Follistatin 344 research links protein to liver and ovary changes

The first study, published in 2026, used a standard model of fatty liver disease ¹. Researchers put mice on a high-fat diet or normal food for 8, 20, or 30 weeks. They then measured follistatin and related proteins in the liver and in the blood. The design allowed them to track changes over time rather than at one moment.

Fatty liver disease, often called MASLD, builds up when fat collects in liver cells. The study wanted to see how the activin-follistatin system, a set of signals that follistatin normally dampens, behaves during that buildup. Activin is a protein that promotes inflammation and scarring in tissues. Follistatin is its natural brake.

The main finding was a mismatch between liver and blood. In the liver, the gene that makes follistatin rose early, but the actual follistatin protein fell at 20 and 30 weeks ¹. In blood, follistatin levels stayed unchanged the whole time ¹. So the liver was making less active brake, but that did not show up in circulation.

At the same time, activin A protein peaked at 20 weeks in both liver and blood ¹. The balance between activin and follistatin shifted toward activin, meaning more of the pro-inflammatory signal relative to its brake ¹. The study also found that the activin receptor ACVR2B stayed elevated, which the authors suggest may favor activin signaling ¹. Another marker, the stress-response protein growth differentiation factor 15 (GDF15), rose in blood and correlated with activin A ¹.

The limitation is that this was a mouse study, not human data. Mice on a high-fat diet develop liver changes that resemble human fatty liver disease, but the timing and severity can differ.

The second study was a human cross-sectional design ². It included 55 women with PCOS, 50 women with idiopathic hirsutism, which is excess hair growth without a clear hormonal cause, and 50 healthy women matched by age and body mass. Researchers measured follistatin, activins, inhibins, and anti-Müllerian hormone, or AMH, in blood.

The finding was that follistatin was significantly higher in the PCOS group than in controls ². Activin A was lower in PCOS than controls, while inhibin A and inhibin B were higher in PCOS ². The study also found that AMH was higher in PCOS than both other groups, but no significant differences in follistatin were found between PCOS and idiopathic hirsutism ². In multivariate analysis, only AMH remained independently associated with PCOS, which means follistatin did not help distinguish PCOS from non-PCOS once other factors were considered ².

The limitation is that this was a cross-sectional study, so it cannot show cause and effect. It measured markers at one point in time and could not tell whether higher follistatin contributes to PCOS or is a response to it.

These two studies add to a growing list of follistatin signals outside muscle. Follistatin is best known as the protein that blocks myostatin, the natural brake on muscle growth ³. That is why the Follistatin 344 research summary focuses on muscle and body composition. But the same protein also appears in metabolic and reproductive research.

In animal models, follistatin affects fat and glucose handling. Mice missing follistatin and a related protein had more fat and worse glucose control, suggesting follistatin normally restrains fat accumulation ⁴. The new liver study shows a different pattern: in a high-fat diet, follistatin in the liver falls, which may allow activin signaling to rise ¹. This is not a contradiction; it shows the system behaves differently in different tissues and over time.

On the reproductive side, follistatin was first identified in ovarian tissue as a protein that suppresses follicle-stimulating hormone ⁵. The PCOS finding that follistatin is higher in women with PCOS is consistent with earlier work showing follistatin gene variants have been studied in PCOS ⁶. However, the new study found that follistatin alone did not distinguish PCOS from idiopathic hirsutism, so its role may be shared across conditions with excess hair growth ².

Follistatin 344 is often discussed in gray-market stacks with BPC-157 and AOD-9604, but neither combination has been tested in the new studies. The new data do not change what is known about stack safety or effectiveness.

A key lesson from the mouse study is that blood follistatin may not reflect what the liver is doing. The gene for follistatin rose early, but liver protein fell later, and blood follistatin never changed ¹. This supports a general caution already present in the literature: circulating follistatin may not track muscle status well. A meta-analysis of older adults found no reliable difference in blood follistatin between those with and without sarcopenia, the age-related loss of muscle ⁷.

The same caution applies to exercise studies. A single bout of exercise raises blood follistatin about seven-fold ⁸, and resistance training plus essential amino acids raised the follistatin-to-myostatin ratio in older women ⁹. The single exercise-bout rise is acute, while the 12-week trial reports a longer-term change, but neither is necessarily a readout of tissue health. The new liver data reinforce that blood levels can stay flat while tissue levels shift ¹.

This matters for anyone trying to interpret follistatin measurements. A blood test may not tell you what is happening inside a specific organ. The protein acts locally by binding activin and myostatin in tissues, and its circulating form is only part of the story ¹⁰.

The two new studies leave open questions. The mouse study did not test whether giving follistatin back to the liver would slow fatty liver disease, so the functional consequence of the drop is unclear ¹. The PCOS study could not determine whether higher follistatin is a cause or a consequence of the condition ². No human trial has tested Follistatin 344 for liver or reproductive outcomes, and the only documented human harm signal remains a case series linking high-dose use to fluid under the retina ¹¹.

It is also unknown whether the synthetic peptide would behave like the body's own follistatin in these tissues. The 344-amino-acid precursor is processed into shorter forms with different tissue binding ¹². Until clinical trials are done, these new findings remain observational clues about the natural protein, not evidence for using the peptide.