — so that they shared a single circulatory system. This is called heterochronic parabiosis. The results were remarkable: in the older mice, the Notch signaling pathway was restored in satellite (muscle stem) cells, hepatocyte proliferation increased, and the regenerative capacity of aged tissue improved toward youthful levels. The paper has now been cited more than 3,000 times and is the cornerstone of every "young blood" story that followed.

But the takeaway the Conboys themselves drew was subtle, and that subtlety got lost. "What we showed in 2005 was evidence that aging is reversible and is not set in stone," Irina Conboy later said. "Under no circumstances were we saying that infusions of young blood into elderly is medicine." The media, entrepreneurs, and a small army of longevity enthusiasts heard only the part about young blood.

The 2016 study that the TIME article was actually reporting

The TIME piece by Alexandra Sifferlin, published November 22, 2016, was anchored to a new paper that the Conboy lab published the same day in Nature Communications. The lab had developed a technique to exchange blood between mice without suturing them together, allowing precise control of volume and timing — something parabiosis could never do. Old and young mice were connected and disconnected at will, sharing 50/50 blood mixtures for defined windows.

The findings flipped the dominant narrative:

  • Old mice did show some improvements — in muscle repair and liver fibrosis, consistent with earlier work.
  • Young mice, however, were measurably damaged by exposure to old blood. They developed worsened cell formation in the brain and impaired motor coordination, and these declines appeared rapidly — within roughly 24 hours of exchange.
  • The net effect, in co-author Michael Conboy's words: "A single exchange hurts the young partner more than it helps the old partner."

This pointed toward a different mechanism than the one the public had latched onto. If young blood were a fountain of youth, the old mouse should have gained more than the young mouse lost. Instead, the data suggested that something in old blood is actively harmful, and that removing or neutralizing those harmful factors — not adding youthful ones — was where the action was.

"Our study suggests that young blood by itself will not work as effective medicine. It's more accurate to say that there are inhibitors in old blood that we need to target to reverse aging."

— Irina Conboy, UC Berkeley, November 2016

The commercial gold rush, 2017–2018

By the time the 2016 paper was published, the commercial train had already left the station. Stanford neurologist Tony Wyss-Coray, whose lab had shown that young blood plasma could improve cognition in old mice, co-founded a company called Alkahest to pursue the idea in humans. Their first small human pilot, run with Stanford's Sharon Sha and reported in 2017, gave 18 patients with mild-to-moderate Alzheimer's weekly infusions of plasma from male donors aged 18–30.

The results were tepid at best. Patients showed no improvement on objective cognitive tests administered by medical staff. Caregiver surveys nudged upward — by about 4.5 points on a 30-point functional scale — which the investigators themselves acknowledged could easily be a placebo effect from the attention of weekly clinic visits. One patient developed a rash; another had an unrelated stroke.

Meanwhile, a far more brazen operation had launched. Jesse Karmazin, a recent medical school graduate, founded Ambrosia and began selling 1.5-liter infusions of plasma from donors aged 16–25 for $8,000 a pop. The pitch was essentially cosmetic — "plastic surgery from the inside" — and the "trial" was pay-to-participate, with no control group and no peer-reviewed publication of efficacy data. By late 2018, Ambrosia was opening a brick-and-mortar clinic in New York.

The FDA steps in (February 2019)

On February 19, 2019, the U.S. Food and Drug Administration issued an unusually blunt safety statement warning the public against receiving plasma infusions from young donors as a treatment for aging, dementia, Alzheimer's, Parkinson's, multiple sclerosis, heart disease, PTSD, and other conditions. The agency noted that there was no proven clinical benefit and that plasma transfusion carries real risks — allergic reactions, transfusion-related acute lung injury (TRALI), transfusion-associated circulatory overload (TACO), and infectious disease transmission.

The warning effectively ended the wild-west phase of the industry. Ambrosia halted patient treatments the next day, according to reporting by CNBC and the Los Angeles Times, and the company's commercial operations collapsed. Alkahest, which had been operating under proper regulatory oversight from the start, continued — but quietly shifted strategy away from whole young plasma and toward refined fractions.

The 2020 paradigm shift: it's about dilution, not transfusion

In May 2020, the Conboy lab published what may turn out to be the most important paper in this whole line of research. In the journal Aging, Melod Mehdipour and colleagues reported a deceptively simple experiment: they replaced half of the plasma in old mice not with young blood, but with saline containing 5% albumin — a "neutral" age blood exchange (NBE). No young factors. No donor plasma. Just dilution of the old plasma, with albumin replacing the protein that would otherwise be lost.

The results were striking. A single NBE:

  • Matched or exceeded the rejuvenative effects of heterochronic parabiosis on muscle repair;
  • Reduced liver adiposity and fibrosis;
  • Increased hippocampal neurogenesis;
  • And, in a follow-up GeroScience paper later that year, improved cognition and attenuated neuroinflammation in old mice.

This was the cleanest demonstration yet that the benefit comes from removing harmful factors that accumulate in old blood, not from adding youthful ones. The same paper showed that an equivalent human procedure — therapeutic plasma exchange (TPE), already FDA-approved for autoimmune and neurological disorders — produced functional "blood rejuvenation" in vitro, abrogating the typical inhibitory effect of old serum on progenitor cell proliferation.

The implications were commercial as well as scientific. If dilution worked, you didn't need young donors at all. You didn't need a supply chain of teenage plasma. You needed an apheresis machine, albumin, and a protocol that already existed in every major hospital.

The shift to plasma fractions: Alkahest's GRF6019 and GRF6021

Alkahest, which was acquired by Spanish plasma giant Grifols in 2021 and now operates as its wholly owned subsidiary, pivoted toward proprietary plasma fractions rather than whole young plasma. The lead candidates:

  • GRF6019 — a fraction of approximately 400 plasma proteins (clotting factors and immunoglobulins removed to reduce risk), tested in Alzheimer's disease.
  • GRF6021 — a related fraction tested in Parkinson's disease dementia.

Per the Alzforum therapeutics database, the first GRF6019 Phase 2 trial enrolled 47 patients with mild-to-moderate AD, comparing 100 mL versus 250 mL doses given as five daily infusions in week 1 and again in week 13. The trial had no placebo group. Top-line results announced in August 2019 showed the treatment was safe and well-tolerated, with participants showing "no decline" on the ADAS-Cog11 and MMSE and "negligible decline" on the CDR-SB and ADCS-ADL over six months. Peer-reviewed publication followed (Hannestad et al., 2020).

A second, placebo-controlled Phase 2 trial in 26 patients with advanced dementia (MMSE below 10) was less encouraging. Reported at CTAD 2020 and published in 2021, it found the treatment was safe but showed no difference between GRF6019 and placebo on cognitive or functional endpoints. A proteomic biomarker analysis did show post-treatment changes in proteins involved in complement, coagulation, nervous system development, apoptosis, axon guidance, and synapse function — but no measurable clinical benefit. Development of GRF6019 for dementia appears to have stalled; Alkahest's current public pipeline is now focused on diagnostic biomarker discovery through its Chronos-PD program in Parkinson's, funded by the Michael J. Fox Foundation.

The 2025 Stanford human trial: a real signal, finally

On January 19, 2025, Stanford's Department of Anesthesiology, Perioperative, and Pain Medicine announced the first genuinely encouraging human data in this field. Led by Dr. Martin Angst, the study was a randomized controlled trial of GRF6021 in 38 elderly patients undergoing major joint replacement surgery. Patients received four infusions — one the day before surgery, two on the day of surgery, and one the day after — of either the young plasma protein fraction or placebo.

The primary finding was at the molecular level, not the clinical level, and it is important to be precise about that distinction. The young plasma fraction produced statistically significant anti-inflammatory immune modulation, affecting the JAK-STAT, NF-κB, and MAPK signaling pathways (p<0.001). The effect was replicated at the cellular level in subsets of T cells — the adaptive branch of the immune system. This is the first time a controlled human trial has demonstrated that young plasma components can measurably move age-relevant biology in a human being.

But the Stanford team was careful about overreach. The release itself states: "The practice of administering young plasma to the elderly for potential health benefits is strongly discouraged in the absence of compelling human data." The trial was not powered to detect clinical outcomes like reduced surgical complications or improved recovery. It was a mechanism study, and it succeeded as one. It is registered as ClinicalTrials.gov NCT 03981419.

The 2025 Buck Institute TPE trial: when the data disagrees with itself

May 2025 brought the most hyped — and most internally contradictory — paper of the decade. In Aging Cell, researchers at the Buck Institute for Research on Aging and a company called Circulate Health published the first placebo-controlled clinical trial of therapeutic plasma exchange for biological aging. Forty-four adults over age 50 were enrolled. The headline numbers, as widely reported:

  • TPE combined with intravenous immunoglobulin (IVIG): average reduction in biological age of 2.61 years, measured across 36 validated epigenetic clocks.
  • TPE alone (monthly): average reduction of 1.32 years.

For the longevity press, this was a triumph. But the broader literature tells a more complicated story. A separate 2025 human trial of plasmapheresis (Borsky et al., PubMed 40592961, cited 11 times) reported "no significant epigenetic rejuvenation was observed based on epigenetic clock measurements" and suggested the procedure may even accelerate epigenetic aging depending on the protocol. Epigenetic clocks themselves remain an active and contested field — different clocks give different answers on the same samples, and a one- or two-year shift across 36 clocks is the kind of result that demands replication before it means anything clinically.

The honest summary: TPE appears to do something to age-related biology, the size and durability of that something is genuinely unclear, and the field has not yet produced a clean clinical endpoint — longer healthspan, reduced frailty, fewer age-related hospitalizations — that would justify paying thousands of dollars out of pocket for it.

What about the specific "youthful" proteins?

Behind the broad narrative, a parallel hunt has been running for the specific molecules responsible. The leading candidates, and their current status:

  • GDF11. Identified by Amy Wagers' lab at Harvard as a "youthful" factor in blood that declines with age; early parabiosis work suggested it could rejuvenate heart, muscle, and brain. The story got messy fast — a 2016 study (Freitas-Rodríguez et al., PMC5302888) found that GDF11 administration did not extend lifespan in mice, and a long-running dispute about whether the assays were even measuring GDF11 versus its close relative myostatin (GDF8) dragged on for years. A 2023 review in the Journal of Cardiovascular Aging (Driss et al.) argues that the major controversies have been "systematically addressed" and that recombinant GDF11 supplementation does reverse age-related muscle dysfunction in preclinical models. A company called Elevian was built around GDF11; its clinical progress has been quiet. No human efficacy data exists.
  • TIMP2. A young-blood factor studied extensively by Wyss-Coray's lab at Stanford; Alkahest's fractions were partially designed around it.
  • β2-microglobulin (β2M). A pro-aging factor — elevated in old blood, contributes to age-related decline in neurogenesis and hippocampal-dependent cognition. Supports the "old blood is bad" hypothesis.
  • TGF-β. Chronically elevated with age; inhibits muscle stem cell function and impairs liver regeneration.
  • Eotaxin (CCL11). Elevated in old blood; impairs brain function and cognition.

The pattern across all of these is consistent with the Conboys' 2016 and 2020 conclusions: the molecules most consistently implicated in driving aging are elevated in old blood, not missing from it. That is a critical distinction for anyone deciding what to do with their own body today.

What the situation looks like in 2026

Ten years on from the TIME article, here is the honest state of play:

  1. The mouse parabiosis data is real and robust. Multiple labs, multiple tissues, multiple endpoints. Heterochronic pairing does rejuvenate old tissues and does prematurely age young ones.
  2. The mechanism is dilution-dominant, not addition-dominant. The 2020 NBE paper is the cleanest single piece of evidence in the field, and it shows you can get most of the parabiosis benefit in old mice without any young blood at all.
  3. Whole young plasma transfusions for anti-aging are a dead end clinically and a regulatory non-starter. The FDA explicitly warned against them in 2019, and Ambrosia — the only company that tried to sell them at scale — was shut down by that warning.
  4. Refined plasma fractions (Alkahest's GRF series) have shown molecular signals in humans but no convincing clinical benefit. The January 2025 Stanford trial is the best human signal yet, and it is a mechanism study in 38 surgical patients, not an efficacy trial.
  5. Therapeutic plasma exchange is the most plausible near-term intervention — it is FDA-approved for other indications, requires no young donors, is mechanistically aligned with the dominant hypothesis, and has produced one positive (if contested) human aging trial. It is also being aggressively commercialized by longevity clinics at price points that bear no relationship to demonstrated benefit.
  6. The "vampire" framing — Peter Thiel, Ambrosia, teenage plasma for Silicon Valley billionaires — was always mostly marketing. The serious science has moved away from it.

Bottom line: if you wanted to act on this today, what can we do

Stripped of hype, here is the honest, evidence-graded menu for an interested adult in 2026:

  1. Do not buy young blood plasma transfusions. The FDA told you not to in 2019, the human efficacy data is absent, the risks are real (TRALI, TACO, infection, immune reaction), and the dominant mechanistic hypothesis says the active ingredient was never the young blood to begin with. Anyone selling you teenage plasma for "rejuvenation" in 2026 is operating outside the mainstream of both regulation and current science.
  2. Be skeptical of TPE at a longevity clinic. Therapeutic plasma exchange is a real medical procedure with real indications — myasthenia gravis, Guillain-Barré, certain autoimmune nephropathies, hyperviscosity syndromes. For pure "anti-aging" it has exactly one positive small trial (Buck Institute, 2025) and one negative trial (Borsky et al., 2025). Clinics charging $5,000–$10,000 per series for it are pricing ahead of the evidence. If you have a legitimate autoimmune or neurological indication, your insurance may cover TPE through a real hospital — that is the proper route.
  3. Track the GRF6021 / Alkahest pipeline. The January 2025 Stanford result is the first time a young-plasma-derived product has moved human age-relevant biology in a controlled trial. Watch for the larger efficacy trials this mechanism study is supposed to enable. They do not exist yet.
  4. Donate blood if you are eligible. The closest thing to a free, safe, FDA-approved "blood renewal" intervention available to an ordinary adult is routine whole-blood donation. It removes some plasma proteins (including ferritin, which accumulates with age and is independently associated with metabolic disease in men), it is rigorously screened, it costs nothing, and it helps someone else. It is not parabiosis. It will not reverse aging. But it is the only blood-modifying intervention with a clean safety record, broad access, and a positive externality.
  5. Do the boring, evidence-backed things. The Conboys themselves, in the 2016 paper, ended with: "A hot fad today may not survive tomorrow, so be conservative on what you try at home in order to live healthier and longer." A decade of subsequent research has confirmed that view. Exercise, sleep, calibrated caloric intake, blood pressure and glucose control, and avoidance of smoking remain — by orders of magnitude — the most powerful age-modifying interventions available to anyone in 2026. Nothing in the young-blood literature has come within sight of their effect sizes.
  6. Watch for the real inflection point. The breakthrough, when it comes, will almost certainly look like a targeted removal or neutralization of specific pro-aging plasma factors (β2M, TGF-β, eotaxin, SASP components from senescent cells), not a transfusion. The 2020 Conboy NBE paper is the conceptual template. Companies working on senolytics (Cleara Biotech, Unity Biotechnology, resTORbio's successors) and on targeted plasma factor depletion are the ones to follow, not the ones selling liters of someone else's youth.

The short version: the mice were telling us something true, and we spent a decade misreading it. The data says old blood is the problem, not that young blood is the cure. Act accordingly.

Sources

  

📝 📜 ⏱️ ⬆️