Skip to content
All life stages
Life stage

Dementia & Alzheimer's MCI

Standard-of-care first for clinical dementia and MCI — biomarker diagnosis, the MAb era (lecanemab, donanemab), the ChEI + memantine ladder, reversible-cause workup, and the narrow peptide adjunct case.

What changes during this transition

Clinical dementia and amnestic-MCI are a different conversation from generic 'cognitive decline' or healthy-adult nootropic use. The diagnostic frame is NIA-AA criteria with biomarker confirmation where available — CSF Aβ42/40 plus p-tau181 or p-tau217, amyloid PET, brain MRI to rule out vascular and structural contributors, and the newer blood-based p-tau217 assays (the Lilly Lumipulse assay received FDA clearance for AD diagnostic use in 2025). MoCA and MMSE remain useful for staging and tracking; caregiver-reported ADL and IADL anchors do most of the load-bearing work in functional terms. The standard-of-care backdrop reshaped substantially between 2023 and 2025. The acetylcholinesterase inhibitor + memantine ladder (donepezil, rivastigmine, galantamine for mild-to-moderate AD; memantine added at moderate-to-severe stages) remains the symptomatic backbone. On top of that, the amyloid-targeting monoclonal antibodies are now the disease-modifying conversation: lecanemab (Leqembi) received FDA approval in July 2023 on the CLARITY-AD trial (27% slowing of CDR-SB decline at 18 months in mild AD/MCI), and donanemab (Kisunla) followed in July 2024 on the TRAILBLAZER-ALZ 2 trial (35% slowing in the low/medium tau subgroup). Both require biomarker-confirmed amyloid pathology, both carry meaningful ARIA-E (edema) and ARIA-H (hemorrhage) risk that demands surveillance MRI scheduling, and both concentrate the highest ARIA risk in APOE ε4 homozygotes — genotyping is now part of the eligibility conversation. Aducanumab was withdrawn from the market in 2024. Where peptides fit, narrowly: cerebrolysin is the library's strongest-case substrate for clinical dementia — a porcine brain-derived multi-peptide preparation registered as a labeled dementia adjunct in 50+ jurisdictions (EU member states, Russia, China, Korea, SE Asia, LATAM, India), with a moderate-quality Cochrane evidence base in vascular dementia and signal in AD across the Ruether 2001, Alvarez 2011, and Heiss 2007 trial program. It sits as an adjunct alongside the standard ladder, not a substitute. Semax carries a Russian Ministry of Health registered indication for cerebrovascular insufficiency that overlaps cleanly with vascular cognitive impairment, with cross-jurisdictional replication absent. B12-methylcobalamin belongs in every MCI workup as a reversible-cause check — Lindenbaum 1988 (NEJM) established that B12-deficiency neurological manifestations occur without macrocytic anemia, and the VITACOG program (Smith 2010, de Jager 2012) showed homocysteine-stratified MCI benefit from B12 + folate + B6 repletion. What's not surfaced here: dihexa and P21. Dihexa's community framing rests on the Benoist 2014 rodent paper that was formally retracted in April 2025 after the January 2025 Athira Pharma DOJ data-integrity settlement, with downstream concerns about c-Met oncogenic activation in an elderly population and the failed ATH-1017 LIFT-AD Phase 2/3 program. P21 (Khavinson-tradition tetrapeptide, sometimes conflated with the separate Iqbal-lab tau-pathology peptide) has no human RCT and no cross-jurisdictional replication. Both are honest-answer substrate entries for users who ask, not discovery options to elevate.

Important caveat

Standard-of-care first, always: NIA-AA criteria diagnosis with biomarker confirmation where available (CSF Aβ42/40 + p-tau181 or p-tau217, amyloid PET, blood-based p-tau217), brain MRI, and a clean reversible-cause workup (B12 with MMA + homocysteine, TSH, depression-pseudodementia screen, normal-pressure hydrocephalus assessment, medication review, sleep apnea, sensory deprivation including hearing aids per Lancet Commission). The symptomatic ladder (ChEI + memantine) and the disease-modifying conversation (lecanemab, donanemab where eligible, with ARIA surveillance MRI scheduling and APOE ε4 genotyping context) belong to the neurologist or geriatrician owning the diagnosis. The 2024 Lancet Commission modifiable risk factors — vascular risk control, hearing aids, social engagement, cognitive engagement, physical activity, sleep, alcohol, air pollution, head-injury prevention — do more population-level work than any peptide will at the individual level. Caregiver burden is a real clinical entity and respite/support planning belongs in the conversation. Peptides are adjuncts, not substitutes, and the cerebrolysin case is strong in its source jurisdictions but doesn't replace the standard ladder anywhere. Dihexa carries a data-integrity history (Benoist 2014 retracted April 2025, Athira DOJ settlement January 2025) and a c-Met oncogenic-mechanism concern; P21's evidence base is essentially absent in this trigger context. Neither belongs in a clinical dementia protocol.

Peptides editorially relevant to dementia & alzheimer's mci

3 peptides from the library — each evidence-tiered honestly.

Want this list to grow? The library is editorial — if there’s a peptide you think belongs on this page with documented or mechanistically-clear evidence, send us a note with the citation and we’ll review it under the same evidence-tier discipline as every other entry.