From the Cornea to the Brain: How Dry Eye Disease May Be Linked to Anxiety

A new study titled 'Comorbidity of Anxiety and Dry Eye Disease: Shared Biomarkers and Immune Responses,' published in Frontiers in Immunology, explores why dry eye disease and anxiety so often appear in the same person. Researchers Xi Long, Xin Peng and their colleagues combined gene-expression analysis, single-cell sequencing and laboratory experiments in mice, uncovering four genes and several immune pathways that could connect discomfort on the eye's surface with changes in the brain linked to anxiety.

Dry eye disease affects an estimated 5% to 50% of people, depending on age, location and diagnostic criteria, causing burning, irritation, unstable vision and damage to the eye's protective surface. Anxiety disorders are also widespread and can involve persistent fear, heightened alertness, muscle tension and avoidance. People living with dry eye frequently report anxiety or depression, but the biology behind this relationship has remained unclear.

Researchers searched for genes shared by both conditions

The researchers examined gene-expression datasets from the Gene Expression Omnibus, using GSE44101 for dry eye disease and GSE98793 for the mental-health component of the analysis. They applied weighted gene co-expression network analysis, differential gene-expression testing and LASSO machine learning to locate molecular patterns appearing across the two conditions.

The dry eye dataset contained 5,939 differentially expressed genes, while the anxiety-related analysis produced 135. Combining the strongest disease-linked modules, altered genes and machine-learning selections revealed 128 candidate genes shared by both conditions. Independent datasets involving Sjögren's syndrome-related tissue and generalized anxiety symptoms were used to check whether the leading signals appeared in other samples.

Many of the shared genes were connected to circadian rhythm, blood-cell development, platelet activation, lysosomal activity, pyrimidine metabolism and long-term potentiation, a process involved in learning and memory. Other functions included microtubule formation, chromosome separation and cellular enzyme activity.

Circadian rhythm stood out because the body clock influences tear secretion, sleep, stress hormones, immune-cell movement and emotional regulation. Disrupted biological rhythms can alter the tear film, weaken the corneal barrier and increase inflammatory signals; at the same time, disturbances in the brain may affect the stress-response system and chemical messengers involved in anxiety. The study did not directly measure clock-gene cycles or time-dependent changes, meaning this finding remains a promising association rather than proof that circadian disruption causes both conditions.

Four genes emerged as the strongest molecular connection

Protein-interaction analysis initially highlighted six hub genes: UBA3, BTBD1, MS4A1, BANK1, HAUS6 and NEDD1. Further validation and single-cell analysis narrowed the central findings to BTBD1, BANK1, UBA3 and NEDD1, all of which were detected across several corneal cell populations, including epithelial cells, progenitor cells and macrophages.

BANK1 produced the most consistent signal as this gene helps regulate B-cell receptor activity and innate immune responses, placing it in a strong position to influence inflammation across different tissues. Its repeated elevation supports the idea that B-cell activity may help connect ocular inflammation with the immune changes produced by chronic psychological stress.

UBA3 controls part of the neddylation process, a protein-modifying system linked to inflammatory disease and activation of the NLRP3 inflammasome. Previous animal research has connected this pathway with stress-related neuroinflammation and anxiety-like behaviour. BTBD1 and NEDD1 are involved in cellular regulation and microtubule-related processes, though their precise roles in dry eye and anxiety need much more investigation.

These genes should not yet be viewed as diagnostic biomarkers or treatment targets. Their altered expression shows that they are associated with the combined condition, but it does not establish that they initiate dry eye, anxiety or the interaction between them.

Immune cells may carry signals between the eyes and brain

Immune-cell analysis connected the shared molecular signature with plasma cells, naïve and memory B cells, macrophages, dendritic cells, mast cells, neutrophils and several T-cell populations. Plasma cells and naïve B cells showed especially notable relationships, while BANK1 was associated with B-cell lineages and plasma-cell activity.

Single-cell sequencing of 12 human corneal samples showed that the four leading genes were not limited to one type of cell. Their presence in both structural corneal cells and immune cells suggests that chronic stress signals could make the ocular surface more vulnerable to inflammation or slow epithelial repair. Persistent inflammation in the eye could also contribute to a wider neuroinflammatory environment associated with anxiety, though the study cannot confirm the direction of this relationship.

The researchers describe a possible biological bridge in which immune activation, stress hormones, circadian disruption and altered neurotransmission interact across the cornea, lacrimal gland and hippocampus. This offers a more connected explanation for the comorbidity than treating dry eye as a purely local eye problem and anxiety as a completely separate mental-health condition.

Mouse experiments revealed inflammation and tissue-specific gene changes

The team tested its computer-generated findings in 20 female mice, dividing them into a control group and a combined dry eye-anxiety group. Dry eye was produced through a low-humidity environment, increased airflow and repeated application of 0.2% benzalkonium chloride for 28 days. Chronic unpredictable mild stress was used to trigger anxiety-like behaviour through changing stressors such as food or water deprivation, overnight light, cold-water swimming, noise and exposure to unfamiliar surroundings.

Model mice produced fewer tears and developed stronger corneal staining, thinning of the corneal epithelium, disorganised stromal tissue and damage to the tear-producing lacrimal gland. Their blood contained higher levels of the inflammatory molecules IL-1β, TNF-α and IL-6, while staining confirmed increased IL-1β and TNF-α activity in corneal and lacrimal tissues.

Changes also appeared in the hippocampus, where researchers observed neuronal loss and disordered cell arrangement. Levels of serotonin and GABA, two chemical messengers associated with mood and calming activity, fell significantly, while excitatory glutamate increased. The mice spent less time exploring central or open areas during behavioural tests, supporting the presence of anxiety-like behaviour, but they showed no clear increase in depressive-like behaviour during forced-swim and tail-suspension tests.

All four hub genes were elevated at both messenger RNA and protein levels in the cornea. BANK1 also remained elevated in the lacrimal gland and hippocampus, making it the only gene consistently increased across all three tissues. BTBD1, UBA3 and NEDD1 were reduced in the lacrimal gland and hippocampus, revealing that the shared condition does not produce one uniform genetic response throughout the body.

The mouse model used benzalkonium chloride, which directly damages the corneal surface and may not reproduce every human form of dry eye. The experiment lacked separate dry eye-only and stress-only groups, used only female mice and did not control for every possible influence from biological timing, movement or body weight. Some immune correlations were calculated without full correction for multiple comparisons, while gene-expression tests confirmed association rather than function. Human clinical samples, male and female animals, gene-manipulation experiments and larger patient groups will be needed before these signals can support diagnosis or treatment.

The research presents dry eye and anxiety as conditions that may meet inside a shared network of immune activity, biological timing and brain chemistry. BANK1 appears to be the strongest cross-tissue candidate, while BTBD1, UBA3 and NEDD1 offer additional clues about how local eye damage and central stress responses may become biologically intertwined.