Dihexa vs PE-22-28: Which Is Better for Memory?
Should I take dihexa or PE-22-28?
Neither, yet. Both have zero human trials, zero human pharmacokinetics and research-only access. PE-22-28 scores 4.9 and dihexa 4.4, and the gap is entirely safety: dihexa potentiates the c-Met pathway that approved cancer drugs exist to block. If you take one anyway, they treat different problems, memory against mood.
- No human trial exists for either compound. Not a failed trial, not an ongoing one. Zero, on both sides, along with zero human pharmacokinetics.
- Dihexa wins memory 5.5 to 1.7, cognition 5.0 to 1.8 and neuroplasticity 5.0 to 2.3. PE-22-28 wins depression 2.5 to 1.0, mood 2.5 to 1.0 and anxiety 2.0 to 1.0. Different problems, not competing answers.
- PE-22-28 scores higher overall despite losing the bigger rows, because dihexa's downside total is 2.276 against 1.606 and its safety subscore is the worst number on this page.
- Dihexa's mechanism is the problem. It potentiates HGF signaling at c-Met, and capmatinib and tepotinib are FDA-approved cancer drugs that inhibit that same receptor.
- FDA states it has not identified human exposure data for dihexa acetate and lacks important safety information. The compounding nomination was withdrawn.
- PE-22-28 has a dosing trap rather than a cancer question. Djillani 2019 showed a biphasic curve where a low dose activates TREK-1 and a higher dose blocks it, and nobody knows where the human line sits.
- Costs are $30 to $60 a month for dihexa and $45 to $138 per vial for PE-22-28, both before the per-lot testing that is mandatory on research-only material.
At a Glance
Dihexa
- Efficacy 3.0
- Breadth 2.0
- Evidence 2.3
- Speed 3.5
- Durability 2.5
- Bioindividuality 2.5
- Safety Risk 3.3
- Side Effects 2.2
- Cost 2.9
- Effort 2.1
- Opportunity Cost 2.5
- Dependency 1.8
- Reversibility 2.9
- Memory
- Cognition Focus
- Neuroplasticity
- Neuroprotection
Angiotensin IV analog acting through HGF and c-Met. BioHarmony 4.4, caution.
PE-22-28
- Efficacy 3.0
- Breadth 2.5
- Evidence 1.8
- Speed 3.5
- Durability 2.2
- Bioindividuality 2.5
- Safety Risk 2.3
- Side Effects 2.2
- Cost 3.0
- Effort 2.8
- Opportunity Cost 3.0
- Dependency 1.8
- Reversibility 2.0
- Depression
- Mood
- Neuroprotection
- Neuroplasticity
Shortened spadin analogue blocking TREK-1. BioHarmony 4.9, neutral.
Head-to-Head Verdict
| Use Case | Winner | Rationale |
|---|---|---|
| Memory | Dihexa | Dihexa 5.5 against PE-22-28's 1.7, the widest gap on the page and still a rodent result. Benoist 2011 showed hippocampal synaptogenesis and spatial memory improvement in rodents, and Sun 2021 rescued cognitive impairment in an APP/PS1 mouse model through PI3K and AKT signaling. Read Wells 2024 alongside them: an angiotensin IV analog failed to protect 40 rats. |
| Cognition Focus | Dihexa | Dihexa 5.0 against 1.8. Its whole reason to exist is procognitive: McCoy 2013 evaluated metabolically stabilized angiotensin IV analogs as procognitive and antidementia agents in rats and cells. PE-22-28 is an antidepressant-class compound and scores 1.8 here because nothing has tested it for focus. This is the row that decides the choice for most people asking. |
| Neuroplasticity | Dihexa | Dihexa 5.0 against PE-22-28's 2.3, and both sides have real mechanistic claims here. Dihexa is claimed to drive dendritic spine formation through HGF and c-Met potentiation. PE-22-28's spadin parent raised PSD-95 and synapsin, increased mature spines and raised BDNF in mice (Devader 2015). Dihexa wins on the strength of the claim, not on evidence quality, because neither has human data. |
| Depression | PE-22-28 | PE-22-28 2.5 against dihexa's 1.0, and this is the compound's actual target. Mazella 2010 discovered spadin blocking TREK-1 and producing antidepressant responses across five tests. Qi 2018 found TREK-1 blockers reversed depressive-like behavior in chronically stressed rats at least a week earlier than fluoxetine. Clean mechanism, absent human evidence. |
| Mood | PE-22-28 | PE-22-28 2.5 against 1.0, on the same serotonergic mechanism. Ye 2015 showed TREK-1 blockade substantially increased serotonin neuron firing in the dorsal raphe and synergized with 5-HT1A signaling. Dihexa has no monoaminergic action at all. Both numbers are low, which is the honest report of two compounds with no clinical data. |
| Anxiety | PE-22-28 | PE-22-28 2.0 against dihexa's 1.0. Both scores sit near the floor and neither has an anxiety trial. PE-22-28 gets the edge because a serotonergic mechanism plausibly touches anxiety, and because Wang 2021 showed hippocampal TREK-1 is upregulated by chronic stress in mice. Plausibility, not proof. |
| Neuroprotection | Dihexa | Dihexa 4.5 against PE-22-28's 2.5, and PE-22-28's case is more interesting than its score. Djillani 2019 found a biphasic dose response where a low dose activates TREK-1 for neuroprotection and improved stroke recovery while a higher dose blocks it for the antidepressant effect. Dihexa wins the row, but PE-22-28's data here is the one place a specific human indication is visible. |
| Nerve Regeneration | Dihexa | Dihexa 3.5 against 1.8, on a single combination study. Weiss 2021 tested stem cells, G-CSF and dihexa together in a rat sciatic nerve damage-repair model, so nothing there isolates dihexa's contribution. That is the entire basis for the row, and it is why the subrating is 3.5 rather than something higher. |
| Stress Resilience | PE-22-28 | PE-22-28 2.0 against dihexa's 1.0, at the floor on both. Wang 2021 showed chronic stress upregulates hippocampal TREK-1 and that inhibiting it altered depression-related behavior and rescued synaptic proteins in mice. That is a stress-relevant mechanism. It is also a mouse. |
| Sleep Quality | Tie | PE-22-28 1.6 against dihexa's 1.0, both at the floor, and neither has a sleep mechanism or a sleep study. This row exists to close the question rather than answer it. If sleep is what you are trying to fix, nothing on this page is the tool. |
| Healthspan | Tie | PE-22-28 1.4 against dihexa's 1.0. Two research chemicals with no human exposure data cannot have a healthspan case, and both scores reflect that absence rather than a measured null. The FDA's own page on dihexa states it has not identified human exposure data and lacks important safety information. |
| Longevity | Tie | PE-22-28 1.4 against 1.0, floor scores on both. No lifespan work exists for either compound in any species. Dihexa's contraindication list names indefinite continuous use with no defined endpoint as a contraindication in itself, which is the opposite of a longevity protocol. |
Cost Comparison
| Intervention | Monthly Cost | Notes |
|---|---|---|
| Dihexa | $30 to $60 | ESTIMATE, priced 2026-09-07, gray-market research-chemical channel. Community sublingual doses of 8 to 12 mg/day run $30 to $60 a month from gray-market vendors. That figure is before the per-lot HPLC and mass spectrometry identity testing the report requires, which can cost more than the peptide itself. Budget the testing as the larger line item, not the smaller one. |
| PE-22-28 | $45 to $138 per vial | ESTIMATE, priced 2026-09-08, research-chemical channel. No validated human dose exists, so this prices one vial per short self-directed course of about a month: an 8 mg vial is listed around $55 by one vendor and another lists PE-22-28 from $44.75 to $138.42 across sizes. The wide range reflects that nobody knows what a course is, not that the product varies that much. |
| The difference | The real cost on both is the testing, not the peptide | PE-22-28 costs more per vial and dihexa is the more expensive commitment, because its assessment window is four weeks against PE-22-28's two and its testing requirement is per lot rather than per purchase.Both prices omit the same thing. These are research-only compounds with purity verification required, and PE-22-28's report puts it bluntly: there is no pharmaceutical-grade option to pay up for, so quality cannot be bought. You are paying for a vial and separately paying to find out what is in it. |
When to Switch
Their evaluation windows are short and mean almost nothing, so treat both numbers with suspicion. PE-22-28 has a two-week window with first noticeable effect and full effect both at about four days, taken from the rodent antidepressant onset in Djillani 2017.
Dihexa has a four-week window with a first change estimated at two days, and that estimate has no human basis at all. Neither figure comes from a person.
Switching between them is not really the decision. They treat different problems: dihexa is a memory and synaptogenesis compound scoring 5.5 and 5.0 on those rows, PE-22-28 is a fast-onset antidepressant-class compound scoring 2.5 on depression and mood.
If you started dihexa for low mood, you picked the wrong one and no amount of waiting fixes that, and the same holds in reverse for anyone taking PE-22-28 for focus.
The switch that matters is off both, and the reasons differ. Stop dihexa if any cancer history exists in you or your family, because it potentiates HGF signaling at c-Met, and capmatinib and tepotinib are approved cancer drugs that block that same receptor in MET-altered lung cancer.
Stop PE-22-28 if you are on an antidepressant without clinician oversight, if you have any history of mania, or if depression requiring evidence-based treatment is what brought you here. Do not stack them: two research chemicals at once means an adverse event cannot be attributed and neither report has human safety data to fall back on.
Who Should Pick What?
Anyone with a personal or family history of cancer
PE-22-28
Dihexa is out, and this is the hardest exclusion on the page. Its mechanism potentiates HGF signaling at c-Met, and Comoglio 2008 frames MET and HGF as a cancer invasion and metastasis pathway. Mathieu 2022 documents the FDA approvals of capmatinib and tepotinib for MET-altered lung cancer. Active malignancy, cancer history and MET-driven tumors are all contraindicated.
Someone whose complaint is low mood or treatment-resistant depression
PE-22-28
Depression 2.5 against 1.0 and mood 2.5 against 1.0. The mechanism is the cleanest thing on this page: Mazella 2010 linked TREK-1 blockade to antidepressant responses across five behavioral tests, and Qi 2018 beat fluoxetine's timeline by a week in stressed rats. Read the contraindication first though: active depression requiring evidence-based treatment is on its own exclusion list.
Someone whose complaint is memory or cognitive decline
Dihexa
Memory 5.5 against 1.7 and cognition 5.0 against 1.8. Sun 2021 rescued cognitive impairment in an APP/PS1 mouse model and Benoist 2011 showed hippocampal synaptogenesis and spatial memory gains. Then read Wells 2024, where an angiotensin IV analog failed to protect 40 rats against cognitive deficits, and the c-Met exclusion above, which rules most older adults out.
Anyone with bipolar disorder or a history of mania
Dihexa
PE-22-28 is out. Bipolar disorder and any history of mania are on its contraindication list, which is the standard exclusion for an antidepressant-class agent with no clinician managing the dose. Dihexa is not a recommendation here so much as the one that is not specifically contraindicated, and its own exclusions still apply.
Person with diabetes or blood-sugar instability
Dihexa
PE-22-28 has a documented metabolic off-target. Hivelin 2016 showed the TREK-1 blocker spadin potentiates calcium influx and insulin secretion in mouse pancreatic beta cells, which makes it an unintended insulin secretagogue. Nobody has measured that in a person. If your glucose control is already unstable, this is not the compound to experiment with.
Anyone who cannot pay for per-lot identity and purity testing
Tie
Neither. Both are research-only with purity verification required, and dihexa's contraindication list names use without per-lot third-party identity and purity testing as a contraindication in its own right. On dihexa that testing can cost more than the peptide. On PE-22-28 there is no pharmaceutical-grade alternative to buy instead.
Competing athlete in a tested sport
PE-22-28
Dihexa is the clearer anti-doping problem. WADA's 2026 list covers growth factors and growth-factor modulators including HGF and related regenerative capacity modulators, which is exactly what dihexa is designed to be. Verify current status yourself before either, because an unlisted research chemical is not the same as a permitted one.
Anyone deciding between these two for general brain health
Tie
Neither, and this is the honest default. Dihexa is 4.4 and rated caution. PE-22-28 is 4.9 and rated neutral. Both have zero human trials and zero human pharmacokinetics, and both source reports carry low confidence. General brain health is the one goal with no urgency behind it, which makes it the worst reason to be first in line.
Research Highlights
-
Mechanism Difference
These two peptides act on unrelated systems and it shows in their scores. Dihexa is an angiotensin IV analog that potentiates hepatocyte growth factor signaling at the c-Met receptor and is claimed to drive dendritic spine formation, which is why it scores 5.5 on memory and 5.0 on cognition and neuroplasticity.PE-22-28 is a shortened spadin analogue that blocks the TREK-1 potassium channel, raising dorsal raphe serotonin neuron firing. Djillani 2017 measured it inhibiting human TREK-1 at about 0.12 nanomolar against 40 to 60 for spadin itself. That is an antidepressant mechanism, which is why it wins depression and mood and loses everything cognitive.
-
Safety Comparison
Dihexa carries the more serious open question. Its own mechanism is a pathway oncology drugs exist to block: Comoglio 2008 frames MET and HGF as a cancer invasion and metastasis target, and Mathieu 2022 documents FDA approvals of capmatinib and tepotinib as MET inhibitors for MET-altered lung cancer.FDA states it has not identified human exposure data for dihexa acetate. Its safety subscore of 3.3 is the worst number on this page.PE-22-28's risks are a dosing trap and a metabolic off-target. Djillani 2019 found a biphasic curve where low doses activate TREK-1 and higher doses block it, so the direction of effect depends on a dose nobody has established in humans, and Hivelin 2016 showed spadin potentiates insulin secretion in mouse beta cells.
-
Cost Comparison
Dihexa runs about $30 to $60 a month at community sublingual doses of 8 to 12 mg/day. PE-22-28 runs about $45 to $138 per vial, priced as one vial per short self-directed course because no validated human dose exists. Both are September 2026 estimates from gray-market vendors.Neither figure includes the testing, and on these two compounds the testing is the real expense. Dihexa's report notes per-lot HPLC and mass spectrometry can cost more than the peptide. PE-22-28's report states there is no pharmaceutical-grade option to pay up for, so quality cannot be bought at any price.
-
Editorial Verdict
PE-22-28 scores 4.9 and dihexa 4.4, and neither number should be read as an endorsement. Both have zero human trials, zero human pharmacokinetics and research-only access, and both source reports carry low confidence. The honest verdict is neither, yet.If one is taken anyway, they answer different questions. Dihexa is the memory and synaptogenesis compound at 5.5 and 5.0, ruled out entirely by any cancer history because of its c-Met mechanism. PE-22-28 is the fast-onset antidepressant-class compound at 2.5 on depression and mood, ruled out by bipolar history, existing antidepressant therapy or unstable blood sugar.
-
Durability
Durability runs 2.5 for dihexa and 2.2 for PE-22-28, both inferred from rodent work rather than measured in anyone. No study on either compound has followed a subject past cessation, in any species, so these numbers describe an expectation rather than a finding.Their stated onsets are equally soft. PE-22-28's four-day figure comes from the rodent antidepressant timeline in Djillani 2017, which also extended the compound's action to about 23 hours in mice compared with under seven for native spadin. Dihexa's two-day first-noticeable figure is marked estimated and has no human basis at all.
Frequently Asked Questions
- Which is better, dihexa or PE-22-28?
- Neither is good, and they are not competing for the same job. PE-22-28 scores 4.9 and dihexa 4.4. Dihexa wins memory 5.5 to 1.7 and cognition 5.0 to 1.8. PE-22-28 wins depression 2.5 to 1.0 and mood 2.5 to 1.0. Both have zero human trials and both source reports carry low confidence.
- Are there any human studies on either peptide?
- No. Not a failed trial, not an ongoing one, none. Neither compound has a human efficacy trial or human pharmacokinetics. The FDA's own compounding page states it has not identified human exposure data for dihexa acetate and lacks important safety information. Every number on this page comes from rats, mice or cells.
- Is dihexa's cancer risk real or theoretical?
- The mechanism is established and the human risk is unquantified, which is the worst combination. Dihexa potentiates HGF signaling at c-Met. Comoglio 2008 frames that pathway as a cancer invasion and metastasis target, and capmatinib and tepotinib are FDA-approved drugs that block the same receptor. All the cancer categories are contraindicated.
- Why does PE-22-28 score higher when dihexa wins more use cases?
- Downside. Dihexa's downside total is 2.276 against PE-22-28's 1.606, and its safety subscore of 3.3 is the worst number on this page, driven by the c-Met question. Dihexa is rated caution and PE-22-28 neutral. Winning cognitive rows on rodent data does not offset an unquantified oncology mechanism.
- What is the right dose for PE-22-28?
- Nobody knows, and getting it wrong may reverse the effect. Djillani 2019 found a biphasic curve on TREK-1 where a low dose activates the channel for neuroprotection and a higher dose blocks it for the antidepressant effect. No human dose has been established. Its own report prices it as one gray-market vial per short self-directed course, which is the honest description of the state of knowledge.
- Can I take them together?
- Do not. Two research chemicals with no human safety data at once means any adverse event cannot be attributed to either, and there is nothing to fall back on when it happens. Their mechanisms do not overlap, so there is no synergy argument. If you take one, take one, and give it a defined endpoint.
- How much do they cost?
- Dihexa runs about $30 to $60 a month at 8 to 12 mg/day sublingual. PE-22-28 runs about $45 to $138 per vial. Both are September 2026 gray-market estimates. Neither includes the per-lot HPLC and mass spectrometry testing, which on dihexa can cost more than the peptide, and PE-22-28 has no pharmaceutical-grade version to buy at any price.
- Who should not take either one?
- Anyone pregnant, breastfeeding or under 18, and anyone who cannot verify identity and purity per lot. Dihexa adds active malignancy, any personal or family cancer history, MET-altered tumors, and indefinite use with no defined endpoint. PE-22-28 adds bipolar disorder or mania history, current antidepressant therapy without clinician oversight, and depression that needs evidence-based treatment.
Evidence Sources
- Preclinical Evaluation of metabolically stabilized angiotensin IV analogs as procognitive and antidementia agents (2013) Rat and cell evidence for dihexa-related synaptogenic and procognitive activity. The foundational dihexa paper.
- Preclinical Facilitation of hippocampal synaptogenesis and spatial memory by C-terminal truncated Nle1-angiotensin IV analogs (2011) Rodent hippocampal synaptogenesis and spatial-memory improvement.
- Preclinical AngIV-analog dihexa rescues cognitive impairment and recovers memory in the APP/PS1 mouse via PI3K/AKT signaling (2021) Independent mouse Alzheimer's-model signal.
- Preclinical Effects of an angiotensin IV analog on 3-nitropropionic-acid-induced Huntington's disease-like symptoms in rats (2024) PNB-0408 (dihexa) did not protect against motor, weight or cognitive deficits in 40 rats. The negative animal result.
- Preclinical Stem cell, G-CSF and dihexa to promote limb function recovery in a rat sciatic nerve damage-repair model (2021) Combination study. Nothing in it isolates dihexa's own contribution to nerve repair.
- Review Drug development of MET inhibitors in oncology (2008) Frames MET and HGF as a cancer invasion and metastasis pathway and an oncology drug target. The basis for dihexa's cancer exclusions.
- Regulatory FDA approval summary: capmatinib and tepotinib for metastatic NSCLC with MET exon 14 skipping alterations (2022) Confirms FDA-approved MET inhibitors exist for MET-altered lung cancer, the same receptor dihexa potentiates.
- Regulatory FDA: certain bulk drug substances for use in compounding that may present significant safety risks (2026) FDA states it has not identified human exposure data for dihexa acetate and lacks important safety information.
- Regulatory FDA GSRS substance record: DIHEXA, UNII 9WYX65A5C2 (2026) Lists DIHEXA, PNB-0408 and ATH-1001, and notes that UNII availability does not imply regulatory review or approval.
- Regulatory WADA 2026 Prohibited List (2026) Growth factors and growth-factor modulators include HGF and related regenerative-capacity modulators.
- Preclinical Spadin, a sortilin-derived peptide, targeting rodent TREK-1 channels: a new concept in the antidepressant drug design (2010) Spadin blocks TREK-1 at about 10 nanomolar, raises serotonin neuron firing and produces antidepressant responses across 5 behavioral tests with increased hippocampal neurogenesis in mice.
- Preclinical Shortened spadin analogs display better TREK-1 inhibition, in vivo stability and antidepressant activity (2017) PE-22-28 inhibits human TREK-1 at about 0.12 nanomolar against 40 to 60 for spadin, reduced immobility in the forced-swim test and induced neurogenesis after 4 days in mice.
- Review Djillani 2019, Pharmacology and Therapeutics: review of the spadin and TREK-1 antidepressant program (2019) Spadin acts within about 4 days against 3 to 4 weeks for fluoxetine, with native spadin stability under 7 hours.
- Preclinical Djillani 2019, Neuropharmacology: mini-spadin biphasic action on TREK-1 (2019) A low dose activates TREK-1 for neuroprotection and a higher dose inhibits it for the antidepressant effect. The dosing trap, with no human dose established.
- Preclinical The specific TREK-1 blocker spadin potentiates calcium influx and insulin secretion in pancreatic beta cells (2016) TREK-1 blockade acts as an insulin secretagogue in mice. A documented metabolic off-target.
- Preclinical Qi 2018, ACS Chemical Neuroscience: TREK-1 blockers versus fluoxetine in chronically stressed rats (2018) Reversed depressive-like behavior at least a week earlier than fluoxetine, with restored hippocampal neurogenesis.
- Preclinical Ye 2015, European Neuropsychopharmacology: TREK-1 blockade and serotonin neuron firing (2015) Antidepressant-like responses in chronically stressed rats with substantially increased dorsal raphe serotonin firing, synergizing with 5-HT1A signaling.
- Preclinical Devader 2015, British Journal of Pharmacology: spadin, apoptosis and synaptic proteins (2015) Activated MAPK and PI3K, protected neurons against apoptosis, raised PSD-95, synapsin, mature spines and BDNF in mice.
- Preclinical Wang 2021, CNS Neuroscience and Therapeutics: hippocampal TREK-1 under chronic stress (2021) Chronic stress upregulated hippocampal TREK-1 in mice, and inhibiting it altered depression-related behavior and rescued synaptic proteins.
- Preclinical Albiston 2001, Journal of Biological Chemistry: the angiotensin IV (AT4) receptor is insulin-regulated aminopeptidase (2001) Establishes the AT4 and IRAP identity that underpins the angiotensin IV analog class dihexa belongs to.
Glossary
Quick reference for the medical and technical terms used in this comparison.
- c-Met MET Receptor Tyrosine Kinase
- The receptor for hepatocyte growth factor. Dihexa potentiates signaling through it, and approved cancer drugs like capmatinib and tepotinib exist to block it. That tension is dihexa's central safety problem.
- HGF Hepatocyte Growth Factor
- The growth factor that activates c-Met, driving tissue repair and, in the wrong context, tumor invasion. WADA's growth-factor modulator category covers compounds acting on it.
- TREK-1 TWIK-Related Potassium Channel 1
- A potassium channel whose blockade raises serotonin neuron firing. PE-22-28 inhibits human TREK-1 at about 0.12 nanomolar, which is the basis of its antidepressant claim.
- Spadin Spadin
- The sortilin-derived peptide that PE-22-28 is a shortened analogue of. Native spadin is unstable, under seven hours, which is why the shortened version was designed.
- AT4 Angiotensin IV Receptor
- Identified by Albiston 2001 as insulin-regulated aminopeptidase. It defines the compound class dihexa belongs to, though dihexa's claimed action runs through c-Met.
- APP/PS1 Amyloid Precursor Protein / Presenilin 1 Mouse
- A standard transgenic mouse model of Alzheimer's disease. Sun 2021 used it for dihexa's strongest cognitive result, which remains a mouse result.
- Biphasic Biphasic Dose Response
- When low and high doses of the same compound produce opposite effects. Djillani 2019 documented this for TREK-1, so PE-22-28's direction of effect depends on a dose nobody has established in humans.
- UNII Unique Ingredient Identifier
- An FDA substance code. Dihexa has one, and the FDA record explicitly notes that having a UNII does not imply regulatory review or approval.
