Peptides for Brain Fog: What the Long-COVID Research Shows

Brain fog used to be a vague complaint. After the pandemic it became a measurable neurological pattern, and the search for compounds that can blunt it has pulled a handful of cognitive peptides into the spotlight. This article walks through what the research actually shows about peptides for brain fog, why post-viral cognitive dysfunction is biologically distinct from ordinary fatigue, and where the evidence is strong, thin, or simply absent.

Key takeaways#

• Long-COVID brain fog is driven by measurable neuroinflammation, blood-brain barrier disruption, and microglial activation, not by perceived laziness or stress alone.
• Long COVID occurs in at least 10% of SARS-CoV-2 infections (Davis et al., 2023), most of them after mild illness, and brain fog is among its most common symptoms.
• Semax and Selank are the two cognitive peptides with the most published human data, both originally developed in Russia for stroke and anxiety indications.
• Dihexa shows extraordinary preclinical potency at driving synaptogenesis but has no human trials in post-viral cognitive dysfunction.
• No peptide is approved anywhere for long-COVID brain fog, and no drug is FDA-approved for it either; the strongest peptide cases remain mechanistic rather than outcome-based.

Brain fog is a real neuroinflammatory signal, not a vague complaint#

The first useful reframe is biological. Cognitive dysfunction represents one of the most persistent and disabling features of Long COVID, yet its molecular underpinnings remain incompletely understood. Key processes include persistent neuroinflammation, blood-brain barrier disruption, endothelial dysfunction, immune dysregulation, and neuroendocrine imbalance.

Microglial activation and cytokine release (e.g., IL-6, TNF-alpha) promote synaptic dysfunction and neuronal injury, while activation of inflammasomes such as NLRP3 amplifies CNS inflammation.

The clinical picture that follows is consistent across cohorts. This constellation of cognitive symptoms is commonly described by patients as "brain fog" and encompasses memory deficiencies, reduced attention span, slowed information processing, and executive dysfunction. Crucially, the deficits are not gated by infection severity. Notably, these impairments have been reported even in individuals with mild or asymptomatic acute infection and may persist for months.

Imaging data backs the symptom reports. Magnetic resonance images and objective cognitive tests recorded before and after SARS-CoV-2 infection have revealed reduced grey matter in the orbitofrontal cortex and parahippocampal gyrus and reduced global brain size correlated with impaired cognitive performance in those who had been infected. There is a partial recovery curve. In a US Veterans Affairs cohort of 138,818 people after SARS-CoV-2 infection and 5,985,227 non-infected controls, the excess risk of 69% of 80 prespecified post-COVID sequelae was no longer statistically significant at 2 years in people who had not been hospitalized (Bowe et al., 2023); 31% remained elevated. That is the natural-history baseline against which any intervention has to be judged.

A 2024 JAMA Network Open randomized trial of lithium aspartate for long-COVID cognitive dysfunction is a useful calibration point. In a randomized clinical trial including 52 participants, lithium aspartate, 10 to 15 mg/d, for 3 weeks provided no significant improvements to fatigue or cognitive dysfunction scores. Even more sobering, the NIH-funded RECOVER-NEURO trial reported that "None of our rehabilitation approaches to treatment for cognitive long COVID proved to be effective. Detecting a benefit for one of the interventions was made more difficult because participants in all 5 treatment arms showed modest improvements," the lead author told the RECOVER initiative. The therapeutic field is genuinely open.

Why peptides became a candidate class for cognitive recovery#

Most cognitive peptides under study share a common thread: they target the neurotrophic and neuroinflammatory axes that the long-COVID literature keeps implicating. They are short amino acid sequences, typically delivered intranasally, that have been shown to modulate BDNF, NGF, and microglial signaling in animal models. None of them is a stimulant in the classical sense.

The mechanistic logic is straightforward. If brain fog is downstream of microglial overactivation and reduced synaptic plasticity, then compounds that dampen neuroinflammation while supporting brain-derived neurotrophic factor signaling are at least biologically plausible candidates. That is a hypothesis, not a clinical guarantee, and the gap between preclinical mechanism and verified human outcomes is the most important thing for readers to keep in view.

Semax: the most clinically studied cognitive peptide#

Semax is a synthetic ACTH(4-10) analog originally developed at the Russian Academy of Sciences. Semax is used in Russia for stroke, dyscirculatory encephalopathy, Parkinson's, ocular nerve atrophy, and for newborns with neurological deficits. That regulatory history is unusual: most "research peptides" lack any approved indication anywhere.

The mechanistic file is reasonably mature. Research suggests Semax upregulates BDNF and NGF and modulates dopaminergic and serotonergic tone. The peptide Semax, which exhibits neuroprotective and nootropic properties, is a long-acting memory enhancer. In a 2025 transgenic Alzheimer model, intranasal administration of Semax or its derivative Heptapeptide improves the cognitive function in the mouse model of Alzheimer's disease. Both Semax and Heptapeptide significantly reduce the amyloid load in the animal brain. The full paper is available through PubMed Central.

Human data, while limited in English-language journals, is not absent. In a small double-blind study of 16 male operators doing shift work (Kaplan et al., Neuroscience Research Communications, 1996), 8 received intranasal Semax at 1 mg (about 16 mcg/kg) and 8 received placebo before testing. On the next morning's test, the Semax group made 71% correct responses against 41% for placebo, and the authors reported a beneficial effect lasting 20 to 24 hours. It is one small study in healthy workers, not in people with brain fog. The honest caveats: most clinical trials were conducted within the Russian healthcare system and published in Russian journals such as Zhurnal Nevrologii i Psikhiatrii. This limits Western access and independent evaluation. The research quality varies: some studies meet international standards with proper controls and blinding, while others lack the methodological rigor expected by FDA or EMA.

No published trial has tested Semax specifically against long-COVID brain fog. The translation from "fatigued shift workers" or "post-stroke cognitive deficit" to "post-viral cognitive dysfunction" is a reasonable mechanistic bet, not a verified outcome.

FDA's own July 2026 review of Semax, written for its compounding advisory committee, lists the open safety questions plainly (FDA, 2026). FDA found no human pharmacokinetic studies by any route. In all but one of the clinical references, adverse events were not discussed. In mice, Semax potentiated amphetamine-induced dopamine release in the striatum, which FDA called concerning, and possible anti-thrombotic properties raise a bleeding question. Semax was also nominated for ADHD, and FDA did not evaluate that use because it found no supporting literature.

Selank: the anxiolytic with a neuroinflammation profile#

Selank is a heptapeptide derived from the immunoglobulin G fragment tuftsin, with a C-terminal Pro-Gly-Pro extension that extends its half-life. The interesting feature for brain fog discussions is its dual profile: anxiolytic plus immunomodulatory.

Selank is a synthetic analog of taftsin extended with a C-terminal Pro-Gly-Pro tripeptide to increase metabolic stability and duration of action. Selank has a pronounced anxiolytic and nootropic effect and is highly effective in the treatment of anxiety and anxiety-asthenic disorders, leading to activation and optimization of cognitive brain function, learning processes, and memory. Along with these functions, Selank also has pronounced immunotropic activity and is able to induce the secretion of interferons.

The cytokine angle is what makes it a candidate worth discussing in the brain-fog context. Preclinical studies report that Selank modulates IL-6 expression and the balance of T helper cell cytokines, cytokine families also implicated in long-COVID neuroinflammation; this has not been tested in people with brain fog. A 2019 study in Bulletin of Experimental Biology and Medicine found that in the object recognition test, Selank produced a cognitive-stimulating effect in 9-month rats not exposed to ethanol and prevented the formation of ethanol-induced memory and attention disturbances developing during alcohol withdrawal. In ex vivo experiments, Selank prevented ethanol-induced increase in BDNF content in the hippocampus and frontal cortex.

Selank and Semax are often discussed as complementary. They hit different receptor systems and are commonly paired in protocols: Semax skews cognition-forward, Selank skews calm-forward. Preliminary evidence suggests amplified effects when used together, but no controlled trial has formally tested the combination against placebo for post-viral cognitive symptoms.

Dihexa: extraordinary potency, almost no human data#

Dihexa is the wildcard. It is a synthetic hexapeptide derived from angiotensin IV, developed at Washington State University. In preclinical models from its originating lab, it induced hippocampal spinogenesis and synaptogenesis at picomolar concentrations and reversed cognitive deficits in scopolamine-treated and aged rats (McCoy et al., 2013). Claims that it is "10 million times more potent than BDNF" circulate widely; treat them with caution, because the same lab's 2014 paper that established its HGF/c-Met mechanism was retracted in 2025 after a Washington State University investigation found falsified or fabricated data (JPET retraction notice).

The mechanism is structurally distinct. Unlike Semax, which acts primarily through BDNF upregulation and dopaminergic enhancement, Dihexa was proposed to work by potentiating the HGF/c-Met signaling axis (hepatocyte growth factor and its receptor), which drives dendritic spine formation and synaptogenesis independent of BDNF. With the 2014 mechanism paper retracted, that mechanism is an open question rather than a settled fact.

The problem is that nearly all of the evidence is preclinical. There are no published randomized trials in humans, and certainly none in long-COVID populations. The same potency that makes it interesting is also a reason for caution: a compound that aggressively drives synaptogenesis and activates growth-factor pathways has unresolved questions around oncogenic risk and maladaptive wiring that simply have not been answered yet. For a post-viral cognitive recovery scenario, this is at the speculative frontier rather than the practical toolkit.

What a careful approach to peptides for brain fog looks like#

A few principles fall out of the literature.

First, do not skip the workup. Brain fog symptoms overlap with thyroid dysfunction, anemia, sleep apnea, depression, and ME/CFS. Studies have shown the importance of comprehensive assessment before any cognitive intervention. Any peptide protocol stacked on top of an undiagnosed thyroid issue or untreated sleep apnea is fighting the wrong battle.

Second, weight the time-course expectations correctly. User reports describe memory and executive-function changes as cumulative over one to two weeks of consistent use; no controlled trial in brain fog has measured that time course. If a user expects a stimulant-like rush from a peptide that works through neurotrophic adaptation, they will draw the wrong conclusion at day three.

Third, plan around the natural recovery curve. The two-year improvement signal in non-hospitalized long-COVID cohorts means that any uncontrolled n-of-1 trial will tend to attribute spontaneous recovery to whatever the user happened to be taking. Track baselines. Use the same cognitive task each week. The Klarovel peptide calculator and the protocol templates in how it works are built around exactly this kind of structured tracking.

Fourth, and most importantly, be honest about the evidence ceiling. The most rigorous public synthesis to date, the RECOVER-NEURO arm summarised by the NIH RECOVER initiative, shows that even well-funded behavioural and device interventions struggled to outperform comparison arms. The peptide evidence base for long-COVID brain fog specifically is preliminary. Anyone marketing peptides as a verified solution is overselling.

The honest position on peptides and brain fog#

The biology lines up. Post-viral brain fog is a neuroinflammatory and neurotrophic problem, and cognitive peptides act on neuroinflammatory and neurotrophic pathways. That makes the category worth taking seriously. It does not make any individual peptide a verified answer for long-COVID cognitive dysfunction. Semax has the deepest clinical file, Selank has the most relevant cytokine profile, and Dihexa has the most striking preclinical numbers and the thinnest human evidence. Klarovel's role is to keep that distinction clear, build protocols from a full health profile, and refuse to oversell what the data does not yet support. If that posture matches how a reader wants to approach their own cognitive recovery, the next step is the baseline workup with your clinician, then tracking symptoms and any changes together over time. A Klarovel account gives you a structured place to record that tracking and bring it to those appointments.