Cholinergic Neurotransmission Has Different Effects on Cerebral Glucose Consumption and Blood Flow in Young Normals, Aged Normals, and Alzheimer's Disease Patients

Cerebral blood flow (CBF) and glucose consumption (GC) are both tracers of brain metabolic activity used to image the human brain in vivo. To know if both tracers reacted in the same manner when brain cholinergic neurotransmission was activated, CBF and GC were measured in young normals (YN), aged normals (AN), and Alzheimer's Disease patients (AD) using positron emission tomography (PET), H2 15O, and 18F-FDG. Each subject was studied twice, under placebo and physostigmine, in randomized order and blind fashion using the maximal tolerated dose of physostigmine individually determined. Under physostigmine CBF increased significantly (P = 0.0007) in posterior regions of the cerebral cortex and in the subcortical structures. Inversely, GC was decreased significantly in most regions. The largest decrease was seen in the prefrontal region of the cerebral cortex (P < 0.0001). Significant regional decreases were registered in all three groups of subjects, but were larger in AD than in controls. Looking at the absolute values of prefrontal cortex metabolism we found no correlation (r = 0.04) between the responses of CBF and GC. After normalization of the regional values for the mean we found a significant positive correlation between the responses of CBF and GC (r = 0.71, P < 0.0001). These findings suggest two components in the CBF response to physostigmine: one metabolic, depressive, and regional which follows the GC response; and one vascular, larger, diffuse, and opposite in direction to the metabolic component. These results have implications for the interpretation of CBF values as tracer of brain metabolic activity when brain cholinergic neurotransmission is manipulated.