Eye Diseases in Chronic Kidney Disease: A Nationwide Longitudinal Case–Control Study in Sweden: Response

Abstract

We appreciate the interest and insights shared by Hsu et al. [1]. regarding our recently published article in Clinical and Experimental Ophthalmology [2]. The points raised are important, and we are grateful for the opportunity to address these questions.

We agree that using repeated identical ICD codes, especially in combination with dialysis procedure codes, would strengthen the specificity of CKD diagnosis. It is also likely that incorporating data on glaucoma medication prescriptions would enhance the robustness of glaucoma diagnoses. However, our study used an existing database that did not include procedure codes or glaucoma medication data, as it was not designed to study these parameters. That said, as highlighted in our article, the Swedish National Patient Register provides comprehensive, high-quality, validated data commonly used in large-scale cohort studies [3].

The study focused on analysing the risk for different eye diseases in CKD for several conditions, including subdivisions based on age groups. Cox proportional hazards model is considered a robust method when investigating the association between two diseases over time.

We acknowledge that a more detailed approach could have provided deeper insights, particularly when analysing different stages of CKD and glaucoma subtypes. However, our study aimed to examine the overall relationship between CKD and eye diseases, rather than focusing on specific types of eye disease or CKD severity. These limitations were discussed in our article, as they arose from the constraints of the pre-existing database, which could not be modified.

It is also important to note that the incidence of glaucoma was higher in age groups II (31–45 years) and III (46–65 years), with HRs of 1.98 and 1.26, respectively, though analysis was not possible for group I due to a limited number of cases. Hence, glaucoma incidence was elevated in patients under 66 years. Differences in ethnicity between Taiwan and Sweden, such as variations in myopia prevalence [4] and the incidence of pseudoexfoliation syndrome (PEX), may also contribute to differences in glaucoma rates. In some regions of Sweden, PEX and open-angle glaucoma are more common, particularly in the aging population [5]. These factors could help explain the differences in the incidence of open-angle glaucoma between the studies and might account for some of the discrepancies in our cases and controls.

In conclusion, the differences in outcomes between those highlighted by Hsu et al. in their letter and our study are likely due to both methodological variations and population differences. Our results underscore the importance of awareness of glaucoma in CKD, and we are thankful to Hsu et al. and the Editor for the opportunity to clarify this point once again.

The authors declare no conflicts of interest.