Abstract
OBJECTIVE: To evaluate the risk of death in relation to incident antiepileptic drug (AED) use compared with nonuse in people with Alzheimer disease (AD) through the assessment in terms of duration of use, specific drugs, and main causes of death. METHODS: The MEDALZ (Medication Use and Alzheimer Disease) cohort study includes all Finnish persons who received a clinically verified AD diagnosis (n = 70,718) in 2005-2011. Incident AED users were identified with 1-year washout period. For each incident AED user (n = 5,638), 1 nonuser was matched according to sex, age, and time since AD diagnosis. Analyses were conducted with Cox proportional regression models and inverse probability of treatment weighting (IPTW). RESULTS: Nearly 50% discontinued AEDs within 6 months. Compared with nonusers, AED users had an increased relative risk of death (IPTW hazard ratio [HR], 1.23; 95% confidence interval [CI], 1.12-1.36). This was mainly due to deaths from dementia (IPTW HR, 1.62; 95% CI, 1.42-1.86). There was no difference in cardiovascular and cerebrovascular deaths (IPTW HR, 0.98; 95% CI, 0.67-1.44). The overall mortality was highest during the first 90 days of AED use (IPTW HR, 2.40; 95% CI, 1.91-3.03). Among users of older AEDs, relative risk of death was greater compared to users of newer AEDs (IPTW HR, 1.79; 95% CI, 1.52-2.16). CONCLUSION: In older vulnerable patients with a cognitive disorder, careful consideration of AED initiation and close adverse events monitoring are needed.