For most people, COVID-19 never required hospitalization. Yet its aftermath can be long and complex, extending well beyond the familiar loss of smell, fatigue, and cognitive symptoms. A new Nature Communications study now suggests that the eye may be an under-recognized site of post-COVID pathology, and that persistent ocular symptoms after mild infection may have measurable neuroimmune signatures.
The researchers – based at Linköping University – prospectively examined 100 individuals in Sweden who developed persistent ocular symptoms after non-hospitalized COVID-19, comparing them with 32 post-COVID controls who had recovered without ocular symptoms. The team combined symptom questionnaires, quality-of-life measures, detailed ophthalmic testing, in vivo confocal microscopy, dynamic pupillometry, corneal sensitivity testing, and tear-film proteomics.
The symptom burden was substantial. Ocular symptoms emerged within 0.2-4 months of initial infection in 68 percent of the affected participants, and by the time of examination had persisted for at least one year in 78.2 percent and at least two years in 33.3 percent of subjects. One-third of the persistent ocular symptoms group were on part-time or full-time sick leave, but only 39 percent of these had a formal long-COVID diagnosis.
On patient-reported assessment, the persistent ocular symptoms group showed significantly greater vision-related disability than controls. Reading text, particularly from screens, was among the most difficult daily tasks. Free-text symptom analysis identified 22 distinct self-described symptoms, most commonly light sensitivity or photophobia, eye pain, eye fatigue, and reduced focusing ability.
Routine clinical examination (i.e. refraction tests and standard anterior eye examinations) did not explain these complaints. The tests showed no major differences in distance best-corrected visual acuity, refraction, tear production, tear break-up time, ocular redness, corneal thickness, endothelial cell density, accommodation, or stereoacuity. However, specialized testing revealed a different picture: deficits in best-corrected binocular near visual acuity, near ocular alignment, and fusional reserve, consistent with latent misalignment and impaired binocular function.
The neuro-ophthalmic findings were notable. In vivo confocal microscopy showed reduced corneal subbasal nerve density, while Cochet–Bonnet esthesiometry demonstrated a weakened blink reflex. Dynamic pupillometry revealed larger dilated and constricted pupil sizes, longer constriction time, and faster recovery of dilation, suggesting ocular dysautonomia. The study authors also found elevated mature dendritic/T-cell density in the corneal subbasal nerve plexus, pointing to persistent ocular surface immune activation.
Tear film proteomics added a mechanistic layer. Among 768 proteins measured, 178 were dysregulated in the persistent symptoms group. The dysregulated network overlapped with signatures reported in severe, fatal, and long COVID datasets. Five proteins remained significant after multiple-comparison adjustment: ITGB6, NFASC, CKMT1A/B, CCN2, and TPSAB1. Correlations linked pupil dysfunction with elevated JUN, and dendritic/T-cell dysregulation with ANGPTL2, SKAP2, and DAPP1.
The study also explored diagnosis. A model using five clinical parameters – based on corneal nerve imaging, blink response, pupillometry, and mature dendritic/T-cell density – achieved 77 percent cross-validated AUC accuracy. Adding six tear proteins increased cross-validated AUC to 91 percent.
The study authors acknowledge limitations, including modest sample size, lack of prevalence estimates, and reliance on self-reported symptoms. Even so, the findings underscore that some patients with unexplained post-COVID ocular complaints may have detectable abnormalities beyond routine examination.
