Symptomatic and Preventive Medication Use Before and After Alzheimer Disease Diagnosis: A 10-Year Matched Cohort Study

Abstract

ObjectiveTo investigate longitudinal changes in symptomatic and preventive medication use among community-dwelling people with and without Alzheimer disease (AD) 5 years pre- and post-AD diagnosis. DesignRetrospective matched cohort study. Settings and ParticipantsThe sample comprised 58,496 people with a geriatrician/neurologist-verified AD diagnosis matched 1:1 for age, sex, and region to people without AD in Finland. MethodsMedication dispensing data were obtained from the Finnish Prescription Register. Prevalence of symptomatic and preventive medication use was evaluated every 6 months from 5 years pre- to post-AD diagnosis. Longitudinal changes in medication use between people with and without AD were compared using ordinal logistic regression. ResultsDuring the 5 years pre- and postdiagnosis, there were differences in symptomatic (P < .001) and preventive (P = .006) medication use between people with and without AD. Over the 5 years prediagnosis, prevalence of symptomatic and preventive medications increased in both people with and without AD. During the 1 year prediagnosis, people with AD had a higher increase in use of ≥3 symptomatic medications (+4.4 vs +2.2%) and ≥3 preventive medications (+6.4 vs +2.9%) compared to people without AD. Over the 5 years postdiagnosis, symptomatic medication use plateaued in both people with and without AD. Meanwhile, people using ≥3 preventive medications decreased (−6.0%) in those with AD, but increased (+6.1%) in those without AD. During the follow-up period, people with AD had a larger absolute percentage increase in prevalence of antipsychotics (+22.7 vs +1.8%) and antidepressants (+19.1 vs +5.0%) than people without AD. During the same period, paracetamol and calcium supplement use increased by 31.1% and 20.4%, respectively, among people with AD. The largest absolute percentage decrease in prevalence of preventive medications over the 5 years postdiagnosis were β-blockers (−9.8%) and statins (−7.0%) in people with AD. Conclusions and ImplicationsAt the point of and following diagnosis, there were population-level changes in medication use among people with AD. Medication assessments during this period appear to coincide with discontinuation of preventive medications whereas minimal changes were observed in symptomatic medication use.