Dry eye disease (DED) is a heterogenous ocular surface condition characterized by a loss of homeostasis and associated ocular symptoms. Tear film instability, hyperosmolarity, ocular surface inflammation and damage and neurosensory abnormalities play etiological roles. A meta‐analysis of United States (US) studies estimated a DED prevalence of approximately 8.1% (95% confidence interval (CI) 4.9-13.1%) in the adult population, acknowledging that different diagnostic definitions (symptoms and/or signs) were used across studies. Despite differences in prevalence reports, DED is widely acknowledged as a common condition that drives a large proportion of eye clinic visits and imposes a substantial burden on patients and the healthcare system. ^,, Functionally, ocular pain symptoms coupled with fluctuating vision can impact the ability to perform activities of daily living and significantly reduce quality of life. ^, Current DED therapies target multiple facets of disease, including tear supplementation with ocular lubricants, anti-inflammatory treatment with limited-duration topical corticosteroids or immunomodulators such as cyclosporine, and Meibomian gland dysfunction (MGD)-directed therapies including lid/physical measures and oral macrolide or tetracycline antibiotics; dietary omega-3/polyunsaturated fatty acids have also been investigated as adjunctive options. However, these therapies vary in effectiveness and are often associated with burdensome side effects. The chronic burden of DED and the limited efficacy of existing treatments highlight the need for effective preventive and new therapeutic strategies.

Vitamin D is a multifunctional hormone whose main physiological role is to maintain calcium and phosphate balance, mineral metabolism, and skeletal health. It also has important anti-inflammatory and immune-modulating effects, mediated through regulation of gene transcription, and it can reduce inflammatory signaling and influence programmed cell death. Given that DED is multifactorial with inflammation as an important component, vitamin D may serve as a potential mediator in its pathogenesis.

Several studies have explored the relationship between vitamin D and DED with mixed findings. Some have found that vitamin D deficiency impacts symptoms and signs of DED. In a prospective clinical study of 34 individuals with serum vitamin D deficiency and 21 controls with normal vitamin D levels, the case group demonstrated significantly lower tear film break-up time (5.18 ± 2.15 vs 7.36 ± 3.10 seconds; p =.01) and Schirmer test scores (12.18 ± 6.44 vs 18.57 ± 8.99 mm; p =.007) compared to controls, supporting an association between vitamin D deficiency and impaired tear film stability and production. Vitamin D deficiency has also been implicated in DED symptoms. A 2020 systematic review and meta-analysis included 10 clinical studies (18,919 participants). Pooling 5 studies with DED as the exposure, heterogeneously defined in each study by a mix of symptoms and/or signs, and 5 studies with serum vitamin D as the exposure, the authors found that patients with DED had mean serum 25-hydroxyvitamin D levels 3.99 ng/mL lower than controls. In subgroup meta-analysis restricted to 5 studies that used vitamin D deficiency as the exposure, individuals with vitamin D deficiency had OSDI scores 10.70 points higher and Schirmer test values 6.38 mm/5 min lower than controls, while tear break-up time was similar between the groups, indicating that vitamin D deficiency is associated with a more severe DED phenotype across symptoms and signs. On the other hand, other studies did not find relationships between vitamin D and DED. A cross-sectional study of 740 Korean adults found no significant relationship between serum vitamin D levels and OSDI-measured dry eye symptoms. A similar pattern was observed in a study of 247 US male Veterans, where serum vitamin D levels showed no relationship with tear parameters. Taken together, these studies suggest that vitamin D deficiency may influence certain aspects of DED, although results across the literature remain inconsistent.

Vitamin D supplementation has been investigated as a treatment for dry eye disease, with predominantly beneficial effects, particularly among individuals with vitamin D deficiency. In a 2024 systematic review and meta-analysis of eight studies including 439 participants, vitamin D supplementation was associated with improved symptoms on the OSDI, with a standardized mean difference (SMD) of −1.10 (95% confidence interval −1.45 to −0.74), and improved clinical signs, including increased tear production on the Schirmer tear secretion test (SMD 1.43, 95% confidence interval 0.81 to 2.05) and improved tear film stability measured by tear film break-up time (SMD 1.19, 95% confidence interval 0.83 to 1.55), supporting a potential therapeutic role for supplementation. Similar findings were noted in smaller studies. In a retrospective observational study, 105 patients with DED refractory to conventional treatment and vitamin D deficiency or insufficiency (mean serum 25-hydroxyvitamin D level 10.52 ± 4.61 ng/mL) received a single intramuscular injection of 200,000 international units of cholecalciferol. Several tear film parameters improved, at least transiently, including tear stability (3.16 ± 2.27 → 5.19 ± 2.34 seconds at 6 weeks, p =.001) and production (6.69 ± 3.92 → 8.92 ± 7.60 millimeters at 6 weeks ( p =.015). Taken together, these findings suggest that vitamin D supplementation may serve as a useful adjunct therapy for DED, but larger, high-quality randomized trials are still needed to define optimal dosing, durability of effect, and the patient subgroups most likely to benefit.

Overall, although significant associations between vitamin D deficiency and DED have been reported, the literature is limited by small sample sizes, cross-sectional designs, and methodological inconsistencies across studies. Large-scale, longitudinal analyses are thus needed to clarify the strength and direction of this relationship. This study aims to address these gaps through a retrospective cohort study using a large, federated database to evaluate the association between vitamin D deficiency and incident DED.

This retrospective cohort study utilized data from the TriNetX US Collaborative Network, a federated research platform that aggregates de-identified electronic health records from more than 180 healthcare organizations across the United States. The platform is fully compliant with the Health Insurance Portability and Accountability Act (HIPAA), with all patient-level data de-identified in accordance with Section 164.514(a) of the HIPAA Privacy Rule. TriNetX’s integrated analytic tools, including incidence estimation, outcomes analyses, and propensity score matching, allow patient-level analyses while reporting only aggregate-level results, thereby preserving data confidentiality. Because this study relied exclusively on de-identified data, it did not constitute human subjects research, and review by an Institutional Review Board was not required. The study adhered to the ethical principles outlined in the Declaration of Helsinki and its subsequent amendments and was conducted in accordance with the STROBE (STrengthening the Reporting of Observational Studies in Epidemiology) guidelines.

The TriNetX database was queried for records spanning September 2005 through September 2025. Eligible participants were adults aged 18 years or older at the time of their most recent qualifying clinical encounter. Two cohorts were defined using International Classification of Diseases, Tenth Revision (ICD-10) codes. Within the TriNetX platform, historical ICD-9 diagnoses are algorithmically mapped to their ICD-10-CM equivalents, allowing cohort definition using ICD-10 codes over the entire study period. Cohort 1 consisted of patients with a documented diagnosis of Vitamin D deficiency (ICD-10: E55), while Cohort 2 comprised patients who had undergone routine adult medical examinations without abnormal findings (ICD-10: Z00.00) and who had no prior history of vitamin D deficiency. The index date was defined as the first recorded encounter for vitamin D deficiency in Cohort 1 and the first documented routine adult examination in Cohort 2. Encounters occurring before September 2005 were excluded. Outcomes were assessed beginning one day after the index event, and all subsequent events recorded through September 2025 were included in the observation period.

To reduce baseline imbalances and minimize confounding, propensity score matching was performed at a 1:1 ratio using the nearest-neighbor method without replacement. Matching was restricted to 5 prespecified variables: age at index and sex in the demographics category; ischemic heart disease (ICD-10: I20–I25) and diabetes mellitus (ICD-10: E08–E13) in the diagnosis category; and body mass index (BMI) in the laboratory category. Other characteristics were captured and reported in the analysis, including race and ethnicity, as well as clinical diagnoses relevant to ocular health such as menopausal and perimenopausal disorders (ICD-10: N95), essential (primary) hypertension (ICD-10: I10), nicotine dependence (ICD-10: F17), other hypothyroidism (ICD-10: E03), thyrotoxicosis/hyperthyroidism (ICD-10: E05), depressive episode (ICD-10: F32), and generalized anxiety disorder (ICD-10: F41.1). Covariate balance after propensity score matching was assessed using SMDs, with an absolute SMD <0.1 considered indicative of ideal balance and values <0.2 regarded as acceptable for residual imbalance. Follow-up duration for both cohorts was reported as the mean (± standard deviation) and the median (with interquartile range), expressed in days.

The primary outcome was the diagnosis of Dry Eye Disease (ICD10: H04.12). Statistical analyses were conducted using the TriNetX analytics platform. Hazard ratios (HRs) with 95% CIs were estimated using Cox proportional hazards regression models, and time-to-event outcomes were summarized with Kaplan–Meier cumulative incidence curves and compared using log-rank test. Differences in incidence proportions between cohorts were assessed with chi-square tests, and effect sizes were expressed as risk ratios (RRs) and odds ratios (ORs) with corresponding 95% confidence intervals (CIs). A 2-sided P value <.05 was considered statistically significant.