For many people who have “recovered” from COVID-19, the illness lingers as brain fog: forgotten appointments, lost words mid-sentence, a mind that feels wrapped in cotton. A growing body of brain imaging research now suggests these symptoms reflect measurable changes in the brain, including reduced metabolism and loss of grey matter in key regions. In one flagship dataset, 785 UK Biobank participants were scanned before and after the pandemic, and 401 who became infected showed clear structural differences compared with uninfected peers.
The Key Studies Behind the Findings
The most closely watched evidence comes from a Flagship Biobank MRI study that followed 785 adults who had brain scans before COVID-19 existed, then again after the virus spread. In that cohort, 401 participants tested positive for SARS infections between scans and were compared with 384 matched controls who remained uninfected, with the second scan taking place an average of 141 days after diagnosis. Researchers reported a greater reduction in grey matter thickness and tissue contrast in orbitofrontal and related regions in the infected group, indicating that even relatively mild cases could be associated with detectable tissue loss.
Those structural findings sit on top of a broader imaging infrastructure that makes this kind of work possible. The Primary Biobank COVID Research Analysis Platform describes how imaging, biomarkers, health records and questionnaires are linked in a single dataset, allowing investigators to compare pre and post infection scans in a way that is rarely possible in neurology. That same infrastructure supports further analysis of tissue contrast changes and other subtle measures that might not show up on routine clinical scans but could still relate to cognition and long-term neurological risk.
Patterns of Brain Changes Observed
If the Biobank work maps structural loss, functional imaging highlights how the brain is working differently in people with long COVID. A Primary PET analysis compared 35 long COVID patients who reported persistent neurological complaints with 44 healthy controls, using voxel-wise techniques to examine metabolism across the whole brain. The researchers found hypometabolism in olfactory and orbitofrontal regions, including the olfactory gyrus, as well as in medial temporal structures, with extension into the thalamus and brainstem, a pattern that aligns strikingly with patients’ reports of smell loss, memory problems and fatigue.
Subsequent work has tried to translate that pattern into something clinicians can recognize in daily practice. A multicenter Primary PET FDG benchmark assessment focused on qualitative and consensus visual review of 18F-FDG PET scans in suspected neurological long COVID, explicitly checking how often the previously described hypometabolic signature appeared. By centering on visual assessment rather than only statistics, that study moves the field closer to operational criteria that nuclear medicine teams could use when they see post COVID patients with unexplained cognitive or sensory symptoms.
Evidence from MRI and Advanced Imaging
Beyond global volume loss, more granular MRI techniques are being tested as potential biomarkers of post COVID condition. A Primary RSNA MRI study prospectively compared 89 Post COVID condition (PCC) patients with 38 unimpaired post COVID controls, using microstructural MRI features to see whether individual-level classification was possible. The peer-reviewed analysis reported that certain combinations of microstructural measures could distinguish PCC from controls, supporting the idea that subtle white matter or cortical changes might help explain persistent symptoms even when conventional MRI appears normal.
Severity of the initial illness also seems to matter. In a multicenter high-field imaging project, a Primary Tesla MRI cohort of 179 recovered participants underwent 7 Tesla scans with detailed cognitive testing. Those who had been hospitalized with COVID showed smaller hippocampal volume and worse performance on memory and global cognition measures, including MoCA and Trail Making Test scores, compared with non-hospitalized participants. One of the lead researchers, quoted in the PubMed summary, linked these hippocampal findings to the kind of forgetfulness and slowed thinking that many patients describe months after infection.
What These Changes Mean for Recovery
The imaging changes do not just sit in isolation; they track closely with real-world symptoms such as brain fog, attention lapses and slowed processing. A mechanistic study using dynamic contrast-enhanced MRI found that long COVID patients with prominent cognitive complaints had evidence of blood brain barrier disruption, with both regional and whole brain leakage metrics elevated compared with expectations. That work, described in Primary COVID MRI BBB research, also tied the leakage to peripheral blood transcriptomics and in vitro endothelial assays, suggesting that vascular and inflammatory processes may underlie at least some of the cognitive sequelae.
Functional imaging hints that some of these abnormalities may improve over time, although the trajectory is far from settled. A Primary US PET FDG COVID Reports cohort followed 45 post COVID patients who underwent FDG-PET/CT, 15 of whom had pre infection scans under identical conditions, and compared them with oncology controls. The researchers observed focal hypometabolism across multiple lobes and the cerebellum that appeared to peak early after infection, with near complete recovery reported at later time points, which they interpreted as a sign that at least part of the metabolic disturbance can be reversible.
Broader Implications and Why It Matters
For clinicians and policymakers, these findings raise concerns about longer term neurological risk in a large population of previously infected people. The Biobank MRI analysis that documented greater reduction in grey matter thickness and tissue contrast in 401 infected participants compared with 384 controls suggests that even community cases might carry subtle but measurable brain effects. When combined with microstructural differences in PCC cohorts and hippocampal volume loss in more severe cases, a picture emerges of COVID as an illness that can touch multiple brain systems, not only the lungs.
Radiology groups have started to frame these changes as a reason for structured follow up rather than one-off reassurance. The Primary RSNA MRI PCC cohort has been cited in calls for routine cognitive screening and imaging-based monitoring in patients with persistent symptoms, arguing that microstructural MRI features could eventually guide rehabilitation or pharmacologic trials. Parallel reporting on young adults, such as a study highlighted by psychiatry-focused coverage of mild COVID cases, reinforces the concern that lasting brain changes are not limited to older or hospitalized groups.
Uncertainties and Future Research Needs
Despite the striking images, researchers are cautious about drawing straight lines from infection to long term brain disease. The multicenter Primary PET FDG COVID assessment explicitly frames its hypometabolic pattern as an association rather than proof of causation, and notes that individual variability and pre existing conditions could contribute. Many cohorts also lack pre and post scans for every participant, which makes the Biobank dataset unusual and leaves open questions about how much of the observed change is directly attributable to SARS infection versus other pandemic era stressors.
Recovery timelines are another open question. While the Primary US PET FDG COVID Reports cohort suggests near complete normalization of metabolism in many patients, other work, including mechanistic COVID MRI BBB disruption data, emphasizes ongoing leakage and inflammation in those with persistent brain fog. Reviews that synthesize this emerging evidence, such as reporting on the biological basis of long COVID brain fog and coverage by European health outlets, stress that full reversibility is not yet demonstrated and that large, long duration studies are needed.
What Advanced Imaging Is Revealing Beneath “Normal” Scans
Several teams are also probing what happens in patients whose standard MRI looks unremarkable. A diffusion microstructure imaging study described as Primary DMI MRI COVID PCR work examined inpatients with neurological symptoms around 29 days after a positive PCR test, and found a shift in volume from intra and extra axonal compartments toward a higher free water fraction. That pattern is consistent with microstructural disruption that conventional imaging might miss, hinting that some neurological complaints could reflect subtle white matter changes rather than overt lesions.
Other investigators are linking these advanced imaging markers to cellular mechanisms. Reporting on long COVID cohorts in which Primary PET COVID analysis is combined with laboratory assays points to persistent immune activation and endothelial dysfunction as plausible drivers of both hypometabolism and BBB leakage. Science communication outlets have highlighted these findings as part of a broader effort to explain why brain fog can persist even after respiratory symptoms fade, with pieces from science news sites and general science coverage bringing the technical imaging work to a wider audience.
Extending the Lens to Prenatal and Pediatric Exposure
Although most imaging data so far focus on adults, parallel work is beginning to examine what COVID exposure means for developing brains. Pediatric researchers have reported that prenatal infection can be associated with altered brain structure at birth, as described in Primary prenatal COVID brain imaging coverage. Follow up stories from child health organizations suggest that these newborn changes might track with later developmental differences, although the evidence is still early and heavily qualified.
Population health groups have also flagged concerns about in utero exposure and later mental health. A synthesis from academic epidemiologists, summarized by infectious disease researchers, links prenatal COVID exposure to brain changes alongside developmental delays, anxiety and depression in early childhood. While these findings sit somewhat apart from the adult PET and MRI data, they point in the same direction: COVID is not just a transient respiratory infection, but a virus with the capacity to reshape brain structure and function across the lifespan, in ways scientists are only beginning to map.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
