What did the researchers actually give people?

Cord blood plasma is the liquid left after cells are separated from blood collected from the umbilical cord. It contains a mixture of proteins. One of them is alpha-fetoprotein, or AFP, a protein made in much larger amounts during fetal development than in healthy adults. Researchers can also manufacture a version of AFP and study it on its own.

This difference matters if someone shows you a study about “young blood.” Ask what the study used. Whole plasma, a concentrated fraction of plasma, a manufactured protein and a transplant of living cord blood cells are different preparations. Even two studies of plasma can involve different donors and routes of administration.

Where did the idea about memory come from?

In 2014, researchers connected the circulations of old and young mice and found changes in the older animals’ hippocampus, a brain region involved in memory. In a separate experiment in the same paper, older mice given plasma from young mice did better on particular learning and memory tests. [1]

A second team tested human cord blood plasma in aged mice. It improved performance on memory tasks. Removing a protein called TIMP2 from the cord plasma reduced the memory benefit, giving researchers a specific factor to investigate. The protein singled out in those experiments was TIMP2. The recipients were mice, so the experiment leaves open whether people would benefit. [2]

What happened in the human aging studies?

A small safety study enrolled 20 older adults. Eighteen completed ten weeks of weekly injections into a muscle. Researchers had concentrated cord plasma before giving it. Everyone received the preparation, so there was no untreated group to compare with. Two people developed redness and heat at the first injection site and received an antihistamine. [3]

One blood test called GrimAge fell by an average of 0.82 years after treatment. GrimAge estimates age-related risk from chemical patterns on DNA. Other age clocks in the study showed no significant reversal. The observed change does not tell us whether the participants lived longer, remembered things more easily or became more independent. For a closer look at what these numbers measure, see Epigenetic Clocks. [3]

A separate South Korean registry entry proposed comparing fresh cord blood, frozen cord blood, frozen plasma and placebo in adults aged 55 and older with signs of frailty. It lists an estimated enrollment of 64. As of November 9, 2026, its status was unknown, its last public update was in January 2016 and it had no posted results. We cannot tell from that record how many people received plasma or how their physical function changed. [4]

Another study often mentioned in this conversation tested plasma from adult donors aged 18 to 30 in people with Alzheimer’s disease. Nine participants entered a randomized crossover comparison with saline. Nine more entered an open-label group after the design changed. The main questions were safety and whether people could complete the infusions. One participant stopped because of hives. With only 18 participants and a short treatment period, the study could not establish whether plasma meaningfully improves memory. [5]

Has AFP itself been tested for rheumatoid arthritis?

There is a reason to study AFP’s effect on immune cells. In a laboratory comparison, low concentrations of AFP from liver tumors inhibited the development of human dendritic cells, which help direct an immune response. The same concentrations of normal AFP from cord blood did not have that effect. The source of the protein and the conditions it was tested under matter. [6]

A clinical trial tested a manufactured form of AFP called MM-093. In a blinded pilot, 12 people with active rheumatoid arthritis continued methotrexate. Eight received weekly MM-093 injections and four received placebo. At 12 weeks, the average improvement in a score that includes joint symptoms and inflammation was about 0.9 points with MM-093 and close to zero with placebo. No serious adverse event was observed. The trial was designed to look primarily at tolerability and was too small to settle efficacy. [7]

Later results changed the picture. In a filing with the U.S. Securities and Exchange Commission, the developer reported that subsequent phase 2 trials missed their main endpoints in rheumatoid arthritis, psoriasis and uveitis, an inflammatory eye disease. The separate rheumatoid arthritis registry record lists no posted results. The sponsor’s filing supplies the overall finding, but no detailed comparison from which to estimate a treatment effect. [8] [9]

Why does AFP also come up in liver cancer?

Some tumors produce AFP. In one phase 1 cancer trial, clinicians gave chemotherapy followed by patients’ own T cells engineered to recognize an AFP target. Thirty-nine people enrolled and 21 received cells. One had a complete response and one had a partial response. All 21 experienced a grade 3 or higher adverse event, and six developed cytokine release syndrome, an inflammatory reaction that was severe in one person. With no comparison group, the study could not measure a survival advantage. [10] [11]

Here, AFP helped the researchers identify a target on cancer cells. The treatment was the engineered T cells. A patient considering a cord plasma or AFP injection would be receiving something different from the intervention in this trial.

What should you ask about a proposed treatment?

Ask for the exact preparation and what it is expected to improve. If the goal is better walking or memory, look for a comparison that measured walking or memory in people with the same problem. A change on a blood test is a different outcome. Find out how donations are screened, what was removed or concentrated and how a reaction would be handled.

In the United States, plasma transfusion has established medical uses, including replacing clotting factors in certain patients who are bleeding. FDA says plasma components are not approved for normal aging or memory loss. Its young-donor plasma update identifies risks of transfusion including allergic reactions, infection transmission and breathing complications. The clinical team should explain the specific risks and authorization for the procedure being proposed. [12]

Identify the preparation

  • Is this whole plasma, a concentrate or a manufactured protein?
  • Which route did the human study use?

Read the outcome

  • Did participants feel or function better?
  • How did the comparison group do?

Clarify the follow-up

  • What should prompt a call after treatment?
  • Who will assess a reaction or a change in symptoms?

Questions people ask

Is cord blood plasma a stem cell treatment?

Plasma is the liquid fraction collected after blood cells are separated out. A cord blood cell transplant delivers cells and has its own indications, risks and evidence. Ask which fraction the proposed procedure uses.

Did 64 people improve in the cord plasma frailty trial?

The number 64 was an enrollment estimate on the registry. The record has unknown status and no posted results. It cannot establish how many participants were treated or whether anyone improved.

What did the GrimAge result mean for participants?

Eighteen people completed a small study of injected cord plasma concentrate. One aging-related blood marker changed over ten weeks. Without a comparison group or a demonstrated improvement in daily function or survival, the clinical meaning remains uncertain.

What did the arthritis trials of AFP find?

A small randomized study found an improvement in disease activity with manufactured AFP added to methotrexate. The developer later reported that phase 2 rheumatoid arthritis testing missed its main endpoint. Those studies tested a specific manufactured protein.

The evidence

Sources checked November 9, 2026. Study types are listed with each source. Visit the Science library for more research.

Mouse study1. Young blood reverses age-related impairments in cognitive function and synaptic plasticity in mice. (2014) PubMed 24793238
Mouse study with human plasma2. Human umbilical cord plasma proteins revitalize hippocampal function in aged mice. (2017) PubMed 28424512
Uncontrolled human study3. Umbilical cord plasma concentrate has beneficial effects on DNA methylation GrimAge and human clinical biomarkers. (2022) PubMed 36052758
Trial registry4. Clinical Trial to Evaluate the Potential Efficacy and Safety of Human Umbilical Cord Blood and Plasma NCT02418013
Human feasibility study5. Safety, Tolerability, and Feasibility of Young Plasma Infusion in the Plasma for Alzheimer Symptom Amelioration Study: A Randomized Clinical Trial. (2019) PubMed 30383097
Laboratory study6. Tumor-derived α-fetoprotein impairs the differentiation and T cell stimulatory activity of human dendritic cells. (2014) PubMed 25355916
Randomized pilot7. A randomised, double-blind, placebo-controlled trial of a recombinant version of human alpha-fetoprotein (MM-093) in patients with active rheumatoid arthritis. (2007) PubMed 17114190
Sponsor report8. Merrimack Pharmaceuticals, Inc., Form S-1, GTC Biotherapeutics agreement and MM-093 clinical outcomes Sponsor filing with SEC
Trial registry9. A Phase 2 Study to Evaluate the Efficacy and Safety of 60mg of MM-093 in Patients With Active Rheumatoid Arthritis NCT00458146
Single-arm phase 1 trial10. Phase I trial of ADP-A2AFP TCR T-cell therapy in patients with advanced hepatocellular or gastric hepatoid carcinoma. (2026) PubMed 40812667
Registry results11. AFPᶜ³³²T in Advanced HCC NCT03132792
Regulatory notice12. Update to Important Information about Young Donor Plasma Infusions Offered for Profit FDA update

This article provides general education. Any treatment decision should be made with a qualified physician. Individual results vary.