What You'll Learn
Table of Contents
Semax And Selank: One Trial, Then Two
Semax is a synthetic fragment related to ACTH, sold for focus, recovery and cognitive protection. Selank is related to tuftsin and sold for anxiety.
The counts:
- Semax: 207 records, 1 randomised trial
- Selank: 68 records, 2
Two hundred and seven papers and a single randomised trial is a ratio worth sitting with. It is the same structure as BPC-157: a large body of mechanistic and animal work, and almost nothing that randomised a person.
These are also genuinely interesting compounds pharmacologically, developed for real clinical purposes, with plausible neurochemistry behind them. That is not in dispute and is not the point.
Where That Literature Comes From
Add a country filter and the picture sharpens.
159 of semax's 207 records are Russian by affiliation or language. For selank it is 55 of 68. Both compounds were developed in Russia, are licensed there, and the great majority of what has been published about them comes from that research tradition.
This is the same observation made about epitalon and one research group, one level up. It is not a claim about the quality of any individual paper, and it is emphatically not a claim about a country's science.
It is a claim about independent replication. The mechanism by which science removes the errors and enthusiasms of any one group is other groups, working in other systems, with other incentives, finding the same thing. A literature that has stayed within one tradition has not had that done to it — not because the tradition is suspect, but because the check has not been run.
The useful question when you see a large paper count is therefore not just how many, but how many independent sources. That question costs one extra search term and it is the one that separates a field from a programme.
Bottom line: Three searches now settle most of this section: filter for randomised trials rather than counting papers, check whether the compound was given or merely measured, and check how concentrated the authorship is — by group, or by country.
DSIP: A 1980s Idea Still On Sale
Delta sleep-inducing peptide has 518 records and 6 randomised trials, which sounds respectable until you notice when they happened. DSIP was isolated in the 1970s and studied heavily through the 1980s, and the research largely stopped.
Research programmes stop for two reasons. Sometimes funding moves on from a good idea. More often, the early promise does not survive contact with better studies, and people quietly go elsewhere.
A compound with forty years of opportunity to accumulate evidence, and six randomised trials mostly from the era of its discovery, is telling you which of those happened.
For sleep specifically, that matters more than usual, because sleep has unusually good non-pharmacological evidence. Cognitive behavioural therapy for insomnia is first-line in clinical guidelines and outperforms most sleep medication over the long term. It is unglamorous and free, which is why nobody markets it.
Noopept And Dihexa: Zero And Zero
Briefly, because there is little to report and that is the report.
- Noopept: 85 records, zero randomised trials.
- Dihexa: 17 records, zero.
Dihexa in particular circulates with claims about being some enormous multiple of BDNF in potency. That figure comes from laboratory work on synaptogenesis. Seventeen papers exist in total, and not one of them randomised a human being.
A number with a large multiplier in it is doing rhetorical work, not clinical work. It is worth asking, every time: a multiple of what, measured in what, in which species.
Cerebrolysin Has 62 Trials, And That Is The Bad News
Now the compound that breaks the pattern, and the reason this article exists.
Cerebrolysin is a peptide preparation derived from pig brain, used clinically in a number of countries for stroke and dementia. It has 633 records and 62 randomised trials. By any standard in this section that is a real evidence base — more randomised trials than BPC-157, TB-500, epitalon, semax, selank, ipamorelin and CJC-1295 combined, several times over.
So the honest question is not whether it has been studied. It is what the studies found when they were pooled.
A systematic review and meta-analysis of animal-derived nootropics in cognitive disorders concluded that although published data suggest potential benefits and relative safety, the supporting evidence is weak, and the size of the effects demonstrated were modest and probably less than would be considered clinically relevant.
Read that as the strongest available answer in this whole category. Where somebody finally ran enough trials to find out, the effect came in below the threshold at which anyone would notice it.
Which is the same result as cagrilintide's 0.16 percentage points, reached by a different route: a real compound, properly studied, producing something too small to matter to the person taking it.
Pooled evidence on animal-derived nootropics in cognitive disorders. Potential benefits and relative safety suggested by published data, but the supporting evidence weak and the effect sizes modest and probably below clinical relevance.
PMID: 36324709 ↗The Complete Pattern, Across Six Articles
This closes the section, so it is worth laying the whole thing out. Every peptide on the market fails in one of five ways, and knowing which one saves you the argument:
- No trials at all. BPC-157, TB-500, epitalon, noopept, dihexa.
- Trials that answer the wrong question. The growth hormone secretagogues — 42 trials establishing that growth hormone rises, and a direct test showing that raising it adds 2.1 kg of lean mass and no strength.
- Trials showing an effect too small to matter. Cagrilintide at 0.16 percentage points, and cerebrolysin below clinical relevance.
- Excellent trials, and you cannot buy it. Retatrutide, and elamipretide for rare disease.
- You can buy it, and the evidence is thin. Copper peptide, tested in 13 people with no objective effect.
Six articles, several hundred papers read or counted, and not one compound that clears all four bars: real trials, meaningful effect, lawfully obtainable, and independently replicated.
That may change. Retatrutide is in phase 3 and cagrilintide is a serious programme, so this section will need rewriting eventually, and that is how it should be. Until then, the things with large replicated evidence for the outcomes people are buying peptides for have not moved: training, protein, sleep, and sunscreen.
Cheap, dull, and the only things on the list that survive their own evidence.
