Abstract
Background/Aim: Aortic stenosis (AS) is a fatal cardiovascular disease associated with metabolic syndrome and inflammation. Ocular surface diseases are also characterized by elevated inflammatory responses. Therefore, this study aimed to investigate the correlation between AS and the subsequent development of ocular surface diseases, including dry eye disease (DED) and superficial keratopathy.
Patients and Methods: A retrospective cohort study was performed via the usage of the TriNetX database. The patients were divided according to the presence of AS and a total of 421,253 patients were enrolled into both the AS and non-AS groups. The primary outcomes were the development of DED and superficial keratitis after AS. The Cox proportional hazard regression was applied to calculate the adjusted hazard ratio (aHR) and 95% confidence interval (CI) of primary outcomes between groups.
Results: There were 11,915 and 8,260 DED episodes in the AS and the non-AS groups, respectively. Also, there were 1,409 and 1,038 superficial keratopathy events in the AS and non-AS groups, respectively. After adjusting for all the confounders, the AS group showed a significant higher incidence of DED (aHR=1.375, 95%CI=1.337-1.414, p<0.001) and superficial keratopathy (aHR=1.286, 95%CI=1.187-1.393, p<0.001) compared to the non-AS group. The cumulative incidences of DED and superficial keratopathy were also significantly higher in the AS group than in the non-AS group (both p<0.001). In subgroup analyses, the risk of ocular surface diseases was significantly higher in all AS subgroups except the Asian population.
Conclusion: The presence of AS associates with a higher risk of subsequent DED and superficial keratopathy development, which positively relate to the duration of AS.
Introduction
Aortic stenosis (AS) is a severe cardiovascular disease characterized by calcification, thickening, and fibrosis of aortic valve leaflets (1, 2). In the Western population, the prevalence of AS is estimated to be 5 percent in people older than 65 years (3). Medical treatments can be applied in mild AS while surgery, like aortic valve replacement, is advocated in severe cases (4, 5). Although the survival time after aortic valve replacement is acceptable (6, 7), postoperative complications of AS may directly contribute to mortality (3).
AS is associated with several disorders according to previous studies (2). Diabetes mellitus (DM) is observed frequently in patients with AS (8). In addition, hypertension also accompanies AS development and the concurrent hypertension can affect the symptoms and severity of AS (9). Moreover, hypercholesterolemia can damage the cardiovascular system and the development and prognosis of AS is also influenced by dyslipidemia (10). In addition to the above metabolic syndromes, the elevated inflammatory response was correlated with AS development (11).
Dry eye disease (DED) is an inflammatory eye disease, which is characterized by tear film homeostasis loss and elevated inflammatory cytokines on the ocular surface (12). In previous studies, DED was associated with the existence of other inflammatory disease including the chronic rhinosinusitis and ankylosing spondylitis (13, 14). However, superficial keratopathy was another ocular surface disease that correlated with Sjögren syndrome and rheumatoid arthritis (15). Nevertheless, the correlation between AS and ocular surface diseases had rarely been evaluated. Because both AS and ocular surface diseases exhibit inflammation (12, 15, 16), the presence of AS may affect the risk of developing ocular surface diseases.
Consequently, the objective of the current study was to evaluate the potential correlation between the AS and subsequent ocular surface diseases via the usage of the TriNetX database. Other predisposing factors for ocular surface disease development were also included in the analysis.
Patients and Methods
Data source. The current study adhered to the 1964 declaration of Helsinki and subsequent amendments. Furthermore, the current study was approved by the Institute Review Board of Chung Shan Medical University Hospital (project code: CS2-23044). TriNetX is a global federated health research network providing access to electronic medical records (diagnoses, procedures, medications, laboratory values, genomic information) across large healthcare organizations. It was generated using the US Collaborative Network, which includes 67 healthcare organizations. TriNetX is a platform that de-identifies and aggregates electronic health record data from contributing healthcare systems, most of which are large academic medical institutions with both inpatient and outpatient facilities at multiple locations across all 50 states in the United States. TriNetX Analytics provides Web-based and secure access to patient electronic health record data from hospitals, primary care and specialty treatment providers, covering diverse geographical locations, age groups, ethnic groups, income levels and insurance types, including several commercial insurances, governmental insurance (Medicare and Medicaid), self-pay and uninsured, worker compensation insurance, and military and Veterans Affairs insurance, among others. The medical information in the TriNetX database includes the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes, age, sex, occupation, region of habitat, educational degree, socioeconomic status, length of hospitalization, image codes, laboratory examination codes and results, procedure codes, surgery codes, and Anatomical Therapeutic Chemical (ATC) codes for medications.
Participant selection. A retrospective cohort study was conducted and patients with the following characteristics were regarded and selected as having AS: 1) the receipt of AS diagnosis according to the related ICD-10 CM codes, 2) the receipt of echocardiogram, electrocardiogram, chest X-ray, cardiac computerized tomography or cardiac catheterization before the diagnosis of AS, and 3) the AS diagnosis was entered by a cardiologist. The index date was set at 6 months after the diagnosis of AS. In addition, the subsequent exclusion criteria were utilized to elevate the homogeneity of the study population: 1) age under 20 or over 80 years, 2) receipt of any ocular surgery before the index date, and 3) outcome (in the following section) developed before the index date. For the comparison, patients with AS were matched to individuals without the AS via the propensity score match (PSM) process. The PSM process integrates demography, systemic diseases and medication prescriptions into specific score and matches different individuals according to that. Finally, a total of 421,253 and 421,253 patients were selected into the AS and non-AS groups, respectively. The flowchart of patient selection is shown in Figure 1.
Flowchart of participant selection. AS: Aortic stenosis; DED: dry eye disease; N: number; PSM: propensity score matching.
Primary outcomes. The primary outcomes of the current study were the development of DED and superficial keratopathy. DED development was defined as the following conditions: 1) the receipt of the diagnosis of DED according to the related ICD-10 CM codes, 2) the performance of tear break-up time exam, Schirmer test, or slit-lamp biomicroscope exam according to the procedure code, and 3) the prescription of artificial tear according to the ATC codes. On the other side, superficial keratopathy development was regarded as the following statuses: 1) the receipt of the diagnosis of superficial keratopathy according to the related ICD-10 CM codes, 2) the performance of slit-lamp biomicroscope exam or fluorescein stain test according to the procedure codes, and 3) the prescription of topical antibiotic eyedrop or ointment according to the ATC codes. Only the DED and superficial keratopathy events that were recorded after the index date were accounted for the outcome achievements in the current study. All the patients were followed till outcome achievement, withdraw from the available health insurance program or the end of TriNetX database: December 31, 2023.
Confounder adjustment. For better elucidating the correlation between AS and subsequent corneal diseases, the effect of the following confounders was adjusted in the multivariable models: age, sex, urbanization, income level, hypertension, DM, hyperlipidemia, ischemic heart disease, chronic kidney disease, cerebrovascular disease, chronic lower respiratory diseases, peripheral vascular disease, estimated glomerular filtration rate (eGFR), high-density lipoprotein (HDL), low-density lipoprotein (LDL), and Troponin I. To ensure the presence of systemic disease was long enough to affect the risk of ocular surface diseases, only co-morbidities that persisted for more than two years were selected into the multivariable analysis of the current study.
Statistical analysis. SAS, version 9.4 (SAS Institute Inc, Cary, NC, USA) was employed in the statistical analyses. The descriptive analysis was employed to present the basic characteristics between the AS and the non-AS groups, and the standard mean difference (SMD) was employed to compare the difference of each parameter between the two groups. An SMD larger than 0.1 was regarded as significant difference in the current study. Then the Cox proportional hazard regression was employed to analyze the incidences of DED and superficial keratopathy between the two groups, and the adjusted hazard ratio (aHR) and related 95% confidence interval (CI) for the incidence of ocular surface diseases were calculated. The age, sex, urbanization, income level, hypertension, DM, hyperlipidemia, cerebrovascular disease, and peripheral vascular disease and laboratory data including eGFR, HDL, LDL, and Troponin I were incorporated into the Cox proportional hazard regression to adjust for their effect on ocular surface disease development. The Kaplan-Meier curve was drawn and the cumulative incidences of ocular surface disease events between the two groups were estimated using the log-rank test. In the subgroup analysis, the patients were divided into different subgroups according to age, sex, HDL, LDL and eGFR. Then the Cox proportional hazard regression was employed again to analyze the primary outcomes in different subgroups. Statistical significance was set at p<0.05.
Results
The basic characteristics of the participants are demonstrated in Table I. The mean age was 65.1±11.7 and 65.4±11.5 in the AS and non-AS groups, respectively. The difference in mean age between the two groups was insignificant (SMD=0.025). The sex distribution and urbanization degree between the two groups were also similar (both SMD <0.1). Regarding co-morbidities, the AS group showed similar rates of hypertension, DM, hyperlipidemia, and other diseases compared to the non-AS group (all SMD <0.1) and the laboratory data values were also similar between groups (all SMD <0.1) (Table I).
Comparison of baseline characteristics between aortic stenosis and control cohorts.
After the whole follow-up interval of 13 years, 11,915 and 8,260 DED episodes occurred in the AS and the non-AS groups, respectively. Also, 1,409 and 1,038 superficial keratopathy events developed in the AS and the non-AS groups, respectively. After adjusting for all the confounding factors, the AS group showed a significantly higher incidence of DED (aHR=1.375, 95%CI=1.337-1.414, p<0.001) and superficial keratopathy (aHR=1.286, 95%CI=1.187-1.393, p<0.001) compared to the non-AS group (Table II). In addition, the cumulative incidences of DED and superficial keratopathy were also significantly higher in the AS group than in the non-AS group (both p<0.001) (Figure 2 and Figure 3).
Comparison of main outcomes between aortic stenosis and control cohorts.
The Kaplan-Meier curve and cumulative incidence of dry eye disease (DED) between groups. *Significant difference between groups.
The Kaplan-Meier curve and cumulative incidence of superficial keratopathy between groups. *Significant difference between groups.
In the subgroup analysis, the incidences of DED in the AS populations were significantly higher than those in the non-AS populations in all subgroups except the Asian population (aHR=1.118, 95%CI=0.980-1.276) (Figure 4). Furthermore, the incidences of superficial keratopathy in the AS populations were significantly higher than those in the non-AS populations in all subgroups except the Asian population (aHR=1.069, 95%CI=0.653-1.749) and those aged from 20-45 years (aHR=1.058, 95%CI=0.765-1.463) (Figure 5).
Risk of dry eye disease in patients with aortic stenosis stratified by age, sex, race, high-density lipoprotein (HDL) and low-density lipoprotein (LDL) levels.
Risk of superficial keratopathy in patients with aortic stenosis stratified by age, sex, race, high-density lipoprotein (HDL) and low-density lipoprotein (LDL) levels.
Discussion
In the current study, the risk of developing ocular surface disease including DED and superficial keratopathy were significantly higher in the AS group than in the non-AS group. Moreover, the risk of developing the above two ocular surface diseases were positively correlated with the disease interval of AS. However, the incidences of DED and superficial keratopathy were not affected by AS in the Asian population.
The AS had been shown to correlate with several disorders in earlier studies (17). The metabolic syndrome, which includes hypertension, dyslipidemia and glucose intolerance, showed positive association with the presence of AS (16). The risk of diabetes is higher in patients with AS, and DM patients with AS showed higher valve calcification than the non-DM population (8). Moreover, hypertension and hyperlipidemia also develop more frequently in patients with AS (9, 10). Except for the metabolic syndrome, prostate cancer was also found more commonly among patients having AS (18). The inflammatory response is an important factor for AS development in which lipid deposits and inflammation are major mechanisms of the initial phase of AS development (19, 20). In addition, the levels of certain inflammatory biomarkers are increased in patients with AS (21). The concentration of interleukin and tumor growth factor beta is higher in patients with AS (22). Also, nuclear factor-kappa B and bone morphogenetic protein pathways were significantly activated during the development of AS (19). However, the ocular surface diseases are also inflammatory disorders where the levels of inflammatory biomarkers, including the caspase-1 and interferon-gamma, increase in the tear film of individuals with DED or corneal damage (23, 24). Furthermore, the application of anti-inflammatory medications like topical steroids and cyclosporine could alleviate the sign and symptoms of DED (25). Furthermore, systemic inflammatory diseases such as Sjögren’s Syndrome, diabetes mellitus and rheumatoid arthritis were significantly associated with the development of DED and superficial keratopathy (15, 26). Because both AS and ocular surface diseases showed inflammation and association with systemic diseases (12, 17, 19, 24), it is possible that the presence of AS influences the incidence of subsequent ocular surface diseases. Our hypothesis was supported by the findings of the current study at least to some degree.
The existence of AS was associated with higher incidence of subsequent DED and superficial keratopathy in the current study. In a previous study, the severity of coronary heart disease was correlated with the subsequent DED degree (27). Also, vasculitis positively correlates with subsequent corneal disease development (28). Still, there was no evidence for the relationship between AS and subsequent ocular surface diseases. To our knowledge, these preliminary results show positive correlation between AS and the subsequent ocular surface developments. Furthermore, we excluded episodes of ocular surface disease that occurred before the index date (i.e., six months after the diagnosis of AS), thereby ensuring a clear temporal relationship between AS and subsequent ocular surface diseases. Moreover, we adjusted the effect of age, sex, metabolic diseases and laboratory data that can influence the occurrence of ocular surface diseases using the Cox proportional hazard regression (29, 30). According to this analysis, AS may be an independent risk factor for the development of subsequent ocular surface diseases. AS has also been correlated with inflammatory reaction including E-selectin elevation and T-lymphocyte aggregation (31). Therefore, it is reasonable to assume that AS associates with ocular surface disease through inflammation (23). In addition, the cumulative probability of DED and superficial keratopathy were significantly higher in the AS group than in the non-AS group. In a previous study, the duration of coronary heart disease was related to the risk of DED (27). Our findings also demonstrate the importance of AS duration in the risk of development of comorbidities.
Concerning the subgroup analysis, AS patients with different age and sex generally demonstrated higher risks of developing DED and superficial keratopathy than the non-AS population. The old age had been shown to be a risk factor for the development of DED (29), and the female sex was also a known risk factor for DED occurrence (32). The aHR of both ocular surface diseases were numerically higher in the old age and female populations, which may indicate that the influences of AS on ocular surface disease is universal regardless of the existing risk factors. Among populations with different laboratory profiles, patients with AS consistently exhibited a higher incidence of ocular surface diseases compared to those without AS. Few studies have reported these findings. Dyslipidemia has been associated with the presence of meibomian gland dysfunction, a known risk factor for ocular surface disease (33). In addition, DED occurred more frequently in those with chronic kidney disease (34). These results concerning the laboratory data analyses further highlight the prominent correlation of AS with the development of ocular surface diseases in different populations. Interestingly, the incidences of DED and superficial keratopathy exhibited insignificant differences between the AS group and the control group in the Asian population. In contrast, a previous study showed that the prevalence of DED in the Asian population was high (35). A possible explanation for this discordance is that the number of Asian patients in the TriNetX database is relatively small, and subsequently the numbers of DED and superficial keratopathy are low. Accordingly, the small numbers of ocular surface diseases may cause statistical bias, although the aHR of both ocular surface diseases were numerically higher in the Asian population with AS.
Regarding the epidemiology, AS is a cardiovascular disease with moderate incidence; however, the incidence is approximately 2.8 percent in patients older than 75 years in the North America region and approximately 4.43 percent of US people are affected by AS (4, 36). In addition, the incidence of AS in the Chinese population aged older than 75 years was 21.9 percent (35). Aortic valve replacement is performed in the majority of patients with AS in developed countries, which is a huge medical cost (37). Regarding the surgical intervention, operative mortality for aortic valve transplantation was above 1% although the survival period would be increased after surgery (4, 11). However, DED is also a prevalent disease, which affects more than 5% of people in the world (12). In the population that frequently uses visual display devices, the incidence of DED can be near 60 percent and can significantly reduce life quality (38). Superficial keratopathy, although not as common as DED, exhibits an incidence of 2% percent in the general population (39). Severe keratopathy can contribute to corneal melting and eyeball rupture in which permanent visual loss may occur (40). Given the substantial prevalence of AS, DED, and superficial keratopathy, along with their significant impact on public health and healthcare expenditures, it is essential to explore the interrelationships between these conditions.
Study limitations. First, we used the claimed database rather than real medical documents for the analysis. Accordingly, multiple important data including the site of AS, the image analysis results of AS, the surgical details of AS, the postoperative condition/reduction of AS, the recurrence of AS, the external eye photography of DED and superficial keratopathy, the results of DED-related exams, the results of microbial culture for keratopathy, the treatment outcomes of ocular surface diseases and the details of co-morbidities cannot be investigated. Second, the retrospective design of the current study would reduce the homogeneity of the study population even after the PSM process compared to a prospective one. Besides, the rates of systemic diseases were significantly higher in the AS group. Although we adjusted for all the diseases in the Cox proportional hazard regression, the uneven co-morbidity distribution may still cause some bias. Finally, most of the patients in the TriNetX database belong to Caucasian population, thus the external validity of the current study to the Asia-pacific or Africa population may be reduced.
In conclusion, the presence of AS is correlated with higher incidence of subsequent ocular surface diseases, including DED and superficial keratopathy. Furthermore, the correlation between AS and subsequent ocular surface diseases is positively associated with the disease period of AS. Consequently, periodic ophthalmic exam may be suggested for the patients with prolonged AS. Further large-scale prospective studies are needed to evaluate the effect of AS on the treatment outcomes of DED and superficial keratopathy.
Footnotes
Authors’ Contributions
SLP participated in data interpretation, writing and revision of the manuscript. JYH participated in data interpretation, carried out statistics analysis and interpretation of the data. CYL participated in data interpretation and writing of the manuscript. CJW participated in data interpretation, statistical analysis, writing and revision of the manuscript. SFY participated in data interpretation, statistical analysis, writing and revision of the manuscript. The Authors declare that there is no conflict of interest with the current publication, and all Authors have approved the final version of the manuscript.
Conflicts of Interest
The Authors declare no competing interests in relation to this study.
Funding
None.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received April 30, 2025.
- Revision received May 14, 2025.
- Accepted May 16, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).











