Summary: A precision neuroimaging study from the University of Turku challenges the widespread idea that chronic, widespread brain inflammation is the main cause of Long COVID. Using advanced PET and MRI techniques alongside blood biomarkers, researchers compared people with Long COVID to healthy controls and to patients with multiple sclerosis (MS). The results show no lasting, global increase in neuroinflammation or neurodegeneration in Long COVID compared with healthy individuals. Instead, inflammation appears to be time-dependent—more apparent in the months after infection and diminishing over time—while persistent symptoms relate more closely to localized cellular activity in emotion and stress-regulating brain regions.
Key findings:
- No sustained widespread neuroinflammation: Long COVID participants did not show higher long-term markers of brain inflammation or degeneration than healthy controls in this PET and MRI study.
- Sensitive imaging and biomarkers: The team used TSPO PET imaging ([11C]PK11195) to detect glial activation, 3T MRI to assess brain structure and white matter, and blood measures of neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP) as markers of neuronal and glial injury.
- Comparison with MS clarified the signal: MS patients, a group with known aggressive brain inflammation, showed higher white matter inflammatory activity than Long COVID participants. Long COVID participants’ PET signals more closely matched healthy controls than MS patients.
- Inflammation is time-dependent: Individuals scanned within 16 months of their initial SARS-CoV-2 infection had higher white matter inflammatory signals than those scanned after a longer interval, suggesting post-infectious neuroinflammation tends to subside with time.
- Localized limbic activation links to symptoms: Greater symptoms of depression, anxiety and reduced quality of life were associated with increased cellular activity in the hippocampus and amygdala—limbic areas that regulate memory, emotion and stress responses.
- Implications for treatment: Because large-scale inflammation appears to decline over time, the authors propose that long-term management should focus more on interventions that target stress processing and emotional regulation rather than relying exclusively on anti-inflammatory approaches.
Source: University of Turku
Long COVID has been suspected to involve ongoing brain inflammation after SARS-CoV-2 infection, which could explain symptoms such as fatigue, cognitive difficulties, anxiety and depression. However, direct, localized evidence for persistent neuroinflammation has been limited and sometimes inconsistent. The Turku team set out to examine these questions using high-resolution PET imaging, MRI and blood biomarkers in well-characterized participants.
“We did not observe evidence of widespread brain inflammation in patients with long COVID when compared to healthy controls,” says Professor Laura Airas, who led the study. Participants included 14 people with Long COVID (LC), 11 healthy controls (HC), and 13 patients with multiple sclerosis (MS). All participants underwent [11C]PK11195 TSPO PET scans that detect glial activation, high-resolution 3T MRI, and blood sampling for NfL and GFAP. Long COVID participants also completed neurological and mental health assessments.
Brain inflammation may be present early after infection
Neuropathological studies of severe acute COVID-19 have previously shown evidence of brain inflammation. In this imaging study, those scanned within 16 months of their initial infection had higher white matter TSPO availability than those with longer disease duration, indicating that post-COVID neuroinflammation may be most prominent during the earlier months and tends to subside over time. The white matter distribution volume ratio (DVR) was lower in Long COVID than in MS (WM DVR 1.03 vs. 1.06; p = 0.007), and individuals imaged within 16 months had higher WM DVR than those with longer durations (1.05 vs. 1.02; p = 0.04).
Importantly, no overall increase in TSPO availability or other imaging markers of neurodegeneration distinguished the Long COVID group from healthy controls across the brain regions measured.
At the same time, the study found strong associations between clinical symptoms and limbic system activity. Lower reported quality of life correlated with higher TSPO DVRs in the hippocampus, amygdala and thalamus (correlation coefficients ρ = −0.83 to −0.70), and depression and anxiety scores correlated positively with hippocampal and amygdalar DVRs (ρ = 0.75–0.97). These links suggest that altered cellular activation in emotion-processing hubs may underlie the severity of neuropsychiatric symptoms in some Long COVID patients.
Toward a clearer understanding of Long COVID and targeted treatments
The findings refine the clinical picture of Long COVID. Rather than a single, sustained process of global neuroinflammation, the disease appears to be more nuanced: inflammation may peak early after infection and then decline, while persistent symptoms can be associated with localized changes in limbic activity. This distinction matters for treatment planning. Anti-inflammatory therapies could be most useful in the early post-infectious window, whereas patients with longer-standing symptoms may benefit more from therapies that target stress pathways, emotional regulation, and neuropsychiatric support.
“This study highlights the need to continue investigating the complex biological mechanisms underlying Long COVID,” adds Airas. “A clearer understanding of timing and regional brain changes will help develop more precise, effective interventions.”
The study, led by Airas and colleagues, is published in the Journal of Neurology. Original research title: “Association between post-COVID-19 neuropsychiatric symptoms and persistent glial activation in the limbic system: a TSPO PET study.” DOI: 10.1007/s00415-026-13842-w.
Frequently asked questions
Q: If Long COVID isn’t causing chronic brain inflammation, why do I still feel depressed, anxious, or exhausted?
A: The study does not dismiss your symptoms. It suggests that the biological drivers change over time. Widespread inflammation appears to peak in the months after infection, but long-term symptoms can be linked to altered cellular activation in limbic regions (hippocampus and amygdala) involved in emotion and stress, which can produce persistent neuropsychiatric effects.
Q: Why compare Long COVID to multiple sclerosis (MS)?
A: MS is a condition with clear, aggressive neuroinflammation. Comparing Long COVID to MS and to healthy controls provides a benchmark: the team found that Long COVID’s white matter inflammatory signal was considerably lower than in MS and similar to healthy controls, helping clarify that Long COVID is not marked by the same degree of chronic inflammation seen in MS.
Q: Should people with long-term symptoms stop anti-inflammatory treatments?
A: The data indicate that anti-inflammatory strategies may be most effective early after infection. For people with symptoms persisting years after infection, treatments focused on stress processing, emotional regulation, rehabilitation and tailored neuropsychiatric care may provide greater benefit. Treatment decisions should always be individualized and discussed with clinical providers.
Editorial notes
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context was provided by editorial staff.
About this research
Author: Tuomas Koivula (University of Turku communications)
Source: University of Turku
Contact: Tuomas Koivula – University of Turku
Image credit: Neuroscience News
Abstract (study overview)
Background: A subset of individuals develop prolonged neurological and psychiatric symptoms following SARS-CoV-2 infection, commonly referred to as Long COVID (LC). Limited evidence has implicated ongoing neuroinflammatory processes. This study used TSPO PET imaging to investigate glial activation in LC.
Methods: Fourteen LC, eleven healthy control and thirteen MS participants underwent [11C]PK11195 TSPO PET and 3T MRI to assess glial activation, white matter pathology and brain volumes. Serum NfL and GFAP were measured as markers of neuronal and glial damage. LC participants completed neurological and mental health assessments.
Results: TSPO availability (DVR) did not differ between LC and healthy controls in any examined brain area, but was significantly lower in LC than in MS (WM DVR 1.03 vs. 1.06; p = 0.007). Individuals imaged within 16 months of infection showed higher WM DVR than those with longer-duration LC (1.05 vs. 1.02; p = 0.04). Lower quality of life correlated with higher DVRs in hippocampus, amygdala and thalamus, while depression and anxiety correlated positively with hippocampal and amygdalar DVRs.
Conclusions: Long COVID participants did not show elevated TSPO availability versus healthy controls, but the higher white matter TSPO signal in earlier post-infection scans suggests that COVID-19-associated neuroinflammation may subside over time. Associations between limbic TSPO availability and symptom severity point to a potential role for limbic activity in Long COVID symptomatology.