Long-term ocular symptoms after mild COVID-19 linked to corneal neurodegeneration and immune dysregulation

New study links long-term ocular symptoms after mild COVID-19 to corneal nerve loss and immune dysregulation

Persistent ocular symptoms lasting up to 3 years after non-hospitalized COVID-19 were associated with corneal nerve loss, pupillary dysautonomia, and tear film immune dysregulation, according to a prospective cross-sectional study published in Nature Communications.¹

Don’t miss out on hearing about our latest peer-reviewed articles, expert opinions, conference news, podcasts and more.

Register Register

Why does this matter?

Ocular symptoms, including light sensitivity, eye pain, blurred vision, and difficulty reading or focusing, have been described in patients with severe COVID-19 and long COVID, but less is known about symptoms that follow milder infections not requiring hospitalization.¹,² Previous work has also identified corneal nerve fibre loss and increased dendritic cells in patients with long COVID, suggesting a possible ocular surface and peripheral nerve component.³ However, objective diagnostic criteria and biomarkers for post-COVID ocular symptoms remain lacking, and the long-term biological basis of these symptoms has not been clearly established.¹

This matters because most cases of COVID-19 do not require hospitalization, meaning that persistent ocular symptoms after milder infection could affect a much larger group than is currently recognized. Longer-term studies of postacute COVID-19 sequelae have also highlighted the need to understand symptoms that persist beyond the acute phase of infection.⁴

What did the study investigate?

Investigators in Sweden conducted a prospective, cross-sectional observational study involving 100 participants who developed new visual or ocular symptoms lasting at least 12 weeks after confirmed or strongly suspected mild, non-hospitalized COVID-19. They were compared with 32 age-matched controls who had recovered from COVID-19 without persistent ocular symptoms.¹

Participants were recruited after national media coverage of the study. Individuals with prior ocular disease, previous ocular surgery, or relevant systemic disease, including diabetes, were excluded.¹

The study included a modified Catquest-9SF questionnaire to assess vision-related quality of life, free-text symptom reporting, and a range of clinical assessments. These included visual acuity, pupillary light responses, corneal sensitivity by blink reflex testing, corneal subbasal nerve density and dendritic/T-cell counts using in vivo confocal microscopy, and tear film proteomic profiling.¹

Group differences were assessed using age-adjusted regression models with false discovery rate (FDR) correction for multiple comparisons. The authors also used penalized logistic regression with bootstrap resampling to develop exploratory models for distinguishing symptomatic participants from controls. Model performance was assessed using area under the receiver operating characteristic curve (AUC) with five-fold cross-validation.¹

What do the data show?

Most participants with persistent ocular symptoms were female (73%). Symptom onset occurred 0.2–4 months after infection in 68% of participants. At the time of examination, symptoms had persisted for at least 1 year in 78.2% of participants and for at least 2 years in 33.3%; the interval from infection to examination ranged from 3 to 42 months.¹

One-third of participants in the symptomatic group were on partial or full sick leave. Of these, only 39% had a formal long-COVID diagnosis. The most frequently reported symptoms were photophobia (42.7%), eye pain (30.2%), eye fatigue (26.0%), and reduced focusing ability (22.9%).¹

Routine anterior eye examination and refraction did not identify several standard abnormalities, but specialized testing showed differences between the symptomatic and control groups. Participants with persistent ocular symptoms had reduced corneal subbasal nerve density, weaker blink reflex responses, altered pupillary responses, and higher mature dendritic/T-cell density in the corneal epithelium.¹

Tear film proteomic analysis identified 178 proteins with nominal differences between groups. Five remained significant after FDR correction: integrin-β6 (ITGB6), neurofascin (NFASC), mitochondrial creatine kinase 1A/1B (CKMT1A-CKMT1B), CCN2/CTGF, and tryptase-α/β-1 (TPSAB1).¹ Several clinical measures also correlated with tear protein levels, including dilated and constricted pupil size with JUN (FDR=0.004 and FDR=0.005, respectively).¹

An exploratory diagnostic model using five clinical parameters achieved a cross-validated AUC of 0.77 for distinguishing symptomatic participants from controls. Adding six tear proteins increased the cross-validated AUC to 0.91.¹

Were the findings consistent across subgroups?

The article reported sex-stratified symptom frequencies, with no significant difference in symptom prevalence between female and male participants (minimum FDR=0.38).¹ However, the study was not designed or powered to draw robust conclusions across subgroups such as age, SARS-CoV-2 variant, vaccination status, or reinfection history.

Why is this relevant for ophthalmologists?

Patients with persistent, otherwise unexplained photophobia, eye pain, or focusing difficulty after COVID-19 may not show clear abnormalities during routine slit-lamp examination. In this study, several of the reported differences were detected only with more specialized assessments, including in vivo confocal microscopy and pupillometry.¹

The findings suggest that post-COVID ocular symptoms may have measurable biological correlates in some patients, rather than being purely subjective or unexplained. They also support greater awareness of this symptom cluster and closer coordination with long-COVID services, particularly because many symptomatic participants on sick leave did not have a formal long-COVID diagnosis.¹

However, the diagnostic models described in the study should be viewed as exploratory research tools. They are not validated clinical tests and should not replace standard ophthalmic assessment.

Limitations

The study included a relatively small sample of 100 symptomatic participants and 32 controls, and it did not estimate the overall prevalence of persistent ocular symptoms after mild COVID-19.¹ Recruitment relied on self-referral after media coverage, which may have enriched the cohort for people with more severe or persistent symptoms and limits generalizability.

Symptom data were self-reported, and the study did not include detailed information on symptom severity, SARS-CoV-2 variant, or vaccination history. There was also no control group entirely free of prior SARS-CoV-2 infection; the authors instead contextualized selected findings against published pre-pandemic normative data.¹

Because tear proteins rather than ocular tissue were sampled, the study could not confirm whether the identified corneal dendritic/T cells directly expressed the implicated proteins or whether the tear film findings reflected broader immune changes. The cross-sectional design, with a single examination performed 3–42 months after infection, also means the study cannot determine how symptoms or biomarkers changed over time within individuals.¹

The diagnostic models were assessed using internal cross-validation only, without an independent external validation cohort. As a result, their clinical utility remains uncertain.¹

Clinical takeaway

In this prospective cross-sectional study, persistent ocular symptoms after mild COVID-19 were associated with corneal nerve loss, altered pupillary responses, and tear film immune dysregulation. These findings were detected using specialized testing rather than routine examination.¹

The combined clinical and tear-protein model distinguished symptomatic participants from controls with a cross-validated AUC of 0.91, but this should not be interpreted as proven clinical diagnostic accuracy. The small, self-referred sample and lack of external validation mean the findings are hypothesis-generating rather than practice-changing. Larger prospective studies with independent validation cohorts are needed to determine how common this symptom cluster is and whether biomarker-based diagnostic approaches have a role in clinical care.

References

1. Moustardas P, Setterud H, Meijer H, et al. Long-term ocular symptoms following COVID-19 linked to immune dysregulation, dysautonomia and peripheral neuropathy. Nat Commun. 2026;17:5624.
2. Davis HE, Assaf GS, McCorkell L, et al. Characterizing long COVID in an international cohort: 7 months of symptoms and their impact. EClinicalMedicine. 2021;38:101019.
3. Bitirgen G, Korkmaz C, Zamani A, et al. Corneal confocal microscopy identifies corneal nerve fibre loss and increased dendritic cells in patients with long COVID. Br J Ophthalmol. 2022;106:1635–41.
4. Bowe B, Xie Y, Al-Aly Z. Postacute sequelae of COVID-19 at 2 years. Nat Med. 2023;29:2347–57.