Objective: To compare rates of cognitive decline between probable Alzheimer's disease (AD) patients treated with long-duration cholinesterase inhibitors (ChE-Is) and those who remained untreated.
Background: ChE-Is, including donepezil and tracrine, have shown beneficial effects on cognition and global functioning in patients with AD. The duration of these benefits is unknown because the longest double-blind placebo-controlled studies reported were only approximately 6 months long. Ethical concerns regarding randomization of patients to placebo for long periods make it difficult to undertake trials of longer duration.
Methods: We identified patients in 4 AD centers who were or were not consistently treated with ChE-Is and who had demographic, psychometric and follow-up data. We compared 205 ChE-I-treated and 218 untreated AD patients on baseline variables hypothesized to differ between these groups, on baseline Mini Mental Status Examination (MMSE) scores and on rates of MMSE change at 1 year. The analysis was performed initially with all ChE-I-treated patients as a single group versus untreated subjects, and then with donepezil versus untreated subjects and tacrine versus untreated subjects.
Results: As expected, treated and untreated patients differed with respect to age, education, ethnicity, percentage of community dwelling and exact days of follow-up (ANOVA and chi2) in several comparisons, but did not differ on baseline MMSE score. These baseline variables were highly intercorrelated. MMSE scores declined significantly more slowly after 1 year of ChE-I treatment compared to untreated patients (p = 0.05) after controlling for baseline differences in age, education, ethnicity and percentage of community dwelling. Slowing of decline was significant in the donepezil-treated patients (p = 0.007) but not in the tacrine-treated group (p = 0.33).
Conclusions: This study, utilizing concurrent, nonrandomized controls, suggests that donepezil continues to have efficacy over at least the first year of therapy. Other studies are needed to determine whether the benefits are maintained beyond 1 year.
Copyright 2001 S. Karger AG, Basel