Brain Peptides and the Problem of Where Evidence Comes From

The compounds sold as nootropic peptides present a different evidence problem from the gut group. There, the difficulty was a thin literature concentrated in one laboratory. Here the literature is genuinely substantial — decades of it, including registered medicines and Cochrane reviews. The difficulty is that much of it cannot easily be appraised from outside the tradition that produced it, and the parts that have been appraised came back weaker than their headlines suggest.

Two different ways evidence can fail to be useful

Evidence can be inadequate because there is not much of it. That is the ordinary case, and it is easy to describe.

It can also be inadequate for reasons that have nothing to do with quantity. A finding published in a language and a registry that reviewers elsewhere do not search is, for practical purposes, unavailable to them. A trial conducted to reporting conventions other than the ones a systematic reviewer expects cannot be assessed for risk of bias, whatever its underlying quality. And a result that has never been attempted by anyone outside the group that produced it remains a single observation however many times that group publishes it.

None of this is an accusation. Research traditions differ, publication infrastructure differs, and a study being hard to appraise from outside is not evidence that it was done badly. But it does mean that the question “how much research is there” and the question “how much can be concluded” come apart sharply in this category.

Semax and Selank: approved somewhere, unappraisable elsewhere

Both compounds came out of the Institute of Molecular Genetics at the Russian Academy of Sciences. Semax is a heptapeptide derived from a fragment of ACTH; Selank is a synthetic analogue of the endogenous peptide tuftsin. They are closely related products of the same programme.

Their regulatory position is unusual and frequently misreported. Semax is a prescription medicine in Russia and appears on the Russian List of Vital and Essential Drugs approved by the government in December 2011. Selank is available in Russian and Ukrainian pharmacies. Neither has been evaluated, approved or marketed in most other countries.

That combination is where the confusion starts. Approval in one jurisdiction is a real fact, and it is routinely presented as though it settled the scientific question generally. It does not. A national regulator's decision reflects the dossier submitted to it and the standards it applies, and those are not automatically transferable — which is why approvals are sought separately in each territory rather than recognised globally.

The underlying literature is predominantly Russian-language and from Russian institutions. For a reader outside that tradition the practical consequences are concrete: the trials are largely absent from the registries normally searched, methods are often not reported in the form international reviewers appraise, and independent replication by unconnected groups is sparse to absent. The honest summary is not that the research is poor. It is that from here it cannot be checked, and an unfalsifiable body of work supports less than a smaller one that has been stress-tested.

Cerebrolysin: what happens when appraisal does take place

Cerebrolysin is the most useful compound in this group precisely because it has been through independent systematic review — twice, by Cochrane, for different indications. The two results together are an education in reading evidence.

For vascular dementia, the 2019 review covered six randomised trials with 597 participants, three from China and one each from Russia and Romania, with one unclear. Its conclusion reads encouragingly: courses of intravenous Cerebrolysin improved cognition and general function, with no suggestion of adverse effects. Quoted on its own, that sentence would sell a great deal of product.

The rest of the review is where the meaning lives. The evidence was graded very low quality under GRADE across every outcome. The included papers carried a high risk of bias, the analyses were limited by heterogeneity, and the authors stated plainly that the data are not definitive, that adequately powered and methodologically robust trials are needed, and that even if the benefits are real the effects may be too small to be clinically meaningful.

For acute ischaemic stroke, the picture reversed. Six trials, 1,501 participants, and a conclusion that the review does not demonstrate clinical benefits — together with moderate-quality evidence that serious adverse events may be more common with Cerebrolysin. The reviewers also recorded that the manufacturer had supported three of the multicentre studies, whether wholly or by supplying drug, placebo, randomisation codes, grants or statisticians, and identified this as a conflict of interest. Selective reporting was judged a high risk of bias in four of the studies and incomplete outcome data in five.

The lesson is not that Cerebrolysin is useless. It is that the same compound produced one mildly encouraging review and one discouraging review with a safety signal, and that in both cases the headline sentence is close to meaningless without the quality grading printed next to it. Anyone can quote the dementia conclusion with perfect accuracy and leave a reader badly misinformed.

DSIP: the compounds that are never disproved

Delta sleep-inducing peptide was isolated in the 1970s by Schoenenberger and Monnier, from the cerebral venous blood of rabbits in induced sleep. It arrived with an evocative name and an obvious hypothesis.

In 2006, a review of the field appeared in the Journal of Neurochemistry under the title “Delta sleep-inducing peptide (DSIP): a still unresolved riddle”. Three decades after discovery, the title was the finding.

This is the most common fate of a candidate compound, and it is widely misread. Most are not refuted; interest simply drains away as results fail to consolidate, and the literature thins to occasional papers. Because nothing was ever formally disproved, the compound retains an indefinite status that reads, to a casual search, as open promise. An absence of recent negative findings usually reflects an absence of recent investigation.

Why cognition is unusually hard to study

There is a reason this category is thick with inconclusive work, and it is methodological rather than commercial.

Cognitive and mood endpoints are soft. The instruments used to measure them vary between trials, which makes results hard to pool. Placebo responses on subjective measures are large and reliable. Repeated cognitive testing produces practice effects that can masquerade as improvement. Trials in this area tend to be short and small, which is the combination most likely to generate a positive result that does not survive replication.

All of which bears on the kind of evidence that circulates about these compounds in practice. Self-reported impressions of clearer thinking, collected without a control group from people who knew what they had taken and expected it to work, sit at the very bottom of the hierarchy. That is not a comment on anyone's honesty. It is that the measurement is incapable of distinguishing the compound from the expectation.

Questions that sort the claims

Six, and they handle most of what circulates.

Can the study actually be located and read, or only cited? Has any group unconnected to the originators reproduced the central finding? Where a systematic review exists, what did it rate the certainty of the evidence — not merely what did it conclude? Who funded and supplied the trials, and did the reviewers flag it? Was the endpoint objective, or self-reported by people who knew what they had received? And is regulatory approval in one country being offered as general scientific validation, which it is not?

Run honestly over this group, those questions produce a consistent answer. There is a great deal of research. Some of it is serious. Very little of it has been independently replicated or graded as better than low certainty, and the one compound subjected to repeated independent review returned a safety signal in the larger indication.

The interesting thing about this category is not that the evidence is thin — it is not thin. It is that volume and provenance are so easily mistaken for each other, and that the distinction is invisible to anyone reading only the conclusions.

References

Documented, batch-tested, research use only

Pioneer publishes a batch-specific certificate of analysis for every compound we supply, matched to the batch code on the vial. We make no claims about what any compound does.

View compounds Batch certificates