Because the clinical picture of Alzheimer disease (AD) is often difficult t
o discriminate from other dementing illnesses, the diagnosis of AD requires
neuropathological confirmation. However, for the pathological diagnosis of
AD, there are no unanimously accepted criteria. The three currently used s
ets of pathological criteria (Khachaturian: Khachaturian, Arch Neural 1985;
42:1097-105; Tierny: Tierney et al., Can J Neurol Sci 1986;13:424-6; CERAD:
Mirra et al., Neurology 1991;41:479-86) for the disease differ from each o
ther considerably. We applied these criteria to the first 43 consecutive su
bjects (37 demented) with no neuropathology other than AD-type pathology fr
om autopsies after longitudinal prospective clinical study in the Oxford Pr
oject to Investigate Memory and Ageing (OPTIMA). The results show that the
CERAD category of definite AD corresponds closely with the cases that fulfi
ll Tierney A3 inclusion criteria for AD. The combined CERAD categories of p
ossible, probable, and definite AD correspond closely to cases fulfilling K
hachaturian criteria for AD. The influence of a clinical diagnosis of demen
tia when Khachaturian and CERAD criteria were applied was considerable beca
use between 93% and 90.7% of patients would have been categorized different
ly depending on whether clinical dementia was present or absent.