What published research reports on Semax in anxiety and stress models

Published August 26, 2026 · Artemis Labs

Semax — answer capsule: Semax is a synthetic seven-amino-acid peptide (PubChem CID 9811102), and its anxiety and stress research record is almost entirely rodent work done in Russia. The main example is a 2024 study at a Moscow institute using a chronic unpredictable stress model in male rats, which reported that Semax reversed or lessened four stress markers and, in the same paper, reported no effect on rat immobility in the forced swim test. No published trial has used an anxiety or stress measure as its primary endpoint in people, and the only randomized controlled trial PubMed indexes concerned motor neuron disease and was negative on its main endpoint. PubMed indexes no meta-analysis, no systematic review, no double-blind study, and no replication of any Semax finding outside Russia.

Key findings

• In a chronic unpredictable stress model in male rats, Semax and a second peptide “reversed or substantially attenuated CUS-induced anhedonia, BW gain suppression, adrenal hypertrophy and a decrease in the hippocampal levels of BDNF” (PMID 39442746, Kurchatov Institute, Moscow).
• The same paper reports a flat null on its behavioural swim measure: “In the forced swim test, no effects of the CUS procedure or peptides on the duration of rat immobility were detected.”
• The authors hedge their own conclusion, writing that these peptides “may have promising therapeutic potential” for stress-related conditions, and scoping it to male rats.
• The nearest human work measured brain scans, not feelings. A 2020 study at a Moscow research centre found resting-state connectivity differences involving the right amygdala in 52 healthy participants, in an analysis its own authors label “Post hoc” (PMID 32342318).

What does the rodent stress research actually report?

The anchor study is a 2024 paper from the National Research Center “Kurchatov Institute” in Moscow, published in the European Journal of Pharmacology. It used a chronic unpredictable stress procedure in male Sprague-Dawley rats, a method in which animals meet varied mild stressors over weeks.

Four measures moved. The stress procedure produced anhedonia, meaning the rats lost interest in something they normally prefer. It held down their body-weight gain. It enlarged their adrenal glands, a physical marker of prolonged stress signalling. And it lowered BDNF, a growth-factor protein, in the hippocampus, a brain region involved in memory and stress response. The paper reports that chronic treatment with Semax “reversed or substantially attenuated” all four.

Every one of those is a measurement in a rat. Anhedonia in a rodent assay is a preference score, not a reported mood, and adrenal weight is an organ measurement. The authors’ own conclusion says these peptides “may have promising therapeutic potential” and scopes it to male rats.

What did the same study fail to find?

Inside that positive paper sits a clean negative. Quoted exactly: “In the forced swim test, no effects of the CUS procedure or peptides on the duration of rat immobility were detected.”

The forced swim test measures one thing, and the sentence above names it: how long a rat stays motionless. That number did not move for the peptides, and it did not move for the stress procedure either. A stress model that fails to shift its own behavioural readout leaves that readout uninformative here.

Most pages about Semax and stress carry the four positive markers and drop this sentence. A null reported by the authors of a positive paper comes from people with every incentive to find an effect.

What happened in the acute stress gene-expression work?

A separate Russian rat study from 2021 read gene activity in the brain after a single episode of acute restraint stress instead of measuring behaviour. It found that Semax “produced > 1500 DEGs, whereas ACTH(6-9)PGP administration led to <400 DEGs at 4.5 h after ARS” (PMID 34576218). A DEG is a differentially expressed gene, meaning one whose activity differed from the comparison condition.

A large number of changed genes shows that something happened, not that it was helpful. The authors keep their own hedge in place, writing that the antistress action of these peptides “may be associated with a correction of gene expression patterns that are disrupted during ARS.” No behavioural outcome was tied to that pattern.

Is there any human research touching emotion circuits?

Two records come close, and neither is an anxiety study. The first is a 2020 imaging study at the Mental Health Research Center in Moscow, which scanned 52 healthy participants three times, once to set a baseline and twice afterwards, with participants receiving Semax, Selank, or placebo. It reported differences in functional connectivity between the right amygdala, a region involved in processing emotional signals, and parts of the right temporal cortex. The authors call that analysis “Post hoc,” meaning the comparison was identified after looking at the data. No anxiety score or mood rating is reported alongside it, so the study cannot say whether anything a participant experienced changed (PMID 32342318).

The second is the only randomized controlled trial in the PubMed record, published in Russian in a Russian neurology and psychiatry journal in 2007, using an intranasal formulation of the Russian pharmaceutical. Its subject was motor neuron disease, and its primary result was negative: Semax “does not influence either the course of CPD or the dynamics of clinical estimates” (PMID 18379501). A secondary quality-of-life measure did move, which the authors attribute to “the improvement of emotional state and motivation” in those patients. That is a secondary endpoint in a trial that missed its primary one, in 27 patients with a specific neurological disease, in a study the abstract itself calls open-label. It is the closest thing to a human emotional finding in the record, and the weakest kind of positive signal.

Why do rodent behaviour results not transfer cleanly?

Rodent behavioural outcomes for this compound have been unstable across strains. A 2012 mouse study of several compounds found that all of them “clearly improved the exploratory behavior in BALB/c mice only” (PMID 23025044), with C57BL/6 mice not showing it. In separate work across five inbred mouse strains, an effect on audiogenic seizure predisposition appeared “in only one-month-old DBA/2J mice” (PMID 19145359) and not in the other four.

The proposed mechanism carries the same problem. The stress paper above reports a change in hippocampal BDNF, but the dedicated rat study of neurotrophin gene expression calls the pattern “multidirectional”: expression fell in the hippocampus and retina and rose in the frontal cortex, depending on region and timing (PMID 19662538). “Semax raises BDNF” is a selective reading of a paper whose own authors chose a word meaning the opposite of uniform.

What the research does not show

No Semax study in people has ever been run outside Russia. PubMed indexes zero meta-analyses, zero systematic reviews, zero double-blind studies, and no independent replication of any Semax result in either direction. A 2026 review by clinicians in the United States states the position plainly for this class of peptides: “there is a current lack of clinical trials” (PMID 41490200).

For this lane specifically: no published trial has used an anxiety or stress measure as its primary endpoint in humans, and the rodent evidence includes a reported null on the anchor study’s one swim-immobility measure. The Russian group behind much of this work has also written that “it is necessary to study influence of the peptide on the brain cells under normal physiological conditions, including understanding the risks of their use,” reporting that in healthy rats the dominant effect was a decrease in immune-related gene expression (PMID 39418522). Separately, one Russian human study logged “episodes of paroxysmal activity” on EEG in some patients (PMID 10199046). The record here is small, one-country, and unreplicated.

For the compound’s identity, sourcing, and analytical specifications, see Semax. Two related pages cover neighbouring parts of the record: Semax vs Selank and Semax cognitive and attention research.

Frequently asked questions

Has Semax been tested for anxiety in people?

Not as a primary question. The human record indexed in PubMed covers stroke, optic nerve conditions, encephalopathy, peptic ulcer, and brain imaging in healthy volunteers, all of it Russian.

What does the forced swim test actually measure?

How long a rat stays motionless. That is the readout, and it is the number the 2024 stress paper reported as unchanged. It is not a measure of anxiety, and no result from it should be described that way.

Does the rat stress research mean Semax reduces stress in humans?

No. Every finding in that study is a measurement in male rats, and the authors scope it to male rats. Nothing in the published record extends it to people.

Has any of this been replicated outside Russia?

No. No independent replication of any Semax result exists in either direction, and no human Semax study has been run outside Russia. The non-Russian primary research is laboratory work in cells and artificial membranes, plus one rat study in Indonesia.

References

1. Antidepressant-like and antistress effects of the ACTH(4-10) synthetic analogs Semax and Melanotan II on male rats in a model of chronic unpredictable stress. Eur J Pharmacol, 2024. PMID 39442746 · DOI 10.1016/j.ejphar.2024.177068
2. PMID 34576218 (2021) — rat brain gene-expression study following acute restraint stress. DOI 10.3390/ijms221810054
3. [The study of chronic partial denervation and quality of life in patients with motor neuron disease treated with semax]. Zh Nevrol Psikhiatr Im S S Korsakova, 2007. PMID 18379501
4. Functional Connectomic Approach to Studying Selank and Semax Effects. Dokl Biol Sci, 2020. PMID 32342318 · DOI 10.1134/S001249662001007X
5. Comparison of the temporary dynamics of NGF and BDNF gene expression in rat hippocampus, frontal cortex, and retina under Semax action. J Mol Neurosci, 2010. PMID 19662538 · DOI 10.1007/s12031-009-9270-z
6. PMID 23025044 (2012) — mouse study of exploratory behaviour in BALB/c and C57BL/6 strains.
7. PMID 19145359 — audiogenic seizure predisposition across five inbred mouse strains.
8. Changes of Transcriptomic Activity in Rat Brain Cells under the Influence of Synthetic Adrenocorticotropic Hormone-Like Peptides. Biochemistry (Mosc), 2024. PMID 39418522 · DOI 10.1134/S0006297924090104
9. [Use of semax at a follow-up of patients with posthypoxic encephalopathy]. Anesteziol Reanimatol, 1999. PMID 10199046
10. PMID 41490200 (2026) — United States review of neuroactive peptides. DOI 10.5435/JAAOSGlobal-D-25-00236

Methodology: this page draws only on primary-source records retrieved from PubChem and NCBI PubMed and transcribed verbatim for the Artemis Labs Semax citation set, last verified August 26, 2026.

All compounds sold by Artemis Labs are for laboratory research use only. Nothing on this page is medical advice, and no statement has been evaluated by the FDA.