Associations Between Dementia Risks and Chronic Exposures to Different Glaucoma Medication Types

Abstract

Purpose : To evaluate whether chronic exposure to different categories of topical glaucoma medications is associated with varying dementia risks.

Methods : We used the data from Adult Changes in Thought, a population-based, prospective cohort study that follows cognitively normal older adults from Kaiser Permanente Washington (KPW) until Alzheimer’s disease (AD) and related dementia development. We included subjects with a diagnosis of glaucoma and KPW pharmacy records of filling topical glaucoma medication (alpha-adrenergic agonists, beta-adrenergic antagonists, miotics, carbonic anhydrase inhibitors, and prostaglandins). The number of days on each medication class in the previous 10 years was computed for the survival analysis, with the most recent 2 years excluded. Multivariable Cox regression was performed with time-varying exposure to each medication category as a predictor of interest with AD and all-cause dementia as outcomes during the 8-year sliding window. Days of exposure were converted to years for calculating hazard ratios.

Results : A total of 612 participants (mean age 75 [range 65-96], 63% female) with a glaucoma diagnosis, pharmacy records, and APOE genotype data were included. APOE ε4 allele was the strongest risk factor for dementia (HR = 1.66, 95% CI= 1.15- 2.38) and AD (HR = 1.81, 95% CI = 1.21- 2.71). Adjusting for other glaucoma medications, demographics, and smoking, each year of use of alpha-adrenergic agonists was associated with higher risks of developing dementia (HR = 1.20, 95% CI = 1.05-1.37).

Conclusions : Older adults with glaucoma and chronic exposure to alpha-adrenergic agonists appear to be at higher risk for developing all-cause dementia.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

 

Fig. 1. Participant Enrollment and Outcome Flowchart

Flow diagram showing the method for selecting participant medical records from the Adult Changes in Thought cohort study. The number of participants excluded from the study at each step is shown. The sum of participants on each medication exceeds the total number of participants because of combination therapy.

 

Table 1. Multivariable Model Results With All-Cause Dementia Outcome


CI = Confidence interval
*These analyses make the assumption that baseline hazard is equal across medications after accounting for demographic characteristics. The assumption of proportional hazards was tenable for all variables but gender