Our results suggest that neocortical Alzheimer’s disease pathology remains significantly correlated with dementia throughout the lifespan. In the oldest subjects in our cohort the relationship between dementia and Alzheimer’s disease pathology changes somewhat, as dementia appears in subjects with low Alzheimer’s disease pathology scores—something that is uncommon in subjects under 90. Moreover, the amount of neurofibrillary tangle pathology needed to cause dementia increases with age. This difference between younger and older subjects, however, does not lessen the impact of Alzheimer’s disease pathology on cognition in elderly subjects with high Alzheimer’s disease pathology scores. Moreover, elderly subjects with low Alzheimer’s disease pathology scores have additional, well described, pathologies to account for their cognitive status, suggesting that dementia in elderly subjects is not dependent on new, unknown, processes. Importantly, atherosclerosis, with and without cerebral infarction, emerges as an important aetiology of dementia in this group, a finding that has recently been explored by our group (Troncoso et al., 2008
). Other pathologies that we have not explored, such as Lewy body pathology, may also play an important role in dementia in the elderly.
Our results are based strictly on neocortical Alzheimer’s disease pathology and not hippocampal pathology. The importance of neocortical Alzheimer’s disease pathology in the clinical manifestations of Alzheimer’s disease was codified by the seminal publications of Braak and Braak (1991)
and Mirra et al. (1991)
, which defined neocortical neurofibrillary tangles and neocortical neuritic amyloid plaques as the key elements in correlating Alzheimer’s disease pathology with dementia. These techniques for scoring Alzheimer’s disease pathology are semi-quantitative and it is hoped that more formal quantitative analyses of Alzheimer’s disease pathology will be even more informative. While hippocampal pathology and atrophy are clearly important in early disease pathogenesis (Reitz et al., 2009
), dementia is a complex clinical phenomenon, dependent on multiple components of the cerebral cortex (Buckner et al., 2005
). Indeed our recent work looking at the importance of infarcts in dementia showed that only neocortical infarcts played a role in cognitive decline (Troncoso et al., 2008
Our study contrasts with a previous retrospective study using brain bank specimens suggesting that Alzheimer-type pathology has a stronger relationship with dementia in the younger-old compared to the oldest-old (Haroutunian et al., 2008
). However in this study, the authors still find significant differences in neocortical Alzheimer’s disease pathology between demented and non-demented subjects even into very advanced age. A limited number of cognitively normal subjects in the older age groups and a reliance on clinical dementia rating scores rather than consensus conferencing for diagnostic determination of dementia also limit this analysis. More recently, Savva et al. (2009)
, in a study from the Medical Research Council Cognitive Function and Ageing Study (MRC-CFAS), found that the association between Alzheimer’s disease pathology and dementia was stronger in younger participants than older participants. Specifically, their data suggested that Alzheimer’s disease pathology in the form of neuritic plaques did not differ significantly between demented and non-demented subjects in their mid-nineties and that the effect of neurofibrillary tangles on cognition was significantly attenuated compared to younger subjects. There are several important differences between our study and the MRC-CFAS. First, hippocampal Alzheimer’s disease pathology in the MRC-CFAS shows little difference between demented and non-demented elderly subjects, especially the amount of neurofibrillary tangles, suggesting a potential ceiling effect in the hippocampus. The MRC-CFAS Alzheimer’s disease pathology scores in neocortical areas of subjects in their early nineties show clear differences between those who are demented and those who are not. Neocortical pathology scores in demented and non-demented MRC-CFAS subjects in their late nineties, in contrast, do show overlap. Importantly, the MRC-CFAS investigators analysed the relationship between dementia and CERAD or Braak Alzheimer’s disease pathology scores by dividing subjects into just two groups (those with high and those with low Alzheimer’s disease pathology scores), rather than analysing the full range of scores, so that their analysis could not detect changes in the amount (threshold) of each of these pathologies which correlated with dementia as a function of age. The characteristics of subjects enrolled in our study and in the MRC-CFAS are also different. While the mean ages in the two studies are similar, all of our subjects were cognitively normal at entry into the study and have high education levels. In contrast, the MRC-CFAS is a population-based sample of urban and rural areas of England and Wales that contain institutions that care for elderly subjects. Many subjects were demented on entry into the MRC-CFAS and the final median Mini-Mental State Examination score of the normal subjects in the MRC-CFAS is 25, while in our study it is 29 (). We feel that the higher educational level and cognitive functioning of our cohort at baseline is more likely to emphasize the effects of Alzheimer’s disease pathology and may be more representative of the future US and European populations (National Centre for Education Statistics, 2008
; Christensen et al., 2009
) that will be better educated and have greater access to health care than generations born in the early 20th century; our cohort does not reflect the demographics of populations at risk for dementia throughout the developing world however (Prince and Jackson, 2009
In conclusion, our findings indicate that Alzheimer’s disease pathology remains a significant predictor of dementia throughout the lifespan. Although other factors contribute to the cognitive decline of patients suffering from Alzheimer’s disease and dementia, including atherosclerosis, Alzheimer-type brain pathology continues to play an important role even in the most elderly.