Dihexa for Sleep Apnea Brain Fog: OSA, Hypoxemia, Hippocampal Damage, Zepbound (Tirzepatide) SURMOUNT-OSA & the 2026 UK Review

Around 13 million UK adults are estimated to have obstructive sleep apnea (OSA) on combined Sleep Apnoea Trust Association and Asthma + Lung UK figures — an overall prevalence of roughly 19.5%, with an estimated 80–85% remaining undiagnosed [1]. 2024–2026 has been an unusually consequential window: on 20 December 2024 the FDA approved Zepbound (tirzepatide) as the first ever prescription medicine for moderate-to-severe OSA in adults with obesity on the back of SURMOUNT-OSA (NEJM, June 2024); in May 2025 the UC Irvine group linked REM-sleep hypoxemia specifically to CA1 hippocampal volume loss and memory impairment; a 2025 systematic review and meta-analysis of 23 studies across 33,226 patients estimated a 36.92% pooled prevalence of cognitive impairment in adults with OSA; and the 2025 Neurology paper from Singh and colleagues connected nocturnal hypoxemia to white matter hyperintensities and temporal lobe changes in older adults. Within this picture, Dihexa — a synaptogenic peptide that activates the HGF/c-Met growth-factor system — is mechanistically interesting for the cognitive aftermath of intermittent hypoxia. It is not, however, a proven treatment for OSA or OSA brain fog. This 2026 UK evidence review summarises the biology, the regulatory landscape, the comparison with licensed treatments, and where the evidence is genuinely absent.

OSA in 2026: From Snoring Disorder to Brain Disorder

The clinical model of obstructive sleep apnea has shifted. For most of the late twentieth century, OSA was understood primarily as a cardiopulmonary problem — loud snoring, daytime sleepiness, hypertension, and an elevated risk of fatal cardiovascular events. The interventions reflected that model: continuous positive airway pressure (CPAP) to mechanically splint the upper airway, lifestyle measures to reduce weight, and surgical options for selected patients. They worked, often. They did not, however, fully explain why so many treated patients continued to report persistent cognitive symptoms long after their apnoea-hypopnoea index (AHI) had normalised.

Modern translational neuroscience reframes OSA as a brain disorder as much as a breathing disorder. Three concurrent insults dominate the cognitive picture:

• Chronic intermittent hypoxia. Recurrent oxygen desaturations — often hundreds per night — produce a hypoxia-reoxygenation injury pattern that resembles ischaemia-reperfusion in stroke biology, driving oxidative stress, neuroinflammation, and endothelial dysfunction.
• Sleep fragmentation. Repeated arousals collapse normal sleep architecture. Patients lose slow-wave (N3) and REM sleep, the stages most important for memory consolidation, glymphatic clearance, and emotional regulation.
• Autonomic surge. Each apnoea triggers a sympathetic burst — tachycardia, hypertension, cortisol release — that elevates allostatic load and damages the same prefrontal-hippocampal circuits described in the Dihexa anxiety and chronic stress review.

The downstream picture is consistent across imaging modalities: hippocampal volume loss (particularly CA1 subregion), temporal lobe atrophy, white matter hyperintensities, thinning of the prefrontal cortex, reduced cerebral perfusion, and elevated brain amyloid burden on PET. Within this circuit-based view, the question is no longer "how loud is the snoring" but "how much structural and functional brain damage has accumulated, and how do we reverse it". That is the question Dihexa's mechanism — covered in detail in the mechanism of action guide — was originally engineered to answer.

REM-Sleep Hypoxemia and the CA1 Hippocampal Story

The single most important neuroscience finding of 2025 for OSA brain fog came from the University of California, Irvine group in May 2025. Until then, the assumption was that overall AHI — the average number of breathing pauses per hour — was the best predictor of cognitive risk. UC Irvine's data turned that assumption on its head.

Working with older adults at risk for dementia, the team showed that hypoxemia during rapid eye movement sleep specifically mediates memory impairment via CA1 hippocampal volume loss. REM accounts for only about 20–25% of overnight sleep, but it is the stage where airway tone collapses most severely — muscle atonia is at its deepest — and where oxygen desaturations are typically longest and most profound. CA1 is the hippocampal subregion most vulnerable to hypoxic injury, the same subfield where early Alzheimer's pathology is most consistently observed, and the region with the highest density of BDNF-dependent synaptic plasticity machinery.

The biological implications are substantial:

• A patient with a "modest" total AHI of 15 events/hour but severe REM-AHI of 60 events/hour may be at higher cognitive risk than someone with overall AHI of 25 spread evenly across the night.
• Standard polysomnography reports often understate REM-specific risk because AHI is averaged across all sleep stages.
• CPAP titration usually optimises against total AHI, not REM-AHI specifically, leaving some patients with residual REM-stage events even on therapy.
• The CA1 region is exactly where any synaptogenic intervention — BDNF mimetic, HGF/c-Met agonist, or ketamine-class plasticity agent — would theoretically be most useful.

This refines the picture from "OSA causes brain damage" to "REM-stage OSA causes CA1-localised structural damage that predicts memory-specific cognitive decline". And it identifies the precise circuit where a synaptogenic compound would have to work to make a difference. The full mechanistic and potency comparison with BDNF, including the interpretation of the often-quoted "10 million times more potent" figure, is in the Dihexa vs BDNF deep-dive.

What Chronic Intermittent Hypoxia Actually Does to the Brain

The structural neurobiology of OSA is now one of the best-characterised sleep-related findings in modern medicine. Across rodent intermittent-hypoxia models and human imaging studies, the recurring pattern is consistent:

• Hippocampal CA1 vulnerability. Pyramidal neurons in the CA1 region undergo dendritic spine loss and dendritic shrinkage under intermittent hypoxia, with corresponding volume reductions on MRI in human OSA cohorts [8].
• White matter hyperintensities. The April 2025 Neurology paper from Singh and colleagues found nocturnal hypoxemia significantly associated with white matter hyperintensities and temporal lobe changes — the small-vessel disease biomarker that also tracks with vascular cognitive impairment [9].
• Reduced BDNF in PFC and hippocampus. Intermittent hypoxia downregulates Brain-Derived Neurotrophic Factor expression in cortical and hippocampal regions, the growth factor most central to synaptic plasticity and to the synaptic-plasticity hypothesis of depression covered in the Dihexa depression review.
• Beta-amyloid accumulation. OSA patients show elevated brain amyloid on PET, with mechanistic studies suggesting impaired glymphatic clearance during fragmented sleep as a key driver — the link between OSA and Alzheimer's pathology covered in the Alzheimer's research page.
• Endothelial and blood-brain barrier dysfunction. Cyclical hypoxia damages cerebral microvasculature, exactly the angiogenesis biology that HGF/c-Met signalling is implicated in repairing — see the vascular dementia review for the full small-vessel story.
• Autonomic dysregulation. Chronic OSA flattens heart-rate variability, blunts the cortisol awakening response, and disrupts the normal nocturnal blood-pressure dip — a physiological signature shared with chronic stress and burnout.

That last cluster matters because it explains why "just use the CPAP" — though clinically essential — is sometimes not enough. The regulatory machinery is itself impaired by the time many patients are diagnosed. The brain is not waiting to reset. It is structurally remodelled, and reversing that remodelling takes either time or active intervention. This is the core insight that motivates interest in synaptogenic compounds: if we have agents that can rebuild hippocampal and prefrontal connectivity faster than spontaneous post-CPAP recovery, the cognitive residua of long-undiagnosed OSA might in principle be addressed.

The 2025 Cognitive-Impairment-in-OSA Meta-Analysis

If a single number captures the brain-fog burden of OSA in 2026, it is the figure from the 2025 systematic review and meta-analysis across 23 studies and 33,226 patients: a 36.92% pooled prevalence of cognitive impairment in adults diagnosed with OSA. Subgroup analyses found a graded relationship: prevalence rose with OSA severity (defined by AHI bands), with age, and with the time spent below 90% oxygen saturation overnight (T90).

For context, that prevalence is broadly comparable with the cognitive-impairment rates seen in long-standing untreated type 2 diabetes, in moderate chronic kidney disease, and in survivors of significant stroke. It is not a small effect.

The same year produced two further evidence updates worth highlighting:

• The PMC12022498 hippocampal-volume study examined 166 patients with polysomnography or home sleep apnoea testing alongside structural MRI. In patients with neurodegenerative, vascular or mixed cognitive impairment, OSA presence, T90 and mean SpO2 were each significantly associated with reduced total hippocampal volumes — though, importantly, not in patients with purely subjective cognitive complaints. This separates "perceived" from "structural" brain fog.
• The 2025 Journal of Neurology systematic review and meta-analysis of longitudinal studies updated the broader sleep-disorders-and-cognitive-decline picture, confirming OSA as one of the most consistent sleep-related modifiable risk factors for dementia.

The clinical implication is uncomfortable: a substantial subgroup of patients with mild cognitive impairment (MCI) covered in the Dihexa MCI and brain aging review in fact have undiagnosed OSA driving or amplifying their decline. The first-line response is screening and treatment of the breathing disorder — not a peptide.

BDNF, HGF/c-Met and the Synaptogenic Hypothesis of OSA Brain Fog

If there is one molecular finding that anchors the synaptic-plasticity hypothesis of OSA cognitive impairment, it is the reduction in BDNF in cortical and hippocampal regions under chronic intermittent hypoxia, paralleled by changes in BDNF-related signalling pathways:

• Rodent intermittent-hypoxia models reliably show reduced BDNF expression in hippocampus, prefrontal cortex and brainstem nuclei involved in respiratory control.
• Human OSA cohorts show altered peripheral BDNF, with directionality complicated by acute versus chronic disease state, treatment status, and comorbid metabolic disease — a pattern partially analogous to the elevated peripheral BDNF observed in fibromyalgia.
• BDNF-dependent long-term potentiation in CA1 is impaired by intermittent hypoxia in animal slice preparations, exactly the synaptic process most relevant to memory consolidation.
• The Val66Met BDNF polymorphism, which reduces activity-dependent BDNF release, may exacerbate cognitive vulnerability to OSA insults — though the literature is still small.

The clinical implication is the same as for depression and anxiety: any compound that supports BDNF-like plasticity in CA1 and the prefrontal cortex is theoretically procognitive after OSA. This is not the same as expanding the airway (the CPAP and surgical model) or improving ventilatory drive (the rare pharmacological respiratory-stimulant model). It is structural — it changes how easily new synapses can form and consolidate after the hypoxic insult has stopped.

Dihexa is not BDNF, but it activates HGF/c-Met, which converges on overlapping intracellular pathways (PI3K/Akt, ERK, mTOR) and produces dendritic spine growth in the same regions. HGF/c-Met biology is also relevant to the vascular and angiogenic component of OSA brain injury: HGF promotes endothelial repair, blood-brain barrier integrity and microvascular angiogenesis, all of which are damaged by chronic intermittent hypoxia. Whether Dihexa produces enough plasticity in CA1 specifically, and whether it engages the vascular HGF biology meaningfully in humans, is unknown — the relevant experiments have not been done.

Zepbound, Tirzepatide and SURMOUNT-OSA: The 2024 Inflection Point

For decades, OSA had no licensed pharmacological treatment. On 20 December 2024, that changed. The FDA approved Zepbound (tirzepatide) as the first and only prescription medicine for moderate-to-severe obstructive sleep apnea in adults with obesity, on the basis of the SURMOUNT-OSA Phase 3 trial published in the New England Journal of Medicine in June 2024.

The trial design and results are worth understanding because they directly shape the cognitive-recovery landscape:

• SURMOUNT-OSA enrolled adults with moderate-to-severe OSA and obesity across two intervention-specific appendices: ISA-1 (no current PAP therapy) and ISA-2 (already on PAP therapy).
• The primary endpoint was change in apnoea-hypopnoea index (AHI) at week 52.
• Tirzepatide produced a reduction of 25–29 apnoea-hypopnoea events per hour versus placebo, with mean body weight reduction of 18–20% versus 2% on placebo.
• The benefit was preserved in patients who continued PAP therapy — i.e. tirzepatide is additive to, not a substitute for, mechanical airway support.
• Secondary cardiometabolic outcomes (blood pressure, hsCRP, lipids) improved in line with the SURMOUNT-1, SURMOUNT-2 and SURMOUNT-3 obesity programmes.
• Patient-reported outcomes, including daytime sleepiness on the Epworth Sleepiness Scale, improved meaningfully.

For the cognitive story, SURMOUNT-OSA did not include a primary cognitive endpoint — that gap will need separate trials, including the Sunnybrook NCT06773416 CPAP-cognition RCT (started February 2025, primary completion September 2027) and the CRESCENDO trial (NCT06983769) comparing CPAP and MAD in cognitively impaired patients (started January 2026). But mechanistically, reducing nocturnal hypoxia by reducing OSA severity should attenuate the hippocampal injury described above. The same drug class is covered in much more depth in our dedicated Dihexa and GLP-1 drugs (Ozempic, Wegovy, Mounjaro) review, including the broader brain-health and APOE4-prevention story and the EVOKE Phase 3 semaglutide-in-Alzheimer's failure.

In the UK, tirzepatide is available under the brand Mounjaro for type 2 diabetes and obesity (NICE TA1026), but as of May 2026 there is no NICE technology appraisal specifically for the OSA indication. Private prescribers and specialist sleep services in 2026 are increasingly using the obesity indication to support patients with comorbid OSA, but this is off the OSA-specific licence in the UK and is not routinely NHS-funded for OSA itself.

Where Dihexa Fits the Sleep Apnea Picture

Dihexa is a small-molecule peptide analogue developed from angiotensin IV (Nle¹-AngIV). Its mechanism, covered in full in the mechanism of action guide, has three features relevant to OSA brain fog:

• Direct HGF/c-Met activation. The hepatocyte growth factor / c-Met receptor system is widely expressed in cortex, hippocampus and cerebral microvasculature, and is one of the strongest endogenous drivers of both synapse formation and angiogenic repair. Activating it pharmacologically provides a route to synaptogenesis and vascular repair that does not require disinhibition (as ketamine does) or weeks of monoamine signalling (as SSRIs do).
• Rapid spine formation. Dihexa has been reported to induce new dendritic spines within hours in cell-culture and slice-physiology preparations — a timescale comparable to ketamine's reported synaptogenic burst, but without NMDA antagonism.
• Blood-brain barrier penetration. Unlike BDNF or recombinant HGF, Dihexa was specifically engineered for oral bioavailability and central penetration. This solves the bioavailability problem that has historically prevented direct neurotrophin therapy for any neurodegenerative or hypoxic-injury indication.

Within the synaptic-plasticity model of OSA cognitive impairment, that mechanistic profile is appealing: a fast, BBB-penetrant compound that drives synaptogenesis and supports vascular HGF/c-Met biology in regions where intermittent hypoxia has caused both synaptic and microvascular damage. On paper.

The translational gap is the same as for anxiety, depression and most other conditions covered on this site. No published clinical trial has tested Dihexa in any sleep-apnea, central-sleep-apnea, chronic-intermittent-hypoxia or post-CPAP cognitive impairment population. No open-label series in OSA patients exists in the peer-reviewed literature. The closest clinical relative — fosgonimeton (ATH-1017) — was developed for Alzheimer's disease and did not test OSA endpoints; its Phase 3 LIFT-AD trial did not meet its primary cognitive endpoint in September 2024, and the Athira SHAPE programme in Parkinson's disease dementia and dementia with Lewy bodies was subsequently discontinued. A plausible mechanism is a reason to investigate, not evidence of effect, and decades of cognitive-enhancement drug development is full of compounds that worked in animal hypoxia paradigms and failed in humans.

The NHS UK OSA Pathway: NICE NG202, CPAP and Inspire

OSA care in the UK is governed by NICE NG202 (Obstructive sleep apnoea/hypopnoea syndrome and obesity hypoventilation syndrome in over-16s), published in August 2021. The pathway has not changed dramatically in 2026, though digital home sleep apnoea testing has expanded and the SURMOUNT-OSA tirzepatide story sits awkwardly outside it.

The current standard-of-care sequence is broadly:

• Screening. Epworth Sleepiness Scale (ESS), STOP-BANG, and clinical suspicion in patients with loud habitual snoring, witnessed apnoeas, BMI > 35, hypertension resistant to two agents, atrial fibrillation, type 2 diabetes, or unexplained daytime cognitive symptoms.
• Diagnosis. Home sleep apnoea test (HSAT) with overnight oximetry and respiratory effort sensing for most patients; in-laboratory polysomnography (PSG) for complex cases, suspected central apnoea, paediatric patients or where HSAT is non-diagnostic.
• Severity grading. AHI 5–14.9 (mild), 15–29.9 (moderate), ≥ 30 (severe). Oxygen desaturation index (ODI) and T90 (time spent below 90% SpO2) provide additional risk stratification.
• CPAP. First-line for moderate-to-severe disease, and for mild disease with significant symptoms or comorbidity. Auto-titrating CPAP (APAP) is now standard.
• Mandibular advancement device (MAD). First-line option for mild-to-moderate disease, mild disease without significant comorbidity, or CPAP intolerance. Custom-made devices outperform thermoplastic boil-and-bite designs.
• Weight loss. Diet, exercise, structured Tier 3 services, bariatric surgery and — increasingly in 2026 — pharmacotherapy with semaglutide (Wegovy under NICE TA875) or tirzepatide (Mounjaro under NICE TA1026 for obesity).
• Positional therapy. For patients with primarily supine-dominant disease.
• Upper airway surgery. Tonsillectomy, uvulopalatopharyngoplasty (UPPP), maxillomandibular advancement (MMA), nasal surgery and bariatric surgery as appropriate.
• Hypoglossal nerve stimulation (Inspire). Under NICE IPG598 (2017), available under special arrangements for moderate-to-severe OSA in adults intolerant of CPAP. UCLH became one of the first NHS centres to offer this technology; volume remains modest as of 2026.

None of these interventions specifically targets the cognitive sequelae once they have developed. The implicit assumption of NG202 is that treating the breathing disorder will protect or restore cognition over time. The May 2025 UC Irvine REM-hypoxemia data, the 2025 CPAP-and-cognition pilot studies and the 36.92% cognitive-impairment prevalence figure suggest that for some patients — particularly those with long-undiagnosed disease, severe T90, or established hippocampal atrophy — that assumption is incomplete. That residual-deficit population is the population a future trial of a synaptogenic agent in OSA would target.

The OSA Comorbidity Cluster: Stroke, Diabetes, Long COVID and Mental Health

OSA rarely travels alone. Anyone presenting with OSA should be screened for, and considered against the backdrop of, a now well-characterised comorbidity cluster:

• Hypertension — particularly resistant hypertension requiring three or more antihypertensive agents.
• Atrial fibrillation. OSA roughly doubles the risk of incident AF, and treating OSA reduces post-cardioversion recurrence rates.
• Type 2 diabetes and prediabetes. OSA worsens HbA1c independent of weight; treating OSA improves glycaemic control modestly. The cognitive overlap is covered in the Dihexa diabetes brain fog review.
• Stroke. OSA increases stroke risk approximately threefold and is also a common consequence of stroke. The post-stroke OSA picture and its implications for cognitive recovery are covered in the Dihexa post-stroke recovery review.
• Heart failure with preserved ejection fraction (HFpEF). OSA contributes to and exacerbates HFpEF; the link with hypoxic encephalopathy compounds cognitive risk.
• Long COVID and post-COVID-19 syndrome. An emerging pattern: previously undiagnosed mild OSA in patients with Long COVID brain fog is being identified at scale in 2025–2026 NHS Long COVID services. See the Long COVID brain fog review.
• Depression and anxiety. OSA roughly doubles risk of major depressive disorder. The bidirectional relationship is covered in the depression review and the anxiety review.
• Erectile dysfunction. Mediated by autonomic dysregulation and endothelial dysfunction; often the presenting complaint that prompts OSA screening in men.
• Refractory migraine and morning headache. Hypoxia-driven nocturnal headache is one of the most cited but under-diagnosed OSA presentations.
• Menopause. OSA prevalence rises sharply at menopause; the cognitive overlap with vasomotor symptoms is covered in the menopause brain fog review.
• ADHD. Particularly in adults, untreated OSA can mimic or amplify inattention and executive dysfunction — the differential and the prefrontal overlap are covered in the Dihexa for ADHD review.

The cognitive picture in any individual OSA patient is rarely "pure". By the time most patients present, they have some combination of vascular small-vessel disease, metabolic brain fog, mood comorbidity and undiagnosed sleep-stage-specific hypoxic injury layered on top of each other. This is why treating OSA "fully" sometimes does not fully resolve the cognitive symptoms — and why the residual-deficit population in OSA looks neurobiologically similar to the residual-deficit populations in stroke, post-CABG cognitive impairment, and chemotherapy-induced cognitive impairment.

CPAP and Cognitive Recovery: What the 2025 Trials Are Showing

The simplest question patients ask their sleep clinicians is: if I use my CPAP, will my brain fog go away? The 2025 evidence base, while still maturing, supports a partial yes with important caveats.

• The 2025 Neurology International pilot study of 21 adults (median age 77) with mild cognitive impairment and Alzheimer's biomarkers found that 12 months of CPAP use was associated with attenuated cognitive decline in those with moderate-to-severe OSA. CPAP users showed significantly slower global decline and improvements in overall cognition compared with non-users.
• The Sunnybrook multi-centre CPAP RCT (NCT06773416) started in February 2025 with primary completion in September 2027 will provide the highest-quality data yet on whether 4 months of web-based sleep education plus CPAP improves cognition, reduces Alzheimer's-associated proteins, prevents cerebrovascular damage and reduces neuroinflammation more than education alone in MCI patients with untreated OSA.
• The CRESCENDO trial (NCT06983769) from National University Hospital Singapore (January 2026–December 2030) compares CPAP and MAD head-to-head in cognitively impaired patients — the first major RCT to compare devices in this specific population.
• The Brain Fog in Sleep Apnea trial (NCT06664450) is among the first to use "brain fog" itself as a primary cognitive endpoint rather than relying on conventional MoCA / MMSE composites.

The clinical picture in 2026 looks like this: in patients with moderate-to-severe OSA, consistent CPAP use (≥ 4 hours per night, ≥ 70% of nights) likely attenuates cognitive decline, modestly improves attention and executive function, and reduces stroke risk. It does not, however, reliably reverse established hippocampal atrophy or white matter hyperintensities, and a substantial subgroup of long-undiagnosed patients have persistent post-CPAP cognitive symptoms that no single intervention currently resolves. That post-CPAP residual-deficit population is exactly where any future synaptogenic adjunct would have to demonstrate value — and where no Dihexa data exists.

Dihexa vs Licensed and Adjunctive Approaches

To place Dihexa in context, it is worth comparing its theoretical action against the interventions most UK patients with OSA brain fog actually receive.

CPAP and Auto-CPAP

The mechanical gold standard. Splints the upper airway open, eliminates apnoeas and hypopnoeas, restores normal oxygen saturation and largely restores sleep architecture in adherent patients. Adherence is the headline limitation: even in optimistic real-world cohorts, only 50–70% of patients are using CPAP ≥ 4 hours per night at 12 months. CPAP does not have a direct neurorestorative mechanism — the cognitive benefit is upstream of the brain.

Compared with Dihexa: CPAP fixes the airway; Dihexa (theoretically) addresses the downstream brain damage. Mechanistically complementary, not substitutable. CPAP has decades of human data; Dihexa has none in OSA.

Mandibular Advancement Devices (MAD)

Custom-fit dental devices that hold the lower jaw forward, enlarging the retroglossal airway. First-line for mild-to-moderate disease and for CPAP-intolerant patients. Adherence is typically higher than CPAP, though efficacy in severe disease is lower.

Tirzepatide (Zepbound / Mounjaro)

The 2024 inflection point. Reduces AHI by 25–29 events/hour in adults with obesity through 18–20% weight loss. Indicated in the US for OSA with obesity, indicated in the UK for obesity (NICE TA1026) and type 2 diabetes. Brain-health implications covered in the Dihexa GLP-1 review.

Compared with Dihexa: Completely different mechanism (GLP-1/GIP agonism vs HGF/c-Met). Tirzepatide treats the underlying obesity driver of OSA in many patients; Dihexa would theoretically address cognitive residua. No interaction data.

Hypoglossal Nerve Stimulation (Inspire)

Implantable device that stimulates the hypoglossal nerve in synchrony with inspiration, advancing the tongue and opening the upper airway. NICE IPG598 (2017) permits use under special arrangements; uptake remains modest in 2026, with UCLH among the lead NHS centres. For CPAP-intolerant patients with moderate-to-severe OSA who meet specific anatomic criteria.

Modafinil and Solriamfetol for Residual Sleepiness

Wake-promoting agents licensed for residual excessive daytime sleepiness in adequately treated OSA. They mask sleepiness symptomatically without addressing the underlying breathing disorder or cognitive residua, and are not currently considered cognitive enhancers in OSA.

Where Dihexa sits in OSA

Dihexa is in a separate mechanistic category: a direct synaptogenic agent acting on HGF/c-Met, not on the airway, not on body weight, not on ventilatory drive, not on wakefulness. It is the only compound in the list with no human OSA or post-OSA cognitive data whatsoever. The position is not "Dihexa is an alternative to CPAP" or "Dihexa is an alternative to Zepbound"; the position is "Dihexa might one day be investigated as an adjunct for the residual cognitive deficit population if the mechanism translates from cell culture to humans".

The UK 2026 OSA Crisis: Why People Are Looking

The interest in unlicensed alternatives for OSA brain fog is best understood against the actual UK landscape in 2026. The Sleep Apnoea Trust Association (SATA), Asthma + Lung UK and the British Sleep Society have all published warnings about the scale of underdiagnosis and the inadequacy of NHS sleep services to meet projected demand:

• Around 13 million UK adults are estimated to have OSA, with overall prevalence around 19.5% and an estimated 80–85% remaining undiagnosed [1][10].
• In England, prevalence among adults aged 40–64 is approximately 17.8%; some occupational subgroups (HGV drivers, shift workers, construction) show prevalence above 30% [11].
• NHS sleep diagnostic waiting times are observing a monthly increase of roughly 2.5%; many trusts report waits in excess of 52 weeks for polysomnography or even home sleep apnoea testing [12].
• Individuals with untreated moderate-to-severe OSA are up to four times more likely to have a heart attack and up to three times more likely to have a stroke.
• Untreated OSA costs UK employers an estimated several billion pounds annually in lost productivity, sick leave and accidents.

The result is a population of people who: have been on a sleep clinic waiting list for over a year, are using CPAP but still have brain fog, cannot tolerate CPAP and are awaiting MAD fitting, or have stopped CPAP for psychological reasons and are looking for "the next thing". That curiosity drives traffic to research-chemical pages like this one. It is also the reason this article is written the way it is — not as a sales pitch, but as a careful evidence summary that takes the underlying biology seriously without overclaiming.

The same audience pressures are visible in adjacent conditions. The 2.7 million-strong NHS adult ADHD waiting list is the subject of the dedicated Dihexa for ADHD review; the NHS post-COVID-19 syndrome service backlog is the context for the Long COVID brain fog piece; the Women's Health Strategy and HRT prescribing patterns shape the menopause brain fog review. OSA sits at the intersection of all three — many ADHD-symptomatic adults have undiagnosed OSA, many Long COVID brain-fog cases have undiagnosed OSA, and OSA prevalence rises sharply through the menopausal transition.

Specific Risks of Self-Experimenting with Dihexa in OSA

Even among the conditions covered on this site, OSA carries specific risks that make self-experimentation particularly inadvisable:

• Masking diagnosis. A patient who attributes their brain fog to OSA but treats it with a peptide instead of a sleep study may continue accumulating cardiovascular, cerebrovascular and metabolic risk for years.
• Masking residual disease on CPAP. Many patients have residual events on CPAP that show up only on download data — a peptide that subjectively improves daytime function could delay an essential titration change.
• Cardiovascular interaction risk. OSA patients are disproportionately on multiple antihypertensives, antiarrhythmics, anticoagulants and lipid-lowering drugs. The interaction profile of Dihexa with these agents is uncharacterised.
• Comorbid diabetes. Dihexa is not known to affect glycaemic control directly, but the underlying angiotensin-IV-derived mechanism and HGF/c-Met biology overlaps with insulin signalling in ways that have not been clinically characterised.
• Theoretical oncogenicity from sustained c-Met activation. Sustained c-Met signalling is a known driver of tumour growth in several solid cancers. Long-term safety in OSA patients (often older, often metabolically unwell, often with pre-existing tumour risk) has not been studied.
• Sleep destabilisation. Anecdotal reports include vivid dreams, fragmented REM and paradoxical insomnia in a subset of users — potentially worsening an already disrupted sleep architecture.
• Mood destabilisation. Covered in the depression and bipolar disorder reviews. A patient with untreated OSA, comorbid depression or burnout, and a research peptide cycle is in a poorly characterised pharmacological state.

A Reasonable Non-Peptide Stack for OSA Brain Fog

Anyone reading this page for solutions, rather than for the science, should consider that the evidence-based interventions for OSA brain fog — in order of effect size — are well-defined:

• Confirm the diagnosis. If you snore loudly, wake unrefreshed, have witnessed apnoeas, severe daytime sleepiness, morning headaches, hypertension or refractory cognitive symptoms, ask your GP for a sleep service referral or self-pay for a home sleep test.
• Use CPAP consistently if prescribed. The dose-response relationship between CPAP adherence and outcome is steep. ≥ 4 hours / ≥ 70% of nights is the standard benchmark; more is better.
• Address weight where appropriate. Diet, exercise, structured Tier 3 services and where indicated tirzepatide (Mounjaro) or semaglutide (Wegovy) via NHS or private routes. See the Dihexa GLP-1 review.
• Sleep hygiene and CBT-I. Often missed; treating coexisting insomnia frequently reduces cognitive symptoms independently.
• Position therapy. Side-sleeping reduces supine-dominant OSA severity significantly in some patients.
• Reduce alcohol within 4 hours of bed. Alcohol potentiates upper airway muscle relaxation.
• Treat nasal obstruction. Persistent nasal congestion materially affects CPAP comfort and OSA severity.
• Address comorbid mood, anxiety and pain. Treating depression and chronic pain improves perceived cognitive function in OSA patients independent of AHI.
• Underlying medical contributors. Hypothyroidism, acromegaly, perimenopause, anaemia, B12 deficiency, restless legs syndrome and chronic pain can all present as or worsen OSA-like cognitive symptoms.

An unscheduled research chemical with no human OSA data is not a substitute for any of the above. It might one day, in a structured trial of patients with adequately treated OSA but persistent cognitive symptoms, become an adjunct that supports residual recovery. It is not currently that.

The Bottom Line in 2026

Dihexa fits the modern, hypoxia-and-circuit-based model of OSA brain fog with reasonable mechanistic precision. Chronic intermittent hypoxia and sleep fragmentation are now understood to drive REM-stage-specific CA1 hippocampal volume loss, white matter hyperintensities, temporal lobe atrophy, BDNF reduction, beta-amyloid accumulation and small-vessel cerebrovascular damage. The CA1 region specifically requires BDNF-driven plasticity for memory consolidation, and the same intracellular pathways (PI3K/Akt, mTOR, ERK) that BDNF engages are activated by HGF/c-Met — the receptor system Dihexa was engineered to target. HGF/c-Met biology is also relevant to the vascular and angiogenic component of OSA injury. A BBB-penetrant compound that drives synaptogenesis and supports endothelial repair in those circuits is, at the level of theory, the kind of agent OSA cognitive neuroscience has been hoping to find.

Theory is where the evidence stops. There is no human trial of Dihexa in obstructive sleep apnea, central sleep apnea, chronic intermittent hypoxia, post-CPAP cognitive impairment or any sleep-related cognitive condition. There is no published case series, no biomarker data, no controlled comparisons against modafinil or solriamfetol, no interaction studies with antihypertensives or with tirzepatide, and no long-term safety data in chronically dosed adults with sleep-disordered breathing. The closest clinical relative fosgonimeton was developed for Alzheimer's disease, not OSA, and its 2024 Phase 3 readout did not meet its primary endpoint, with the SHAPE PDD/DLB programme subsequently discontinued. Community self-reports are heterogeneous and do not rise above the quality threshold of hypothesis-generation.

For anyone with significant OSA brain fog reading this in 2026, the honest reading of the evidence is: Dihexa is biologically coherent within the synaptic-plasticity-and-vascular-repair model of OSA cognitive impairment, but clinically unverified. It cannot currently be recommended for sleep apnea brain fog. Evidence-based interventions — confirmation of OSA diagnosis with a sleep study, consistent CPAP or MAD use, weight loss including tirzepatide where appropriate, treatment of comorbid mood, hypertension, diabetes and nasal disease, and sleep hygiene — are the routes with measurable human effect sizes. If you are curious about Dihexa as a compound, the what is Dihexa, benefits and research and studies pages are the right starting points. If your brain fog is severe and unexplained, please speak to a GP about an NHS sleep service referral, contact the Sleep Apnoea Trust Association for patient support, or use the Epworth Sleepiness Scale and STOP-BANG questionnaires as a starting point.

Frequently Asked Questions

Selected References & Outbound Sources

1. Sleep Apnoea Trust Association. UK patient and professional resource: prevalence, screening and signposting. Accessed May 2026.
2. Asthma + Lung UK. Obstructive sleep apnoea (OSA): UK patient information. Accessed May 2026.
3. Sun H, Wei L, Hua D et al. Prevalence of cognitive impairment among adults with obstructive sleep apnea: a systematic review and meta-analysis. 2025.
4. Mander BA, Dave A, Lui KK et al. Hypoxemia during rapid eye movement sleep mediates memory impairment in older adults at risk for dementia via CA1 hippocampal volume loss. 2024–2025.
5. University of California, Irvine News. Study links REM sleep apnea to brain changes, memory loss in older adults. May 2025.
6. Malhotra A, Grunstein RR, Fietze I et al. Tirzepatide for the Treatment of Obstructive Sleep Apnea and Obesity (SURMOUNT-OSA). New England Journal of Medicine, June 2024.
7. US Food and Drug Administration. FDA Approves First Medication for Obstructive Sleep Apnea. 20 December 2024.
8. Examining the association between sleep apnea and total hippocampal volumes in cognitive impairment. PMC12022498, 2025.
9. Singh M et al. Association of Hypoxemia Due to Obstructive Sleep Apnea With White Matter Hyperintensities and Temporal Lobe Changes in Older Adults. Neurology, 2025.
10. NICE. Obstructive sleep apnoea/hypopnoea syndrome and obesity hypoventilation syndrome in over 16s (NG202). August 2021.
11. Watson NF et al. Estimated prevalence of obstructive sleep apnea by occupation and industry in England. SLEEP Advances, 2024.
12. EMJ Reviews. Untreated Sleep Apnoea: A Hidden Multi-Billion Hit to Productivity. 2024.
13. NICE. Hypoglossal nerve stimulation for moderate to severe obstructive sleep apnoea (IPG598). 2017.
14. Wright JW, Harding JW. The brain hepatocyte growth factor/c-Met receptor system: a new target for the treatment of Alzheimer's disease. J Alzheimers Dis, 2015.
15. Benoist CC et al. The procognitive and synaptogenic effects of angiotensin IV-derived peptides are dependent on activation of the hepatocyte growth factor/c-Met system. J Pharmacol Exp Ther, 2014.
16. Treatment of obstructive sleep apnea using CPAP and cognitive impairment (medRxiv preprint, August 2025). medRxiv 2025.08.07.25333263.
17. ClinicalTrials.gov. NCT06773416 - Treating Sleep Apnea to Improve Cognitive Function (Sunnybrook RCT). February 2025 start.
18. ClinicalTrials.gov. NCT06983769 - CRESCENDO: CPAP vs MAD for OSA in Patients With Cognitive Impairment. January 2026 start.
19. ClinicalTrials.gov. NCT06664450 - Brain Fog in Sleep Apnea.