The paper by Dominic Holland et al. reporting that AD-associated decline slows with advancing age is interesting. Although no autopsy confirmation was available in this study, these data agree with neuropathologic studies in the oldest-old (Haroutunian et al., 2008).
However, there are some limitations of this study, which only in part have been mentioned by the authors:
1. The importance of confounding pathologies, in particular, cerebrovascular lesions including the importance of CAA (the authors only mentioned microvascular pathology), Lewy body pathology, argyrophilic grain disease, hippocampal atrophy, etc., which are frequent in aged human brains (see "mixed dementia," e.g., Kovacs et al., 2008; Jellinger and Attems, 2011). It should be emphasized that up to two-thirds of aged human brains contain non-AD type pathologies (Nelson et al., 2007, and others).
2. A high percentage of demented persons aged 80+ do not meet the morphological criteria of AD and are classified "dementia of unknown etiology" (Crystal et al., 1988; Jellinger, 2001; Corrada et al., 2012),...