Left hippocampal volume loss in Alzheimer's disease is reflected in performance on odor identification: A structural MRI study

Abstract

The very high sensitivity and specificity of odor identification tasks in discriminating between Alzheimer's patients and normals suggests that they reflect the presence of underlying neuropathology. Significant neuropathological changes are seen in areas critical to processing olfactory information, even in the early stages of Alzheimer's disease (AD). The current study was designed to investigate whether performance on olfactory tasks (odor threshold and odor identification) was related to volumetric MRI measures of mesial temporal areas central to olfactory information processing and important in the neuropathology of AD. Participants were 8 male and 5 female patients with probable AD, and 10 male and 12 female normal age-matched controls, diagnosed at the UCSD Alzheimer's Disease Research Center. The study investigated correlations between volumetric measures of hippocampus, the parahippocampal gyrus and the amygdala, and the psychophysical measures of olfactory function. Robust relationships were observed between mesial temporal lobe volumes and olfactory functional measures. The finding of a strong relationship between left hippocampal volume and performance on the odor identification task (r 5 .85) is compatible with a left-hemisphere superiority for verbally mediated olfactory tasks. The findings suggest a neural substrate for the breakdown in functional performance on verbally mediated odor identification tasks in Alzheimer's disease and suggest the utility of quantitative MRI measures and psychophysical performance in the assessment of AD. These results support the potential clinical utility of inclusion of odor identification tests in diagnostic batteries for detecting AD. (JINS, 2003, 9, 459-471.)

The human brain representation of odor identification

“…In the anatomic MTL ROI the statistical threshold for the OI event-related analysis was set to voxel level z ϭ 1.645 and P Ͻ 0.05, uncorrected for multiple comparisons, with a minimum of 9 continuous voxels. A more liberal statistical analysis was used in the OI analysis because of the lower number of observations in the event-related analysis, and due to the expected lower BOLD signal from the MTL due to susceptibility (Ojemann et al 1997(Ojemann et al , 2010) and coil effects (Kaza et al 2011). Moreover, entorhinal BOLD activity in response to odor stimulation has been shown to be substantially smaller than in piriform cortex and amygdala (Tabert et al 2007).…”

“…In the within-subject whole brain group level analysis of OI, entorhinal cortex activity was restricted to a small region and had a low z value, whereas many voxels were active with high max parameter estimate values in the anatomic entorhinal ROI. This discrepancy probably stems from a combination of MR technical (Kaza et al 2011;Ojemann 1997Ojemann , 2010 issues combined with local anatomic-physiologic characteristics of odor processing in the entorhinal cortex. In rats entorhinal neurons display odor selectivity (Young et al 1997), and thus both inter-and intraindividual activity to the identified odors may differ, thus contributing to a limited area of activation in the third level group analysis, although activity for each individual was high, as shown in the functional and anatomic entorhinal ROI analyses.…”

“…Since AD pathology is first observed in entorhinal cortex and subsequently in the hippocampus (Braak and Braak 1992), OI impairments may arise from medial temporal lobe (MTL) pathology. Indeed, the early and specific OI deficit in AD correlates with the number of tangles in entorhinal cortex and the hippocampus (Wilson et al 2007), and left hippocampal atrophy (Murphy et al 2003). Structural changes in the entorhinal cortex and hippocampus are also present in patients with schizophrenia (Baiano et al 2008;Bogerts et al 1985;Ebdrup et al 2010;Schultz et al 2009;Witthaus et al 2009), another group of patients with a specific OI deficit (Atanasova et al 2008;Moberg et al 1997Moberg et al , 2006Rupp 2010).…”

Feasibility of Multi-site Clinical Structural Neuroimaging Studies of Aging Using Legacy Data

“…The terms of this arrangement have been reviewed and approved by the University of California, San Diego in accordance with its conflict of interest policies. Work based on these MRI data have been published separately for studies performed within each site locally, including: UCSD (Jernigan et al 2001a, b;Murphy et al 2003;Jernigan and Fennema-Notestine 2004;Jernigan and Gamst 2005;Fennema-Notestine et al 2006); MGH/BWH (Killiany et al 2000, Killiany et al 2002; and WashU (Buckner et al 2004, Head et al 2005. Preliminary findings related to the combined data analysis were presented at the Society for Neuroscience 2005 meeting (FennemaNotestine et al 2005); work related to the combined data has not been published elsewhere.…”

“…Data from the UCSD and WashU were collected as part of on-going Alzheimer's Disease Research Center (ADRC) studies; data from MGH/BWH were collected through an on-going study of prodromal Alzheimer's disease. Previous within-site published studies based on these samples include: UCSD (Jernigan et al 2001a,b;Murphy et al 2003;Jernigan and Fennema-Notestine 2004;Jernigan and Gamst 2005;Fennema-Notestine et al 2006); MGH/BWH (Killiany et al 2000(Killiany et al , 2002; and WashU (Buckner et al 2004, Head et al 2005). Data from individuals over 60 years of age were included in this investigation.…”