Recent neuroanatomical and functional neuroimaging studies indicate that the anterior part of the hippocampus, rather than the whole structure, may be specifically involved in episodic memory. In the present work, we examined whether anterior structural measurements are superior to other regional or global measurements in mapping functionally relevant degenerative alterations of the hippocampus in Alzheimer's disease (AD).
Twenty patients with early AD (MMSE 25.7 ± 1.7) and 18 healthy controls were studied using magnetic resonance and diffusion-tensor imaging. Using a regions-of-interest analysis, we obtained volumetric and diffusivity measures of the hippocampal head and body-tail-section as well as of the whole hippocampus. Detailed cognitive evaluation was based on the CERAD battery.
All volumetric measures as well as diffusivity of the hippocampus head were significantly (p < 0.01) altered in patients as compared to controls. In patients, increased left head diffusivity significantly (p < 0.01) correlated with performance on free delayed verbal recall test (DVR) (r = −0.74, p = 0.0002) and with the CERAD global score. Reduced volume of the left body-tail was also associated with performance on DVR (r = 0.62, p = 0.004). Stepwise regression analyses revealed that increased left head diffusivity was the only predictor for performance on DVR (R2 = 52%, p < 0.0005).
These findings suggest that anterior hippocampus diffusivity is more closely related to verbal episodic memory impairment than other regional or global structural measures. Our data support the hypothesis of functional differentiation in general and the specific role of the anterior hippocampus in episodic memory in particular. Diffusivity measurements might be highly sensitive to functionally relevant degenerative alterations of the hippocampus.
There is strong empirical evidence that the hippocampus plays a crucial role in encoding and retrieving new memories (Squire, 1992). Yet, functional specialization within the hippocampus remains controversial. To explain hippocampal function in relation to the cytoarchitectural divisions, more attention has focused on the horizontal axis than on the longitudinal axis of the hippocampal formation. During the last 15 years, however, it has become clear that functional specialization along the longitudinal axis is fundamental to hippocampal function. This evidence has come primarily from lesion experiments on animals (Moser et al., 1993 and Moser et al., 1995), but it has rapidly been supported by functional neuroimaging studies in human (see Schacter & Wagner, 1999 for review and references). More specifically, functional magnetic resonance imaging (fMRI) findings indicated a relationship between the anterior part of the hippocampus and associative memory. Recently, Chua, Schacter, Rand-Giovannetti, and Sperling (2007) highlighted a specific role of the anterior hippocampal formation in successful memory encoding.
The hippocampal formation is the first brain structure to be affected by Alzheimer's disease (AD) (Braak & Braak, 1991) and, correspondingly, prospective studies have found that memory deficits are the first clinical signs of AD (Jacobs et al., 1995). Therefore, this limbic area has been the focus of special research attention. AD-associated pathological alterations of the hippocampus at very early stages of the disease have been systematically demonstrated using various neuroscience techniques. Among neuroimaging methods, MRI-based hippocampal volumetry is likely to be the most widespread diagnostic tool (de Leon et al., 2007). Generally, regional cerebral volume reduction is regarded to reflect neuronal cell loss. Accordingly, total numbers of neurons in the hippocampus strongly correlate with MRI-determined hippocampal volume in AD (Bobinski et al., 2000). Numerous studies have consistently demonstrated hippocampal atrophy not only in patients with very mild AD, but also in subjects with mild cognitive impairment (MCI) whose cognitive deficits are limited to the memory domain (i.e., amnestic MCI, aMCI) (Convit et al., 1997 and Wolf et al., 2001). Furthermore, hippocampal size reduction predicted conversion to AD in aMCI and in normal elderly (Jack et al., 2005 and Wolf et al., 2003).
While a critical role of hippocampal atrophy in early diagnosis of AD is apparent, fewer consistencies exist in respect of its functional implications. The association between hippocampal volume and episodic memory performance (as assessed by free delayed verbal recall test, DVR) is controversial. Presence (e.g., Devanand et al., 2007 and Laakso et al., 2000) as well as absence (e.g., Basso et al., 2006 and Walhovd et al., in press) of a (positive) correlation has been reported in AD and aMCI. Notably, within one study paradigm, Fjell et al. (2008) found a positive association between the hippocampal volume and performance on DVR in one MCI sample, but not in another. This suggests that the situation is complex and factors beyond methodological differences (e.g., vascular, hormonal) (Geuze, Vermetten, & Bremner, 2005) may account for the variability in volumetric measurements in general and their relationship with memory in particular. Overall, the existing evidence indicates that hippocampal atrophy might not be an ideal marker of hippocampus dysfunction.
Diffusion-tensor imaging (DTI) measures the random, unidirectional motion of water molecule protons in biologic tissue (Le Bihan et al., 2001). The mean diffusivity (MD) is a measure of randomized mean water diffusion. As any neurodegenerative process is accompanied by a progressive loss of barriers that restrict water molecule motion (e.g. neuronal loss), it can be sensitively detected by DTI as pathologically elevated MD (Beaulieu, 2002). Accordingly, early DTI studies documented increased MD in hippocampal regions of patients with AD and aMCI (Fellgiebel et al., 2004a and Kantarci et al., 2001). Later our group showed that increased hippocampal diffusivity was associated with impaired verbal memory performance in aMCI (Müller et al., 2005). Furthermore, Kantarci et al. (2005) and Fellgiebel et al. (2006) found that hippocampal diffusivity was superior to hippocampal volume measurements in predicting conversion of aMCI to dementia.
In the present study, we examined whether anterior hippocampal diffusivity or volume measurements are superior to global or posterior measurements in mapping functionally relevant degenerative alterations of the hippocampus in early AD. Guided by the concept of functional differentiation, we hypothesized that AD-associated increased hippocampal head diffusivity would be a more sensitive marker of episodic memory impairment than other regional or global diffusivity or volumetric measurements.
