There are no peptides proven to treat Alzheimer’s or Parkinson’s disease. The best-tested peptides for Alzheimer’s and Parkinson’s are GLP-1 drugs, and the biggest trials were negative: oral semaglutide did not slow early Alzheimer’s in 3,808 people over 104 weeks,[1] and exenatide did not slow Parkinson’s in a 96-week phase 3 trial.[4] One smaller trial of lixisenatide showed a modest motor effect in early Parkinson’s.[3] Cerebrolysin has small, short positive trials abroad, and Semax, Dihexa and P21 have no meaningful human evidence for either disease.
This guide grades each option by evidence. For the wider topic, see our nootropic peptides hub.
Evidence at a glance
| Peptide or drug | Evidence level | What was shown | FDA status | Our profile |
|---|---|---|---|---|
| Approved Alzheimer’s and Parkinson’s drugs (e.g., donanemab, levodopa) | FDA-approved for this use | Donanemab slowed decline in mild disease[15] | Approved (not peptides) | n/a |
| Semaglutide (oral) | Human RCTs | No slowing of early Alzheimer’s in EVOKE and EVOKE+[1] | Not approved for Alzheimer’s | Semaglutide |
| Liraglutide | Human RCTs | Primary outcome negative in Alzheimer’s (ELAD)[2] | Not approved for Alzheimer’s | Liraglutide |
| Lixisenatide | Human RCTs | Small slowing of motor decline over 12 months[3] | Not approved for Parkinson’s | n/a |
| Exenatide | Human RCTs | No benefit in phase 3 Parkinson’s trial[4] | Not approved for Parkinson’s | Exenatide |
| Cerebrolysin | Small human studies | Short-term gains in small Alzheimer’s trials[6] | Not approved | Cerebrolysin |
| Semax | Animal or lab only (for these diseases) | Raised BDNF in rats[10] | Not approved | Semax |
| Dihexa | Animal or lab only | Improved memory in rat models[12] | Not approved | Dihexa |
| P21 | Animal or lab only | Reduced tau changes in mice[14] | Not approved | P21 |
Start with what works
Approved treatments come first. For Alzheimer’s, FDA approved donanemab in July 2024 for people with mild cognitive impairment or mild dementia; it slowed decline in its main trial.[15] Donanemab is an antibody, a large protein, not a peptide. Older drugs such as cholinesterase inhibitors ease symptoms. For Parkinson’s, levodopa and related drugs treat symptoms, and exercise and physical therapy help with movement. None of these slows the disease fully, which is why peptide research gets so much attention.
GLP-1 drugs in Alzheimer’s: the big negative trials
GLP-1 drugs are peptides that copy a gut hormone. Brain cells have GLP-1 receptors, and animal studies showed less inflammation and better nerve-cell survival. That made them leading candidates.
- EVOKE and EVOKE+ (2026): 3,808 people with amyloid-confirmed early Alzheimer’s took oral semaglutide up to 14 mg daily or placebo. After 104 weeks, the main dementia rating score worsened by about the same amount in both groups (2.3 versus 2.3 points in EVOKE). Both trials were stopped.[1]
- ELAD (2026): in 204 people with mild to moderate Alzheimer’s, daily liraglutide for 52 weeks did not change brain glucose use, the main outcome. One executive-function test favored liraglutide, but daily-living and dementia scores did not differ.[2]
We cover these trials in detail in GLP-1 for Alzheimer’s and Parkinson’s.
GLP-1 drugs in Parkinson’s: one signal, one failure
- LIXIPARK (2024): in 156 people diagnosed within 3 years, daily lixisenatide kept motor scores essentially flat for 12 months, while scores on placebo worsened by about 3 points. Nausea affected 46% of people on the drug.[3]
- Exenatide-PD3 (2025): in 194 people, weekly exenatide 2 mg for 96 weeks did not slow motor decline. It was well tolerated.[4]
- Pooled analysis (2026): across four randomized trials and 667 people, GLP-1 drugs did not improve motor scores, non-motor symptoms, cognition or quality of life. Weight loss was more common.[5]
Cerebrolysin: used abroad, unproven here
Cerebrolysin is an injected mix of small peptides and amino acids made from pig brain. Its maker says it is not registered with FDA or approved for sale in the US.[9]
- A 2015 meta-analysis of six placebo-controlled trials in mild to moderate Alzheimer’s reported better cognition at 4 weeks and better global ratings at 6 months.[6] The cognition effect was no longer significant at 6 months.
- In a 28-week trial of 197 people, Cerebrolysin, donepezil and the two combined performed similarly on cognition and daily function.[7]
- A Cochrane review in vascular dementia, a related condition, found six trials with 597 people. It reported small gains in cognition but called the evidence weak, citing high risk of bias, and said any benefit may be too small to matter clinically.[8]
There are no good trials of Cerebrolysin in Parkinson’s disease.
Semax, Dihexa and P21: lab and animal data only
- Semax is a nasal peptide prescribed in Russia, mainly for stroke. In rats it raised BDNF, a growth factor for nerve cells.[10] It has not been tested in Alzheimer’s or Parkinson’s trials, and FDA staff recommended against adding it to the compounding list in 2026.[11] See Dihexa vs Semax vs Selank.
- Dihexa improved memory in rat models of dementia.[12] FDA has found no human exposure data for it.[16] A drug acting on the same HGF pathway, fosgonimeton, missed its main and secondary goals in a 287-person phase 2/3 Alzheimer’s trial.[13]
- P21 reduced tau changes and memory problems in Alzheimer’s mice.[14] It has never been tested in people.
What’s next in the research
After EVOKE, the big open questions are timing and brain access. Researchers want to know whether starting years before symptoms would work better, and whether newer injectable or dual-action drugs reach the brain more effectively than oral semaglutide. Smaller studies continue: for example, a 60-person phase 2 trial at Imperial College London plans to test oral semaglutide in Alzheimer’s and was listed as not yet recruiting in October 2026.[17] In Parkinson’s, a larger and longer lixisenatide-style trial would be needed to confirm LIXIPARK’s small effect.[3] For Cerebrolysin, independent, well-blinded trials with long follow-up are still missing.
Safety and the risk of “research peptides”
Approved GLP-1 drugs have known side effects, mainly nausea, vomiting and weight loss, which can matter for frail older adults. Cerebrolysin requires repeated injections or infusions. Dihexa, P21 and Semax sold online as “research chemicals” are not checked for purity or dose, and their long-term effects in people are unknown. Dihexa activates a growth pathway also involved in some cancers, which is a theoretical concern with no human safety data to settle it.
Questions to ask your neurologist
- Am I a candidate for an approved disease-modifying treatment?
- Are there GLP-1 or other peptide trials I could join?
- Could any supplement or peptide interact with my current medicines?
For related reading, see best peptides for cognition, peptides for stroke recovery, peptides for TBI and concussion and our comparison of Selank vs Semax.
