Cognitive decline after long COVID: what the evidence shows (2024–2026)

How common is cognitive decline after COVID-19?

The 2024 systematic review cited World Health Organization figures of more than 774,075,242 reported COVID-19 cases and noted that approximately 10–35% of survivors experience persistent symptoms.^[1] Under the NICE definition, long COVID describes symptoms that persist beyond 12 weeks (3 months), and roughly one in three people with COVID-19 was diagnosed with neurological symptoms within six months.^[1] The focus, "brain fog," was a cluster of confusion, short-term memory loss, and difficulty concentrating, and a 2022 Global Burden of Disease analysis in JAMA confirmed that a persistent cognitive symptom cluster affected a large share of people after symptomatic COVID-19, a population-scale problem rather than a rare complication.^[3]

What interventions have been tested for brain fog?

The most comprehensive 2024–2026 map is a PRISMA systematic review by Gorenshtein and colleagues, pre-registered on PROSPERO (CRD42024502977) and published in Neurological Sciences.^[1] Searching five databases for 2020–2023 articles, the reviewers screened 5,770 records down to 287 eligible articles and included 17 studies covering 806 patients across 13 countries.^[1]

The interventions sorted into four families. Noninvasive brain stimulation (6 studies) all reported symptom improvements; hyperbaric oxygen therapy (3 studies) improved brain perfusion on MRI and reduced symptoms in all three, in one randomized trial, 73 patients assigned to 40 daily oxygen sessions or control showed a significant gain in global cognition, attention, and executive function.^[4] The PEA-LUT compound (2 studies) improved cognitive impairment in both, while rehabilitation (6 studies) was the most mixed, 4 reported gains and 2 reported none.^[1]

The review reported that rehabilitation carried the safest profile but that its efficacy remained uncertain given the split results.^[1] A full study-by-study breakdown is on the companion page, Review: 17 studies tested interventions for long-COVID brain fog.

Does cognitive training help long-COVID brain fog?

Within the rehabilitation family, the strongest individual signal came from personalized computerized cognitive training. A 2023 before-and-after feasibility study by Duñabeitia and colleagues (ClinicalTrials.gov NCT05571852) enrolled adults reporting brain fog at least three months after infection; of 262 eligible volunteers, 73 (mean age 46.1; 66 women) completed a minimum of 10 sessions across 8 weeks of home-based training tailored to each person's baseline cognitive profile.^[2]

On the Cognitive Assessment Battery, 17 tests normed on 1,282,242 healthy test-takers, participants improved on all five domains by +14 in attention, +18 in memory, +18 in coordination, +17 in perception, and +11 in reasoning percentile points (all p < 0.001), and gains scaled with minutes of training (interaction p = 0.012).^[2] Participants completed a mean of 51 sessions and 435 minutes of training, showing people with brain fog could run an unsupervised program. Full methods and the per-domain table are on the companion page, 8 weeks of personalized cognitive training in long-COVID brain fog.

It was, to its authors' knowledge, the first study to extend to long COVID a research line earlier meta-analyses had built in stroke and Parkinson's disease.^[2] Separately, the largest rehabilitation cohort in the review (208 patients, Barrow Neurological Institute Screen) saw the share below the cut-off fall from 54% to 33% at follow-up.^[5]

Why did personalized programs outperform generic ones?

The two rehabilitation studies that showed no improvement both used a one-size-fits-all multimodal protocol; the authors attributed the null results to a lack of personalization and to low adherence, in one telerehabilitation trial only 25% of participants used the mobile application meaningfully. The studies that did show gains, including the feasibility study, tailored task selection and difficulty to each person's measured profile.^[1][2] Across the wider literature, the 2024 Lancet standing Commission on dementia treated structured cognitive activity as one part of a multidomain brain-health approach, alongside fourteen modifiable risk factors associated with an estimated 45% of dementia cases worldwide, and the long-COVID evidence fits that framing, in which training is a low-risk supportive option, not a disease treatment.^[6]

How does the evidence compare across brain-training programs?

For long-COVID brain fog specifically, no head-to-head randomized trial has compared the major commercial cognitive-training programs, and as of this writing no consumer program has a published randomized controlled trial demonstrating that it treats long COVID. The 2024 review's strongest-rated families were noninvasive brain stimulation and hyperbaric oxygen, both clinical, not consumer, interventions, and the reviewers could not rank any intervention as best.^[1] The honest comparison is therefore not "which program works best" but "which programs have any peer-reviewed cognitive-outcome data at all", and the feasibility evidence for personalized training is among the few published signals, with the caveat that it lacked a control group.^[2]

What the evidence does NOT show

No trial shows cognitive training cures or prevents long COVID. The studies here measured cognitive-performance outcomes over weeks to months; none demonstrated that any program treats the underlying condition or prevents future decline.^[1][2]

The strongest training result came from a design that cannot prove cause. The 73-participant feasibility study had no control group, so practice effects, natural recovery, and placebo response cannot be separated from the training effect; its authors called for randomized controlled trials, and the sample was self-selected and 90% female.^[2]

The overall evidence base is small, heterogeneous, and mixed. Only 17 studies met the criteria, several were case reports with samples as small as one patient, no effect sizes were pooled, and two of the six rehabilitation studies showed no benefit. The reviewers cited "high heterogeneity … and the lack of meta-regression analysis" among their limitations and stressed that further research is necessary to establish efficacy.^[1]

What this means in practice

For someone navigating brain fog after COVID-19, the 2024–2026 evidence supports a measured reading: several low-risk interventions, including personalized cognitive training, were associated with cognitive gains in early studies, but the field is still too small to name a best treatment, and none establishes a cure. Decisions about persistent symptoms should be made with a qualified clinician.

See the underlying studies, the 17-study intervention review and the 8-week feasibility study, and related guides on healthy older adults, stroke, TBI and surgery, multiple sclerosis, chronic insomnia and memory, mobility and driving, and stress and memory. The hub is the cognitive-decline evidence guide; every study is in the research index.