Physostigmine effects in Alzheimer's disease: relationship to dementia severity.

Eleven demented patients were administered .004, .009, and .013 mg/kg physostigmine intramuscularly, and placebo, double-blind, in Phase 1. The most effective dose, in terms of showing the best memory score as compared to saline, was repeated during Phase 2. Five patients improved their verbal memory scores in both Phases 1 and 2 after the most effective dose of physostigmine; these five "responders" were found to be significantly more demented than the six "nonresponders." Drug-induced increases in memory scores were significantly correlated with illness severity. Intrusions, which were not a factor in selection of the most effective dose, were reduced in the group as a whole, with the responders showing the most improvement and the nonresponders the least. The association between physostigmine effect and degree of dementia suggests to us that the severe cases may have more permeable blood-brain barriers, and that drug availability to the brain is an important factor in evaluating treatment of SDAT with cholinergic substances.