Abstract

High alcohol consumption for long periods of time causes significant hippocampal neurodegeneration in rodents. A single study using neuronal density measures has reported similar findings in humans. The present study aims to substantiate these findings in human alcoholics using unbiased stereological techniques. Both amnesic (n = 5) and nonamnesic (n = 7) chronic alcoholics were selected and compared with nonalcoholic controls (n = 8) and patients with marked memory loss and hippocampal neurodegeneration caused by Alzheimer's disease (n = 4). Hippocampal volume was significantly reduced in the alcoholics and in patients with Alzheimer's disease. However, in alcoholics the volume reduction occurred exclusively in the white matter, whereas both the gray and white matter were reduced in the patients with Alzheimer's disease. Neuron loss occurred exclusively from the CA1 and subiculum subregions of the hippocampus in Alzheimer's disease. No neuron loss occurred from any subregion of the hippocampus in alcoholics. There were no correlations with age and any of the volume or neuron number measures. Hippocampal volume correlated with brain volume and with the regional gray and white matter volumes within the hippocampus. In addition, hippocampal gray matter volume correlated with the number of CA1 pyramidal neurons. These results do not support the theory that chronic alcohol consumption is neurotoxic to hippocampal pyramidal neurons in humans. Further, the present results suggest that changes observed in rodent models of alcoholism do not parallel those observed in humans, questioning the validity of such models.