THE HYPOTHESIS that aluminum is involved in the pathogenesis of Alzheimer disease (AD) rests on 3 connected issues: the toxicity of aluminum, elevation of levels of aluminum in the brains of patients with AD, and increased prevalence of AD in geographic regions where levels of aluminum in drinking water are high. Although the title of this article refers only to the third issue, all 3 are so interlocked that they must be examined together.
The aluminum hypothesis originated in 1965 with the discovery that injections of aluminum salts in the brains or cerebrospinal fluid of rabbits induce a progressive encephalopathy that is associated with the development of lesions reminiscent of neurofibrillary tangles.1,2 The scientific community's excitement in response to this discovery waned as more sophisticated electron microscopy and the development of immunohistochemistry showed that aluminum-induced tangles are made up of bundles of normal neurofilaments, in sharp contrast with neurofibrillary tangles in AD in which paired helical fibrils mostly consist of abnormal tau proteins.3 Furthermore, it was demonstrated that changes induced by aluminum predominantly affect motor neurons and spare the hippocampus, a reversal of the lesion topography in AD.4 Dialysis encephalopathy syndrome, a disease associated with excessive amounts of aluminum in tap water, is the human counterpart to the rabbit's aluminum-induced encephalopathy.5 This syndrome, characterized by dysarthria, myoclonus, seizures, and markedly abnormal electroencephalographic findings, bears no similarity to AD.6,7 An examination of the brains of patients with this disease reveals a total absence of neuritic senile plaques and tau-immunoreactive neurofibrillary tangles; some reports describe neurofibrillarylike structures made of neurofilaments, similar to those reported in the brains of the rabbits.8 In patients with dialysis encephalopathy syndrome, levels of aluminum are elevated in serum and the brain; its distribution in the latter bears no correlation to that of diffuse β-amyloid deposits in elderly patients.9 Parenteral aluminum is a neurotoxin, as are at least a dozen other elements and hundreds if not thousands of compounds. However, the mechanisms by which parenteral aluminum damages the brain are radically different from those involved in the pathogenesis of AD.