Dihexa for Mild Cognitive Impairment (MCI) & Brain Aging: The 2026 UK Cognitive Decline Review

On 31 March 2026 NICE published its third draft guidance against routine NHS use of lecanemab and donanemab, the first two licensed disease-modifying drugs for early Alzheimer’s disease. The consultation closed on 28 April 2026 and the committee meets again on 10 June 2026. An estimated 70,000 people in England with mild cognitive impairment (MCI) or mild Alzheimer’s would have been eligible — and now have no disease-modifying option on the NHS. Predictably, search interest in “peptides for memory loss” and “Dihexa for cognitive decline” has surged. This 2026 review walks honestly through the question: what does Dihexa actually do at the synapse, how does that map onto the synaptic loss that defines MCI and the ageing brain, where does it stop being a substitute for licensed care, and what do clinicians, patients and families need to know before going anywhere near it?

MCI in the UK in 2026: Why the Treatment Gap Just Got Worse

For decades, mild cognitive impairment was treated as a research category with no useful clinical handle. That changed in 2024 when the MHRA licensed the first two anti-amyloid antibodies, lecanemab (Leqembi, August 2024) and donanemab (Kisunla, October 2024), for adults with MCI or mild dementia due to Alzheimer’s disease. The hope was real. For the first time, UK patients in the earliest symptomatic stage of the most common dementia had a drug that did something to the underlying biology.

That hope did not survive the cost-effectiveness review. NICE’s draft guidance in summer 2025 did not recommend either drug for routine NHS use. Following an appeal, NICE opened a third consultation on 31 March 2026, which closed on 28 April 2026; the committee meets on 10 June 2026 to evaluate the evidence again. The position published by Alzheimer Europe describes the regulator’s reasoning bluntly: the benefits are too small for the costs. The trials showed lecanemab slowed Alzheimer’s clinical progression by approximately 27% versus placebo over 18 months on the CDR-SB scale, and donanemab by approximately 36%. In absolute terms, NICE estimated this translates to a four-to-six-month delay in disease progression. Set against the cost of monthly intravenous infusions, amyloid PET imaging, regular MRI safety monitoring for amyloid-related imaging abnormalities (ARIA), and APOE genotyping, NICE concluded the cost-effectiveness threshold could not be met.

The arithmetic is brutal for the estimated 70,000 people in England NICE itself identified as eligible. Patients who were diagnosed in 2024-2025, told a treatment was coming, and made plans accordingly, now find themselves with no disease-modifying option. The House of Commons Library briefing on developments in dementia treatments documents the policy timeline in detail. The Alzheimer’s Research UK statement is unsurprisingly frustrated. The UK Dementia Research Institute response reads as a careful balance of disappointment and acceptance of the regulatory logic.

An additional methodological wrinkle has surfaced in 2026. A Cochrane review published in late 2025 questioned whether the cognitive benefits of anti-amyloid antibodies are clinically meaningful at all, arguing the differences fall below the minimum clinically important difference (MCID) on standard scales. If that interpretation prevails, the case for funding the drugs at scale weakens further. If it does not, the appeal may yet succeed. Either way, in April 2026 the practical position is the same: no NHS disease-modifying drug for MCI.

Layered on top of this is the structural pressure on UK memory services. NHS waits for a memory clinic assessment in 2026 average eight to twelve weeks in most areas, with longer waits in parts of the East of England and the Midlands. Some integrated care boards have explicitly deprioritised MCI follow-up while concentrating on confirmed dementia diagnoses, given budget constraints. The Alzheimer’s Society’s briefings in 2026 have repeatedly flagged the gap between policy ambition (the National Dementia Strategy) and operational reality (memory services running at 60-70% of needed capacity).

It is in this context — high-profile drug rejection, a 70,000-strong eligible population now with no NHS option, and growing public awareness that synaptic loss matters before plaques do — that interest in nootropic peptides, particularly Dihexa, has accelerated. The honest question is whether the science supports any of that interest, or whether it is mostly people understandably trying to do something while the system catches up.

MCI, Subjective Cognitive Decline & the Path to Dementia

Before assessing whether Dihexa could plausibly help, it is worth being precise about what mild cognitive impairment is — and what it isn’t.

The Three Stages: SCD, MCI, Dementia

Subjective cognitive decline (SCD) describes self-perceived cognitive decline that does not yet show on standard neuropsychological testing. People with SCD know something has changed; the tests do not yet detect it. Longitudinal studies show SCD approximately doubles the risk of progression to MCI compared with people who report no concern. SCD is now recognised as the earliest reported symptomatic stage of the Alzheimer’s continuum, although by definition it is also non-specific. Persistent SCD is a reasonable trigger for a memory-service referral.

Mild cognitive impairment (MCI) is the next stage. Per the American Academy of Neurology practice guideline and broadly compatible UK criteria, MCI is diagnosed when there is concern about a change in cognition (from the patient or a knowledgeable informant), objective impairment in one or more cognitive domains on testing, preserved independence in functional abilities, and an absence of dementia. The diagnosis is clinical, supported by neuropsychological testing.

MCI splits into amnestic MCI (memory predominant, more likely to progress to Alzheimer’s dementia) and non-amnestic MCI (executive, language or visuospatial predominant, more likely to progress to non-Alzheimer’s dementias such as frontotemporal dementia, dementia with Lewy bodies or vascular dementia). Roughly 5-15% of people with MCI progress to a dementia diagnosis each year. A non-trivial minority — perhaps 20% — revert to normal cognition over time, often because their MCI was driven by reversible contributors (depression, medication, sleep apnoea, B12 deficiency, thyroid dysfunction) rather than by neurodegeneration.

Dementia requires both objective cognitive impairment and impairment of independent functional abilities. Once functional impairment crosses that threshold, the diagnostic category and (in some places) the available pharmacology shift.

The Reversible MCI Workup — Why It Matters Before Any Peptide

A meaningful share of MCI cases are reversible if the contributor is identified and treated. A standard memory-service workup screens for: depression and anxiety; obstructive sleep apnoea; B12 and folate deficiency; thyroid dysfunction; uncontrolled hypertension or diabetes; alcohol misuse; polypharmacy with anticholinergic burden (a common iatrogenic cognitive impairment in older adults); chronic pain and chronic insomnia; and untreated hearing or vision loss. The 2024 Lancet Commission emphasises that several of these contributors operate even in confirmed neurodegenerative MCI, and addressing them improves cognition independently of the underlying disease process.

This matters before considering any unlicensed compound. A patient who self-treats suspected MCI with Dihexa without first having an evaluation may inadvertently mask a treatable cause — depression, sleep apnoea, B12 deficiency — while attributing any improvement to the peptide. The worst-case scenario is a missed diagnosis of a reversible condition; the second-worst is mis-crediting a peptide for what regular sleep and B12 supplementation actually fixed.

Biomarker Staging in 2026: Amyloid, Tau and the Blood Tests

One of the major changes in MCI care between 2020 and 2026 is the rise of blood-based biomarkers. Plasma p-tau217 and the Aβ42/40 ratio now have CE-marked assays available in the UK and are being deployed in research-active memory services and some private clinics. A positive biomarker pattern alongside MCI symptoms now supports a diagnosis of MCI due to Alzheimer’s disease. This staging matters because it identifies the population in whom anti-amyloid drugs would, in principle, work — and in whom a synaptogenic peptide’s mechanistic case is most defensible.

NICE NG97 was updated in 2024 to recognise the role of these biomarkers, but routine NHS deployment remains patchy in 2026. Private memory clinics in London, Manchester and Edinburgh increasingly offer plasma p-tau217 testing in the £200-£400 range. For families looking to understand “is this Alzheimer’s, vascular or something else?”, biomarker testing is becoming a meaningful clinical tool.

The Real Neurobiology: Synaptic Loss Drives Cognitive Decline

The dominant clinical story of Alzheimer’s disease for thirty years was the amyloid hypothesis: plaques cause everything. The dominant neuropathological story has always been more nuanced. Two seminal papers from the late 1980s and early 1990s established what every textbook now repeats: synapse loss is the strongest neuropathological correlate of cognitive impairment.

Robert D. Terry and colleagues, in their 1991 paper, found that synaptic density in the frontal and temporal cortex correlated more tightly with cognitive scores than plaque or tangle counts in Alzheimer’s patients. DeKosky and Scheff in 1990 reached the same conclusion in frontal cortex biopsies. The implication was profound: even if you can’t address plaques, addressing synaptic loss should address symptoms.

Synaptic Loss Is Already Present in MCI

Quantitative electron microscopy and immunostaining studies of the hippocampal CA1 region and entorhinal cortex show that synapse density is reduced by approximately 25% in MCI compared with cognitively normal age-matched controls, and by approximately 50% in mild Alzheimer’s dementia. These differences are not subtle. They predate measurable plaque burden in many cases. They appear in tau-positive neurons before those neurons die. And they are the most plausible direct cause of the memory and executive symptoms patients actually report.

The biological logic is straightforward. Cognition is a network property. Networks degrade gracefully with sparse pruning — cognitive reserve buffers individuals for years. But once a synaptic-density threshold is crossed, network performance falls off sharply. MCI is, on this reading, the threshold-crossing point.

The Aging Hippocampus and Dendritic Spines

The hippocampus is the brain region most vulnerable to age-related synaptic loss. Even in cognitively normal ageing, hippocampal volume declines by approximately 1-2% per year after age 60, with steeper decline in those who progress to MCI and Alzheimer’s. Dendritic spine density on CA1 pyramidal cells — the primary excitatory synapse type in episodic memory — declines in parallel.

This is precisely the substrate that BDNF and HGF support. BDNF promotes spine maturation and stabilisation through TrkB receptors. HGF promotes spine formation and density through c-Met receptors. The two pathways are partially redundant in the developing brain and partially divergent in the adult, with HGF signalling often more potent at the structural-spine end.

BDNF, MCI Progression and What the 2025 Evidence Says

The mechanistic case for BDNF in MCI is no longer just preclinical. A 2025 prospective cohort study tracked serum BDNF and progression to MCI in cognitively normal older adults; lower mid-life BDNF was associated with higher MCI risk. The 2025 NeuroMolecular Medicine review on BDNF signalling in Alzheimer’s neuropathology and treatment concluded that BDNF deficiency is mechanistically central to the disease. The same year saw a paper in Frontiers in Cellular Neuroscience proposing that stimulating myelin restoration with BDNF may be a promising therapeutic approach.

None of these papers tested Dihexa. But all of them strengthen the framework within which Dihexa’s mechanism becomes plausible: if MCI is a synaptic-loss disease, and if BDNF/HGF signalling is the pathway most directly responsible for spine maintenance, then a positive modulator of that pathway should — in theory — help. Theory is not evidence. But the gap between theory and evidence here is smaller than for any other indication on this site.

Where Dihexa Fits: HGF/c-Met and the Synaptic Hypothesis of Cognitive Decline

If you want to understand why Dihexa emerged as a candidate for MCI and Alzheimer’s disease at all, the path runs through the Wright laboratory at Washington State University and a series of papers in the late 2000s and early 2010s. The original observation was that angiotensin IV (AngIV) and its analogues facilitate hippocampus-dependent learning in rodent memory models. The Wright group, working with John Harding, refined the molecule across several iterations to produce N-hexanoyl-Tyr-Ile-(6) aminohexanoic amide — later named Dihexa or PNB-0408 — with improved oral bioavailability and brain penetration.

The HGF/c-Met Synaptogenic Pathway

The mechanism that emerged from this work is the central biological claim about Dihexa. The pivotal demonstration came in Benoist and colleagues: AngIV-derived peptides require activation of the hepatocyte growth factor (HGF) / c-Met system to produce their procognitive and synaptogenic effects. Block c-Met, and the cognitive benefit disappears. The pathway downstream of c-Met activation involves PI-3K/AKT signalling, which is the same intracellular cascade that supports neuronal survival, dendritic spine formation and synaptic plasticity in many other contexts.

Hepatocyte growth factor is, despite the name, a major growth factor in the central nervous system. The 2021 Frontiers review of HGF and MET in brain development and neurological disorders documents its roles in neuronal survival, dendritic outgrowth, axon guidance and synapse formation. The c-Met receptor is expressed on cortical and hippocampal neurons, including the CA1 pyramidal cells most relevant to memory.

The Alzheimer’s Mouse Evidence

The most relevant preclinical paper for the MCI question is Wright and colleagues (2021), which showed that Dihexa rescues cognitive impairment and recovers memory in the APP/PS1 mouse model of Alzheimer’s disease, with the effect blocked by inhibition of the PI-3K/AKT pathway. The APP/PS1 model overexpresses mutant amyloid precursor protein and presenilin 1, develops amyloid plaques and shows memory deficits in spatial tasks. Dihexa treatment partially restored performance to wild-type levels. The 2022 systematic review by Royea and Hamel on cognitive benefits of AngIV and Ang-(1-7) places this finding in the wider context of the angiotensin-derived peptide literature, where the procognitive signal is broad and consistent across deficit-state models.

The catch — and it is a real catch — is that reversal of cognitive deficit in a transgenic mouse is not the same as treatment of human MCI. The APP/PS1 model produces amyloid pathology but does not capture the tau, vascular and inflammatory components that characterise human Alzheimer’s. Animal-to-human translation in dementia drug development has a notoriously poor track record. The Athira/fosgonimeton story, covered next, is the most relevant cautionary parallel.

Why MCI Is the Strongest Mechanistic Fit for Dihexa

Across the indications covered on this site — ADHD, anxiety, depression, TBI, Long COVID brain fog, menopause brain fog and sleep-related vivid dreams — MCI is, on first principles, the cleanest fit. The pathology is dominated by synaptic loss; the mechanism of Dihexa is synapse-building. The age and demographic of the target population overlaps with the population in whom angiotensin-system biology is most prominent (older adults, often hypertensive). The translational distance from a deficit-state mouse model to a human MCI patient is shorter than from any other animal-to-human pairing on this site.

The catch, again, is the absence of any controlled human trial. We will turn to that next.

The Fosgonimeton Cautionary Tale: When the Mechanism Doesn’t Translate

The single most important data point for assessing Dihexa’s clinical prospects in MCI is the story of fosgonimeton (ATH-1017), the most clinically advanced HGF/MET positive modulator ever taken into humans. Fosgonimeton was developed by Athira Pharma, a Seattle-based biotech founded in 2008, with intellectual property and scientific lineage tracing directly back to the Wright laboratory work on Dihexa. It is Dihexa’s closest pharmacological cousin.

Athira advanced fosgonimeton through Phase 1 and Phase 2 trials with encouraging biomarker and exploratory cognitive signals, then launched LIFT-AD, a pivotal Phase 2/3 trial in patients with mild-to-moderate Alzheimer’s disease. The trial missed its primary endpoint in 2024. There were exploratory signals in subgroups (notably APOE4 non-carriers and patients not on acetylcholinesterase inhibitors) but the overall result was clearly negative. Subsequent share-price collapse and corporate restructuring at Athira was severe. The full editorial treatment is on the fosgonimeton page; the implications for Dihexa-class compounds are these:

The mechanism is real but not sufficient. HGF/MET activation does not, in mild-to-moderate Alzheimer’s disease, translate to a clinically meaningful cognitive benefit at the doses and durations tested. This may be because (a) by mild-to-moderate AD the synaptic substrate has been too eroded for synaptogenic intervention to rescue, (b) the dose was wrong, (c) the mechanism is genuinely insufficient in human disease, (d) trial design issues (placebo response, comorbid AChE inhibitor confound) obscured a real signal, or some combination.

The MCI window may be where Dihexa-class drugs would actually work, if they work. The therapeutic window for synaptogenic intervention is logically earlier in the disease than fosgonimeton was tested. By mild-to-moderate AD, plaques and tangles are mature, neuronal death is established, and the substrate for spine-rebuilding may simply not be there. MCI — especially MCI with biomarker evidence of underlying Alzheimer’s pathology — is theoretically a more favourable window. But this is hypothesis, not data.

Anyone considering Dihexa for MCI should know the fosgonimeton result. The closest clinical-stage relative of Dihexa, in the most rigorous large-scale trial ever conducted of an HGF/MET-activating compound in dementia, did not deliver a clinically meaningful benefit. Self-experimentation in 2026 with an unlicensed peptide of the same family, on the assumption it will work where its better-characterised cousin did not, is a strong claim that needs equally strong justification.

Dihexa vs Lecanemab and Donanemab: Different Targets, Different Risks

The comparison most frequently made in 2026 user forums is between Dihexa and the anti-amyloid antibodies. The comparison is misleading at the mechanism level but worth working through.

What Lecanemab and Donanemab Actually Do

Lecanemab (Eisai/Biogen, brand name Leqembi) and donanemab (Eli Lilly, brand name Kisunla) are monoclonal antibodies that bind beta-amyloid in the brain and recruit microglia to clear it via Fc-mediated phagocytosis. They are administered by intravenous infusion. They reduce amyloid plaque burden on PET imaging by 60-80% over 12-18 months. They slow clinical progression on the CDR-SB scale by approximately 27% (lecanemab) and 36% (donanemab) over 18 months versus placebo. The absolute differences are real but small — on the order of a four-to-six-month delay in clinical progression.

The dominant safety issues are amyloid-related imaging abnormalities, ARIA-E (oedema) and ARIA-H (haemorrhage). These appear in roughly 12-21% of treated patients depending on the drug and APOE genotype, with APOE4 homozygotes at substantially higher risk. The need for MRI safety monitoring at multiple timepoints is what drives much of the cost.

Dihexa Doesn’t Touch Amyloid

Dihexa has no documented effect on amyloid plaque burden. It does not bind amyloid; it does not recruit microglia; it does not alter the amyloid trajectory of the disease. Whether it would compensate for established amyloid-driven synaptic loss by promoting new spine formation is the open question. The APP/PS1 mouse data say yes, in mice, in early disease. Human data: none.

This means Dihexa and anti-amyloid antibodies operate on completely different parts of the pathological cascade. Anti-amyloid drugs target the upstream pathology (assuming the amyloid hypothesis is correct). Synaptogenic drugs target the downstream consequence. The two could in principle be complementary — clear the plaque, then rebuild the synapses. But there are no human data on this combination.

Risk Profile Comparison

Anti-amyloid antibodies have well-characterised risks: ARIA, infusion reactions, and the financial and logistical cost of monthly infusions. Dihexa’s risk profile is, by contrast, poorly characterised. The theoretical concerns flagged on the side effects page — activation of the c-Met pathway in tissues where c-Met overexpression is implicated in cancer (breast, ovarian, lung, gastric, colorectal), interactions with the renin-angiotensin system, cardiovascular effects from a peptide affecting blood pressure regulation pathways — remain theoretical because the data simply are not there. For older adults with multiple comorbidities and on multiple medications, “poorly characterised” is not a comforting risk descriptor.

The fundamental asymmetry is this: lecanemab and donanemab have a known modest benefit and known modest harms, both quantified to a clinical-trial standard. Dihexa has a hypothetical benefit and a hypothetical harm profile. Substituting hypothetical-for-known is rarely a sound clinical move.

The 2024 Lancet Commission: 14 Modifiable Risk Factors and the 45% Opportunity

The single most important non-pharmacological evidence base for the MCI conversation in 2026 is the 2024 Lancet Standing Commission on dementia prevention, intervention and care, chaired by Gill Livingston. It is the third in the series of Lancet Commissions (after 2017 and 2020) and represents the consensus of leading dementia researchers across the UK, US, Europe and Asia.

The 14 Modifiable Risk Factors

The 2024 Commission identified 14 modifiable risk factors that together account for approximately 45% of dementia cases worldwide when addressed across the life course. Two factors were added in 2024 versus the 2020 Commission: untreated vision loss and high LDL cholesterol. The full list, with the Commission’s population-attributable fraction estimates:

• Early life: Less education (5%) — the largest single early-life factor.
• Mid-life (age 45-65): Hearing loss (7%), high LDL cholesterol (7%), depression (3%), traumatic brain injury (3%), physical inactivity (2%), diabetes (2%), smoking (2%), hypertension (2%), obesity (1%), excessive alcohol (1%).
• Late life (over 65): Social isolation (5%), air pollution (3%), untreated vision loss (2%).

The Alzheimer’s Disease International summary sets out the headline finding clearly: tackling these factors at the population level could prevent or delay 45% of dementia cases globally. Alzheimer’s Research UK’s response emphasised the policy implication that brain health is a public health intervention, not just a clinical one.

Effect Size: Lifestyle Beats Any Drug Available in 2026

The blunt comparison: anti-amyloid antibodies slow clinical progression by ~30% in those who already have early Alzheimer’s. The 14 modifiable risk factors, addressed at the population level across the life course, prevent or delay 45% of cases entirely. The effect size of population lifestyle and vascular optimisation exceeds the effect size of any drug currently licensed or in development for dementia, including the anti-amyloid antibodies that have just been NICE-rejected.

This is not an argument against pharmacology. It is an argument that the most cost-effective intervention for an individual concerned about future cognitive decline in 2026 is, with high confidence: address the modifiable risk factors first. Hearing aids if hearing is reduced. Aggressive blood pressure control. Statins if LDL is high. Stop smoking. Treat depression. Move daily. Drink within UK guidelines (under 14 units/week). Address sleep apnoea. Stay socially engaged. None of this is glamorous and none of it is novel. It is also more evidence-supported than any peptide.

The FINGER Trial and Multi-Domain Lifestyle Interventions

The FINGER trial (Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability), published in The Lancet in 2015, was the first large randomised controlled trial of a structured multi-domain lifestyle intervention in older adults at risk of cognitive decline. The intervention — diet, exercise, cognitive training and vascular risk monitoring — produced a clear cognitive benefit over two years compared with usual-care control. The US-POINTER trial, an FINGER replication, reported positive results in 2025. The UK-FINGERS feasibility study reported in 2024 and is informing NHS pilot programmes in 2026.

The implication for Dihexa users is direct: the structured-lifestyle effect size in randomised trials of the most relevant population (older adults with cognitive concern but not yet dementia) is established, replicated and roughly comparable to the modest gains seen with the anti-amyloid antibodies — without the cost, the infusions or the ARIA risk. Anyone reaching for an unlicensed peptide before exhausting the structured-lifestyle option is, at the population level, optimising in the wrong order.

APOE4, Cognitive Reserve and Individual Risk Stratification

For readers who have already had genetic risk testing — increasingly common in 2026 — or who have a strong family history of dementia, an additional layer matters: APOE genotype.

What APOE4 Does

APOE encodes apolipoprotein E, a major lipid transport protein in the brain. The three common alleles are APOE2, APOE3 and APOE4. APOE4 is the strongest genetic risk factor for late-onset Alzheimer’s disease. One copy of APOE4 (heterozygous) approximately triples the risk; two copies (homozygous) raise it ten- to twelvefold. Approximately 25% of the UK population carries one copy of APOE4 and 2-3% carry two copies.

The biological mechanisms are multiple: APOE4 carriers have less efficient amyloid clearance, more aggressive synaptic pruning by complement proteins, lower BDNF signalling efficiency, more vascular pathology, and accelerated hippocampal atrophy from middle age onwards. APOE4 homozygotes are now considered, in some 2024-2025 reframings, to have a near-deterministic form of Alzheimer’s disease rather than just elevated risk — though this framing is contested.

APOE4 and Anti-Amyloid Antibody Eligibility

The anti-amyloid antibody trials demonstrated substantially higher ARIA risk in APOE4 homozygotes. As a result, the MHRA-licensed indications for both lecanemab and donanemab in the UK explicitly exclude APOE4 homozygotes, restricting use to non-carriers and heterozygotes. This means even before the NICE rejection, the highest-genetic-risk Alzheimer’s population was already excluded from the new drug class. NICE’s 70,000-eligible figure refers to non-homozygous APOE4 patients only.

APOE4 and Dihexa: A Speculative Position

There is no human Dihexa data stratified by APOE genotype. On the mechanism, an HGF/c-Met-targeted synaptogenic peptide should be APOE-agnostic in its primary action: it builds spines via PI-3K/AKT, not via amyloid clearance. In theory this could make synaptogenic peptides more attractive for APOE4 homozygotes, who have no anti-amyloid option. In practice, the absence of any safety or efficacy data in this population, combined with the higher cardiovascular and synaptic-vulnerability profile of APOE4 carriers, makes self-experimentation a particularly poor bet for this group. The fosgonimeton subgroup signal in APOE4 non-carriers further muddies the picture.

Cognitive Reserve and Why It Matters

Cognitive reserve — the brain’s capacity to maintain function in the face of pathology — is one of the most robust findings in dementia epidemiology. People with higher cognitive reserve, built from education, occupational complexity, lifelong learning, multilingualism, social engagement and physical activity, can carry substantially more amyloid and tau pathology before showing clinical symptoms. Cognitive reserve does not stop the underlying disease, but it shifts the threshold at which symptoms appear, sometimes by years.

The implication is significant: building cognitive reserve through lifestyle does not require a peptide, has the largest evidence base of any intervention, and is available to anyone. Dihexa could in theory contribute by promoting structural-substrate maintenance, but it is one input into a much larger, much-better-evidenced framework.

Practical Realities: If Someone Decides to Self-Experiment Anyway

This site exists because people will research Dihexa whether or not we cover it. The honest editorial position is that for MCI — as for the other indications covered — clinical assessment first and structured lifestyle work second is the right answer. But pretending self-experimentation does not happen would be dishonest. So here is the harm-reduction frame, as soberly as we can put it.

Dosing Considerations in Older Adults

The community dose range for Dihexa, drawn from the 2026 review of self-reported protocols, is typically 8-25 mg/day orally or sublingually for 8-16 weeks, often cycled. There are several reasons to consider the lower end of this range in older adults:

• Renal and hepatic function often declines with age. Peptide pharmacokinetics in older adults are not characterised for Dihexa.
• Polypharmacy is the rule rather than the exception for over-65s. Dihexa interactions with statins, antihypertensives, anticoagulants, SSRIs, gabapentinoids and acetylcholinesterase inhibitors are not characterised. The risk picture is one of unknowns multiplied.
• Cardiovascular comorbidity is common; Dihexa’s direct cardiovascular profile in humans is not established.
• Vivid dreams — a consistent Dihexa effect documented in the sleep and memory review — can become parasomnia or REM behaviour disorder in older adults, which is itself a prodromal feature of synucleinopathy. The interaction between Dihexa-driven dream intensity and pre-existing dream-enactment behaviour is wholly uncharacterised.

Cycling and Monitoring

The community consensus, such as it is, favours pulse-cycling (8-12 weeks on, 4-8 weeks off) rather than continuous use. The biological argument is partly speculative (down-regulation of c-Met receptors, opportunity for natural homeostasis) and partly cautious (limit duration of an unstudied exposure). For anyone in the MCI risk window who chooses to experiment, the cycling pattern is at least no less defensible than continuous dosing.

Monitoring should include, at minimum: subjective cognitive symptom diary (any clear change should be discussed with the GP and not attributed to the peptide without testing), blood pressure, sleep quality (a journal capturing dream content and any dream-enactment behaviour), and a baseline blood panel (FBC, U&Es, LFTs, lipids, B12, folate, ferritin, TFTs, HbA1c) before and after a cycle. Plasma p-tau217, where available privately, can be a useful longitudinal marker for the MCI cohort specifically.

When to Stop Immediately

Sudden onset of confusion, severe headaches, vision changes, focal neurological symptoms, syncope, palpitations or new chest pain require urgent medical attention regardless of cause. New or worsening dream-enactment behaviour (acting out dreams, kicking, falling out of bed) is a separate red flag specifically relevant to older adults and requires neurological evaluation. Any confirmed cancer diagnosis or strong family history of c-Met-implicated cancers warrants immediate cessation pending oncology input.

Who Should Not Consider Dihexa for MCI

A reasonable list, drawing on the broader side effects review and the specific characteristics of MCI populations:

• Anyone with a confirmed dementia diagnosis where capacity to consent to research-chemical use is compromised.
• Anyone with a personal or strong family history of breast, ovarian, lung, gastric or colorectal cancer (c-Met implicated tissues).
• Anyone on warfarin or DOAC anticoagulation without prescriber oversight.
• Anyone with poorly controlled hypertension or recent acute coronary syndrome.
• Anyone on multiple psychiatric or neurological medications without prescriber oversight.
• Anyone who has not yet had a memory-service workup — the diagnostic clarity of an unmasked symptom picture matters more than experimental neuroprotection.
• APOE4 homozygotes specifically: the genetic risk profile makes the absence of safety data more concerning, not less.

The Evidence-Based 2026 MCI Plan: What to Actually Do

If you take one section of this article seriously, take this one. For someone with MCI symptoms, suspected MCI or active concern about brain aging in the UK in 2026, here is the order of operations with the strongest evidence base.

1. Get assessed. GP referral to your local memory service (NHS) or to a private memory clinic offering biomarker-supported diagnostic workup. The aim is to identify reversible contributors and to establish, where possible, whether MCI is due to Alzheimer’s disease (biomarker-supported) or another aetiology. Alzheimer’s Society UK explains what to expect.
2. Treat reversible contributors. Depression, sleep apnoea, B12 deficiency, thyroid dysfunction, alcohol overuse, anticholinergic burden, untreated hypertension. Each of these can produce or worsen MCI symptoms, and each has well-evidenced treatment.
3. Address the 14 Lancet Commission risk factors. Hearing aids if hearing is impaired, statins if LDL is high (NICE QRISK threshold), aggressive blood pressure control, smoking cessation, alcohol below 14 UK units/week, structured aerobic exercise (150+ minutes/week), Mediterranean or MIND-style diet, cognitive engagement, social engagement, treatment of vision impairment.
4. Consider biomarker testing. Plasma p-tau217 and Aβ42/40 ratio are increasingly accessible privately. A negative biomarker result in someone with cognitive concern is reassuring; a positive result allows for early intervention planning.
5. Engage with research. The UK has a strong dementia research infrastructure. Join Dementia Research (joindementiaresearch.nihr.ac.uk) connects volunteers with clinical trials. Trials of next-generation anti-amyloid drugs, anti-tau drugs, and synaptic-protection drugs are recruiting in 2026. This is the route by which experimental drug access is meaningful and supervised.
6. Plan ahead. Lasting power of attorney, advance care planning, and discussions with family. These are uncomfortable conversations and they are easier to have early.
7. Stay engaged with the policy story. The NICE consultation on lecanemab and donanemab has not yet concluded as of late April 2026. The 10 June 2026 committee meeting may revisit access. If the appeal succeeds, the access landscape will change rapidly.

The BDNF/HGF Axis Across the Aging Brain: A Deeper Look

For readers who want to understand the underlying biology more thoroughly, the BDNF/HGF axis deserves a closer treatment. The story matters because it bridges what we know about MCI pathology with what Dihexa, in principle, addresses.

BDNF Across the Lifespan

BDNF expression in the cortex and hippocampus peaks in young adulthood and declines progressively through middle and late life. The decline is exacerbated by chronic stress, sedentary behaviour, sleep disruption, low-grade systemic inflammation, type 2 diabetes and several neurodegenerative diseases. Cross-sectional data find serum BDNF levels in cognitively normal 70-year-olds approximately 25-40% lower than in cognitively normal 30-year-olds.

The ProBDNF/mature-BDNF balance also shifts with age, in a direction that disfavours synaptic strengthening and favours synaptic pruning — the opposite of what an ageing brain actually needs. Several papers in 2024-2025 have refined the model that disrupted BDNF signalling, particularly an unfavourable proBDNF/mBDNF ratio, is a mechanistic driver of age-related cognitive decline rather than just a correlate.

HGF and c-Met in the Older Brain

HGF expression in the brain is more dynamic than BDNF’s. It rises in response to injury (the original reason fosgonimeton was tested) and is upregulated in early Alzheimer’s disease as part of an apparent endogenous compensatory response. The compensation is, however, insufficient. c-Met receptor density on hippocampal neurons appears to be relatively preserved in early disease but declines with progression.

The therapeutic implication: the HGF/c-Met system is biologically receptive in early disease and less so in advanced disease. This argues, mechanistically, that an HGF-positive modulator is more likely to work in MCI and very early AD than in mild-to-moderate AD — which is, again, the trial population in which fosgonimeton failed. Whether this argues for an MCI trial of a synaptogenic peptide, or against one (because the biological signal in MCI is too small to be statistically detectable), is exactly the question that randomised controlled trials would answer if they existed.

The Exercise-BDNF Connection

Worth noting because it is the most evidence-supported way to raise BDNF in older adults. Aerobic exercise, particularly at moderate-to-vigorous intensity for 30-45 minutes three to five times per week, raises peripheral BDNF and is associated with better cognitive outcomes, slower hippocampal atrophy, and reduced dementia risk. Resistance training adds independent benefit. The effect sizes from structured exercise in older adults exceed those reported for any pharmacological cognitive enhancer in non-deficit populations.

Anyone considering a peptide that targets BDNF/HGF biology should, on first principles, have a structured exercise programme already in place. Building synapses on top of a sedentary baseline is a much weaker proposition than building them on top of a metabolically active, vascularly healthy brain.

2026 News Context: What Is Actually Happening This Year

For SEO transparency and reader value, the following 2026 events frame the current MCI conversation in the UK and are referenced where relevant in this article:

• 31 March 2026 — NICE third draft guidance: NICE opens consultation on its third version of draft guidance against routine NHS use of lecanemab and donanemab. Consultation closes 28 April 2026. Committee meets 10 June 2026. Pharmaceutical Journal coverage.
• 20 March 2026 — Alzheimer’s Society response: Alzheimer’s Society’s comment on the NICE review process explicitly framed the consultation as a critical opportunity for patients with mild Alzheimer’s and MCI. Alzheimer’s Society UK comment.
• Late 2025 — Cochrane review: A Cochrane review questioned whether anti-amyloid antibody benefits exceed the minimum clinically important difference. UK DRI response.
• 2024 — Lancet Commission, third edition: 14 modifiable risk factors, 45% prevention potential, two new factors added (vision loss, high LDL cholesterol). Full open access summary.
• October 2024 — Donanemab licensed by MHRA: Second anti-amyloid antibody to receive UK regulatory approval, immediately referred to NICE for cost-effectiveness review.
• August 2024 — Lecanemab licensed by MHRA: First anti-amyloid antibody licensed in the UK; NICE rejection followed in summer 2025.
• 2024 — Athira LIFT-AD Phase 3 readout: Fosgonimeton missed primary endpoint in mild-to-moderate Alzheimer’s disease. The most directly relevant clinical signal for the entire HGF/MET-positive-modulator class.

Together these events define the contemporary policy and evidence environment. The trajectory across 2024-2026 has been: regulatory approval of anti-amyloid drugs, then NHS cost-effectiveness rejection, then Cochrane scepticism on clinical meaningfulness, then a re-opened NICE consultation. The fosgonimeton failure sits in the background as a reminder that synaptogenic-peptide mechanism does not, by itself, deliver clinical benefit in human dementia.

The Bottom Line: Synaptic Logic, Zero Human MCI Evidence

The 2026 reading on Dihexa for MCI is more nuanced than for any other indication on this site. The mechanism is the cleanest fit. The pathology of MCI is dominated by synaptic loss, and Dihexa is, in mechanism, a synapse-builder. The 2024 Lancet Commission’s 14 modifiable risk factors include several that converge on BDNF-driven plasticity (exercise, depression, alcohol, sleep), suggesting the BDNF/HGF axis is biologically central to cognitive decline. The 2025 evidence linking lower mid-life BDNF with higher progression to MCI strengthens the framework further.

And yet: there is no controlled human trial. The closest clinical-stage relative, fosgonimeton, missed its Alzheimer’s Phase 3 primary endpoint in 2024. The therapeutic window in which a synaptogenic peptide should plausibly work (early MCI, biomarker-positive Alzheimer’s) has not been formally tested for any HGF/c-Met modulator. The safety profile in older adults — the population in whom MCI exists — is wholly uncharacterised, with theoretical concerns around c-Met activation and cancer-relevant tissues, plus the practical concern of polypharmacy interaction.

The honest 2026 reading: clinical assessment first, treatment of reversible contributors second, structured-lifestyle work always, anti-amyloid antibodies via clinical trial or private route if biomarker-positive, off-label adjuncts only with prescriber oversight, research chemicals essentially last — if at all. The right next step for most readers of this page is not a peptide vendor. It is a GP appointment, a memory-service referral letter, and a structured-exercise programme starting tomorrow.

Frequently Asked Questions

Related Reading on Dihexa.co.uk

• Dihexa & Alzheimer’s Research — the academic origin story of why Alzheimer’s was the original Dihexa target.
• Fosgonimeton & Athira — the closest clinical-stage relative, the Phase 3 failure and what it means for the synaptogenic-peptide class.
• Dihexa vs BDNF: What “10 Million Times More Potent” Actually Means — the central potency claim, examined.
• Dihexa for TBI, Concussion & Stroke Recovery — the neurorepair indication that shares the synaptic-rebuilding rationale.
• Dihexa for Menopause & Perimenopause Brain Fog — the estrogen-BDNF axis and a closely related cognitive-symptom phenotype.
• Dihexa for Long COVID Brain Fog — another major cognitive-symptom indication where the synaptic case is increasingly clear.
• Dihexa for Depression & Mood — the synaptogenic hypothesis of depression, which overlaps with MCI risk via the depression-as-risk-factor pathway.
• Dihexa for Anxiety & Chronic Stress — chronic stress accelerates hippocampal atrophy and is a Lancet Commission risk factor.
• Dihexa, Sleep & Memory Consolidation — how sleep-stage memory consolidation depends on the same BDNF/HGF biology.
• Dihexa for ADHD — the prefrontal cortex angle on the same broader synaptic story.
• Dihexa Review 2026 — effects timeline, oral vs sublingual, cycling protocols.
• Dihexa Stacking Guide — why combining with cholinesterase inhibitors or memantine needs prescriber oversight.
• Dihexa for Cognitive Enhancement — the broader healthy-adult conversation.
• Mechanism of Action — HGF/c-Met, PI-3K/AKT, dendritic spines.
• Benefits Overview — the broader claimed-benefit landscape, evidence-rated.
• Dosage Guide — community dose ranges and considerations.
• Side Effects & Risks — the general safety picture.
• UK Legal Status — where Dihexa sits in UK law and MHRA advertising rules.
• Research & Studies — the human and animal evidence base reviewed.
• Dihexa vs Other Nootropics — how Dihexa compares with Semax, Selank, Noopept, BPC-157 and racetams for cognitive indications.
• Glossary — technical terms used on this page.
• Full Site FAQ — the broader question set across the site.

External Authoritative Sources Cited

• Livingston G et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet (2024).
• Alzheimer Europe. NICE final guidance recommending against routine NHS use of donanemab and lecanemab.
• Alzheimer’s Society UK. Comment on NICE reviewing lecanemab and donanemab (March 2026).
• House of Commons Library. Developments in dementia treatments briefing.
• Alzheimer’s Research UK. NHS spending watchdog to review evidence on Alzheimer’s treatments.
• Alzheimer’s Research UK. Second Alzheimer’s drug, donanemab, not available on NHS.
• UK Dementia Research Institute. Response to NICE guidance on donanemab and lecanemab.
• UK Dementia Research Institute. Response to Cochrane review on anti-amyloid Alzheimer’s drugs.
• The Pharmaceutical Journal. NICE opens consultation on previously-rejected Alzheimer’s treatments donanemab and lecanemab.
• Alzheimer’s Disease International. Lancet Commission identifies two new risk factors for dementia.
• Alzheimer’s Research UK. Addressing 14 health and lifestyle factors could prevent nearly half of global dementia cases.
• Terry RD et al. Physical basis of cognitive alterations in Alzheimer’s disease: synapse loss is the major correlate of cognitive impairment (Annals of Neurology, 1991).
• DeKosky ST & Scheff SW. Synapse loss in frontal cortex biopsies in Alzheimer’s disease: correlation with cognitive severity (Annals of Neurology, 1990).
• Petersen RC et al. Practice guideline update summary: Mild cognitive impairment (American Academy of Neurology, 2018).
• NICE NG97. Dementia: assessment, management and support for people living with dementia and their carers.
• Alzheimer’s Society UK. What is mild cognitive impairment?
• Serum BDNF and progression to MCI in cognitively normal older adults: a prospective cohort study (2025).
• Unlocking Alzheimer’s Disease: The Role of BDNF Signaling in Neuropathology and Treatment (NeuroMolecular Medicine, 2025).
• Stimulating myelin restoration with BDNF: a promising therapeutic approach for Alzheimer’s disease (Frontiers in Cellular Neuroscience, 2024).
• Wright JW et al. AngIV-Analog Dihexa Rescues Cognitive Impairment and Recovers Memory in the APP/PS1 Mouse via the PI3K/AKT Signaling Pathway (2021).
• Benoist CC et al. The Procognitive and Synaptogenic Effects of Angiotensin IV-Derived Peptides Are Dependent on Activation of the HGF/c-Met System.
• Royea J & Hamel E. Cognitive benefits of Angiotensin IV and Angiotensin-(1-7): a systematic review of experimental studies.
• HGF and MET: From Brain Development to Neurological Disorders (Frontiers in Cell and Developmental Biology, 2021).
• Join Dementia Research (NIHR).
• NHS dementia information hub.
• Dementia UK Admiral Nurse helpline.