Do Pinealon and Epitalon improve REM sleep, and what did Huberman really claim?

The short answer

Huberman did say Pinealon doubled his REM sleep, and that claim is being repeated without the three caveats he attached to it. He noted Pinealon has very little human data, said he had stopped taking it, and was explicit it was personal use rather than a recommendation. The measurement came from a consumer sleep tracker.

Evidence tier: Tier 4 for Pinealon and Epitalon sleep effects — no completed human efficacy trials, evidence confined to one research lineage. Tier 2 for the limitations of consumer sleep-tracker REM staging. Educational content, not medical advice.

The key points:

• The Huberman quote is real — but stripped of every qualification he gave it.
• It is n=1, tracker-measured — and consumer wearables are poor at scoring REM specifically.
• He stopped taking it, which almost never survives the retelling.
• Neither peptide has completed human efficacy trials, and nearly all evidence traces to one lab.
• The FDA classified Epitalon as a Category 2 bulk drug substance — flagged for safety concerns or lack of adequate evidence.

What did Huberman actually say?

Evidence tier: 4 — single self-report.

On a Huberman Lab episode with peptide physician Dr Craig Koniver, Andrew Huberman described taking Pinealon over roughly four to six months and seeing his REM sleep rise from about 1–1.5 hours to nearly 3 hours per night, tracked with a wearable. He said something close to: never before had he found something that improved the amount of REM sleep he got.

That much is accurately reported. Here is what gets left out.

He said Pinealon has very little human data. He raised the limitation himself, unprompted.

He said he had stopped taking it. A detail that almost never survives into the viral version, and an odd omission — if something doubled your REM sleep, stopping is informative.

He was explicit that this was personal use, not a protocol he was recommending. The distinction he drew is the entire difference between a data point and advice.

And his own observation undercuts the mechanism. He noted his sleep stayed improved on nights he did not take Pinealon. He and Koniver speculated this might reflect regenerated pinealocyte function — but no published research demonstrates that for Pinealon, and the same observation is equally consistent with expectancy effects, regression to the mean, seasonal change, or any of the other things that shift sleep over a four-to-six-month window.

A recent post framing these compounds as "the closest thing to an Ozempic for sleep" reached over 74,000 views on the strength of that quote. The comparison is worth pausing on: semaglutide's effects rest on multiple large randomised trials with tens of thousands of participants. Pinealon's rest on one man's wearable.

Can a sleep tracker actually measure REM?

Evidence tier: 2 — established validation literature.

This is the part nobody addresses, and it undermines the quantitative claim directly.

Consumer wearables infer sleep stages from movement, heart rate and heart-rate variability. Against polysomnography — the clinical standard, which measures brain activity via EEG — they perform reasonably at distinguishing sleep from wake, and considerably worse at staging. REM is among the stages they get wrong most often, because the physiological signatures they rely on overlap heavily between REM and light sleep.

So "my REM doubled from 1.5 to 3 hours" is a statement about what an algorithm estimated, not about measured brain activity. A change of that size in tracker-derived REM is well within the range that device firmware updates, a changed sleeping position, or a shifted bedtime can produce.

None of this means nothing happened. It means the instrument cannot support the precision of the claim. If you want to know whether a compound changes your REM sleep, the honest answer is that you would need polysomnography, or at minimum a pre-registered n=1 design with washout periods — which is a reasonable thing to run, and covered in our guide to designing a cognitive self-experiment.

What is the actual evidence for Pinealon?

Evidence tier: 4 — preclinical only, single research lineage.

Pinealon is a tripeptide (Glu-Asp-Arg, "EDR") from Vladimir Khavinson's group at the St Petersburg Institute of Bioregulation and Gerontology, part of the family marketed as "bioregulators." The proposed mechanism is that short peptides enter cells and nuclei and modulate gene expression directly.

The preclinical work is real but thin. A 2011 study of cultured neuronal cells under oxidative stress reported reduced free-radical levels, improved viability and greater proliferative activity versus controls. There is additional animal work on neuroprotection.

There are no completed human efficacy trials for Pinealon, and none at all for sleep. The wider problem is structural: the evidence base clusters almost entirely around a single research network, largely in Russian-language publications, with minimal independent Western replication. When a compound's entire literature traces to one institute and its collaborators, the appropriate reading is hypothesis-generating rather than confirmatory — not because the researchers are assumed to be wrong, but because independent replication is what converts a finding into knowledge, and it has not happened here.

There is also no published research proposing a mechanism by which Pinealon would specifically increase REM sleep. The sleep claim is downstream of one person's tracker, not of a mechanistic hypothesis anyone has tested.

What about Epitalon?

Evidence tier: 3–4 — animal and cell-culture data, no controlled human efficacy trials.

Epitalon (epithalon, AEDG — Ala-Glu-Asp-Gly) is the better-known compound from the same lab, a synthetic simplification of the bovine pineal extract epithalamin. Its reputation rests on telomerase.

What exists: telomerase activation and telomere elongation in human cell cultures, including an independent replication at Brunel University London in 2025 — which matters, because it is one of the few findings reproduced outside the originating network. A 2001 mouse study from Anisimov, Khavinson and colleagues reported slower ageing and extended lifespan in female CBA mice.

What does not exist: any peer-reviewed, independently replicated controlled trial in healthy humans showing that systemic Epitalon meaningfully lengthens leucocyte telomeres. Human data is limited to observational work and cohort studies largely using the parent extract. The mouse lifespan work has never been independently replicated outside Khavinson's lab.

And the regulatory position is unfavourable. The FDA classified Epitalon as a Category 2 bulk drug substance — the designation for substances presenting demonstrable safety concerns or lacking adequate evidence of clinical utility for compounding.

That last point sits awkwardly beside recent developments. Epitalon was among the compounds an advisory committee recommended for the 503A bulks list in 2026, on a split vote. A Category 2 classification and a positive committee recommendation are not straightforwardly compatible, and anyone following this compound should watch how that resolves rather than assume the favourable reading.

Neither compound has any published evidence connecting it to sleep architecture.

Limitations

This is educational content, not medical advice.

• The central claim is a single person's self-report, measured with an instrument not validated for REM staging.
• Neither Pinealon nor Epitalon has completed human efficacy trials. Neither has any published sleep-outcome data.
• The evidence base is concentrated in one research lineage, largely in Russian-language literature, with limited independent replication.
• The Brunel cell-culture replication is meaningful but narrow — telomerase activation in vitro is a long way from a clinical effect.
• "Thousands of people say" is not an evidence claim. Self-selected reports from people who bought a compound and expected it to work are the least reliable data available on sleep, which is unusually responsive to expectation.
• Epitalon's FDA Category 2 status reflects an unfavourable regulatory assessment and should not be dismissed.
• Long-term safety data is absent for both compounds.
• Marko Maal, MSc Pharmacy reviewed this article. Reviewer attribution does not constitute a doctor-patient relationship.

The bottom line

The Huberman quote is real, and that is exactly why the way it travels is a problem. He reported a large personal change in tracker-estimated REM, and in the same conversation flagged that the compound has very little human data, that he had stopped using it, and that he was not recommending it. The viral version keeps the number and discards all three qualifications — which converts a carefully hedged anecdote into something it was never intended to be.

Underneath, both compounds sit in the same position: interesting preclinical mechanisms, essentially no human efficacy data, and an evidence base that has not been independently replicated in any meaningful way. Epitalon has slightly more behind it, including one 2025 cell-culture replication outside the originating lab, and also carries an unfavourable FDA classification. Neither has published sleep data of any kind.

If you sleep badly, that is worth taking seriously — and the interventions with actual randomised evidence behind them are unglamorous and well established. If you want to try these anyway, the honest framing is that you are running an uncontrolled experiment on yourself with a compound nobody has properly studied, and that your wearable is not equipped to tell you whether it worked.

Related on this site

• Epitalon
• Why most longevity peptide stacks don't survive evidence scrutiny
• Peptides for longevity (2026): what's backed by evidence
• How to design a cognitive self-experiment with metrics that aren't BS
• Are you supposed to feel Semax & Selank? What nootropic peptides feel like
• Our evidence-tier framework

References

• Huberman Lab podcast, episode with Dr Craig Koniver (2024) — the original Pinealon REM self-report, including Huberman's own caveats about limited human data and having discontinued use.
• Khavinson VK, et al. Short peptide bioregulators — Pinealon (EDR) neuroprotection and oxidative-stress work, St Petersburg Institute of Bioregulation and Gerontology. Preclinical; predominantly single-lineage.
• Anisimov VN, Khavinson VK, et al. Epithalon and lifespan in female CBA mice (2001) — not independently replicated outside the originating laboratory.
• Independent replication of Epitalon telomerase activation in human cell culture, Brunel University London, 2025.
• FDA bulk drug substances nominations — Epitalon classified Category 2 (demonstrable safety concerns or insufficient evidence of clinical utility for compounding).