We studied a possible correlation between autonomic cardiac activity and the level of the red blood cell acetylcholinesterase (AChE) in patients with probable Alzheimer disease (AD). The influence of cholinesterase inhibitor treatment on this autonomic activity was evaluated. Twelve patients satisfying the NINCDS-ADRDA criteria of probable AD and 10 healthy controls were studied. Autonomic cardiac activity was evaluated by means of power spectral analysis (PSA) of heart rate variability (HRV) using an autoregressive algorithm on 250 consecutive electrocardiographic R-R intervals. All patients received oral eptastigmine, a new cholinesterase inhibitor, for 1 month. Before treatment, a simultaneous recording of the electrocardiographic and respiratory activities was performed at rest and subsequently during head-up tilt test at 700. Recording was repeated on the last day of treatment. The level of AChE activity during each recording was also evaluated. Spectrum power was calculated in three main frequency bands: high frequency (HF), 0.15-0.4 Hz; low frequency (LF), 0.04-0.15 Hz; very low frequency (VLF), <0.04 Hz. In addition, we calculated the total spectrum power (TSP) and the LF/HF ratio. The TSP and the absolute value of each spectral component were significantly lower in AD patients than in controls. In contrast with controls, AD patients did not show any significant change before treatment in either the LF and HF components or in the LF/HF ratio during the tilt test. However, the modification in the LF component, induced by tilting, showed a significant correlation with the level of AChE activity (p < 0.03). During the tilt test, the treatment caused changes in LF and HF components and in the LF/HF ratio similar to those observed in controls. These results suggest that the presence of autonomic cardiac dysfunction in AD patients might be due to a cholinergic deficit in the peripheral autonomic nervous system. Some aspects of this autonomic dysfunction can be normalized by cholinesterase inhibitor treatment.