Even in nonfamilial cases of dementia there is some evidence of a genetic factor. This may be linked to defective expression of neurofilament protein and also abnormal phosphorylation of cytoskeletal proteins. In this respect there may be a link with accumulation of tangles and amyloid which have some degree of homology. It may be speculated that neurons containing tangles or undergoing granulovacuolar degeneration would not be able to release trophic factors and that transneuronal degeneration would result. However, the environmental or aetiological factors associated with Alzheimer’s disease are not known. Although there has been a failure to transmit Alzheimer’s disease to primates, it is possible that as in postencephalitic Parkinson’s disease virus may be implicated at some stage in the pathogenesis. Finally, free radical formation has been considered as an alternative mechanism for death of large neurons within the CNS. Although tangles are found in several other dementing conditions (e.g. dementia puglistica, Parkinson-dementia complex of Guam), Alzheimer-type plaques and tangles are not invariably found in cases of cognitive deficit. For example, in dementia of Parkinson’s disease there is a low neuritic plaque count and normal population of tangles. In addition, memory loss is not necessarily associated with defects in the cholinergic system and/or loss of nucleus basalis nerve cells. We have proposed that damage to or loss of cortical cells may be a more general finding in dementing illness.