Increasing the diagnostic accuracy of medial temporal lobe atrophy in Alzheimer's disease

Medial temporal lobe (MTL) atrophy is considered to be one of the most important predictors of Alzheimer's disease (AD). This study investigates whether atrophy in parietal and prefrontal areas increases the predictive value of MTL atrophy in three groups of different cognitive status. Seventy-five older adults were classified as cognitively stable (n = 38) or cognitively declining (n = 37) after three years follow-up. At follow-up, the grey matter of the MTL, inferior prefrontal cortex (IPC), and inferior parietal lobule (IPL) was delineated on MRI scans. Six years later, a dementia assessment resulted in distinguishing and separating a third group (n = 9) who can be considered as preclinical AD cases at scan time. Ordinal logistic regressions analysis showed that the left and right MTL, as well as the right IPC and IPL accurately predicted group membership. Receiver Operating Curves showed that the MTL was best in distinguishing cognitively stable from cognitively declining individuals. The accuracy of the differentiation between preclinical AD and cognitively stable participants improved when MTL and IPL volumes were combined, while differentiating preclinical AD and cognitively declined participants was accomplished most accurately by the combined volume of all three areas. We conclude that depending on the current cognitive status of an individual, adding IPL or IPC atrophy improved the accuracy of predicting conversion to AD by up to 22%. Diagnosis of preclinical AD may lead to more false positive outcomes if only the MTL atrophy is considered.