Trials will be done of therapeutic agents whose possible efficacy in Alzheimer disease has been suggested by laboratory research in the Dementia Research Service. The ability of high dose thiamine to slow the progress of Alzheimer disease will be tested, because of evidence for deficiencies of thiamine-dependent enzymes in Alzheimer brain and because of a statistically significant effect of high dose thiamine on cognition in a pilot clinical study. Patients receiving the diagnoses of probable or possible Alzheimer disease over the course of 2 years on the Geriatric Evaluation Service (GES) will be given the opportunity to take part in this out-patient trial. Each participant will be given either high dose thiamine (1 gm 3 times a day) or niacinamide placebo (3 times a day in identical capsules) for one year. Since the aim is to test the effect on the natural history of a progressive disorder, the design will be double-blind A:B rather than cross-over. Patients will be followed at 4 month intervals, and tested on mental status tests (MMS, MSQ, and ADAS) and behavioral ratings (Blessed and Haycox). Outcome variables will be mental status and behavioral scores. Statistical analysis will be by paired t-test. An open, dose finding study of the effect of 3,4-diaminopyridine will be done. Fifteen patients with probable Alzheimer disease will be recruited, from the GES population. Each patient will be admitted to the Burke Rehabilitation Center for 3 weeks. The first week will be an equilibration period, to stablilize these slowly adaptable patients cognitively and behavioraly to their hospital setting. During the second and third weeks, doses of 3,4- diaminopyridine will be progressively increased from 5 mg twice a day to 15 mg 3 times a day (orally, in gelatin capsules with sucrose carrier). Patients will be tested at regular (2-4 day intervals) with the MMS, ADAS, and Blessed behavioral scales. These scores will be the outcome variables. Statistical analysis will be by an ANOVA multiple comparisons procedure.
Showing the most recent 10 out of 41 publications
