Hippocampal mean diffusivity for the diagnosis of dementia and mild cognitive impairment in primary care.

Background: Hippocampal mean diffusivity (MD) measured by Diffusion-Tensor Imaging is a promising diagnostic marker for Mild Cognitive Impairment (MCI) and dementia. Its performance has yet to be evaluated in primary care patients, who vary systematically from patients visiting specialized care settings. Objective: We assessed the diagnostic accuracy of hippocampus diffusivity for detecting MCI and dementia in a sample recruited from primary care, compared to a sample from specialized care. Method: One sample was recruited from a primary care intervention trial (DelpHi-MV) (n=70), and the other sample was recruited from our memory clinic (n=70). The samples were matched pairwise for diagnosis, MMSE, age, gender, and education. They included dementia patients, MCI patients and healthy subjects. Mean MD was calculated for the left and right hippocampus, corrected for partial volume effects. Within each sample, left or right hippocampal MD served as predictor for diagnostic group in logistic regressions, which were additionally controlled for white matter lesions. Results: In the primary care sample, hippocampal MD detected dementia with high cross-validated accuracy (left: AUC=.92; right: AUC=.85), but did not classify MCI with an accuracy above chance (left: AUC=.58; right: AUC=.44). In the memory clinic sample, hippocampal MD classified both dementia (left: AUC=.91; right: AUC=.91) and MCI (left: AUC=.86; right: AUC=.83) with high cross-validated accuracy. Conclusion: Hippocampal MD supported the identification of dementia but did not contribute to the detection of MCI in the primary care patient population.