Distinct patterns of olfactory impairment in Alzheimer's disease, semantic dementia, frontotemporal dementia, and corticobasal degeneration

Abstract

Performance on tests of odour discrimination, naming, and matching was compared in patients with four distinct forms of neurodegenerative disease: Alzheimer's disease (AD), semantic dementia (SD), frontotemporal dementia (FTD), and corticobasal degeneration (CBD). The SD patients were found to have a severe impairment of identification from olfaction despite having normal discrimination, consistent with the multimodal semantic impairment characteristic of this patient group. The AD patients’ poor odour discrimination suggests that a perceptual impairment is the root of their poor odour identification. Mild impairments in odour identification observed in FTD and CBD are consistent with their generalised executive dysfunction. The findings illustrate that breakdown in olfaction can occur at a perceptual or semantic level, analogous to the distinction between apperceptive and associative forms of deficit in the visual and auditory modalities. The findings add further insights into the nature of the semantic deficit in SD by exploring a hitherto neglected modality and may have relevance in explaining the altered eating habits commonly associated with SD.

Introduction

The purpose of the present study was to examine smell perception and recognition in four neurodegenerative disorders: frontotemporal dementia (FTD), semantic dementia (SD), Alzheimer's disease (AD), and corticobasal degeneration (CBD). There are anatomical, neuropsychological, and clinical reasons why investigation of smell in these disorders is important—yet it has never been done so before in FTD and SD.

In the olfactory system, axons of the mitral cells of the olfactory bulb course through the olfactory tracts and then via the lateral olfactory striae terminate in the amygdala and the “primary” olfactory cortex of the uncus (Nauta & Feirtag, 1986). Projections then pass to the “association” olfactory cortex of the parahippocampal gyrus and entorhinal area. The parahippocampal gyrus receives fibres via the cingulum from the cingulate gyrus and projects into the hippocampus. These “primary” and “associational” areas are referred to as the pyriform cortex. The amygdala receives connections from middle and inferior temporal gyri (inferior association cortex), known to be critical for semantic processing, and in turn projects into the hypothalamus. Damage to pyriform cortex would be expected to lead to impairment in odour at a sensory or perceptual (apperceptive) level, whereas damage to temporal cortical-amygdala connections would be expected to lead to impairment at a semantic (associative) level. Lesion and functional neuroimaging studies (Van Hoesen, Parvizi, & Chu, 2000) indicate that the medial orbitofrontal cortex – which receives input from the piriform cortex – is also implicated in several olfactory functions, including olfactory discrimination.

In neurodegenerative disease, olfaction has been examined most extensively in AD (Doty, Shaman, & Dann, 1984; Kesslak et al., 1988; Knupfer & Spiegel, 1986; Koss, Weiffenbach, Haxby, & Friedland, 1988; Larsson et al., 1999; Lehrner, Brücke, Dal-Bianco, Gatterer, & Kryspin-Exner, 1997; McCaffrey, Duff, & Solomon, 2000; Rezek, 1987; Royet et al., 2001; Schiffman, Clark, & Warwick, 1990; Serby, Larson, & Kalkstein, 1991). Studies have examined threshold, familiarity, feature discrimination, identification, and recognition (see Liberini & Parola, 2001; Martzke, Kopala, & Good, 1997; Mesholam, Moberg, Mahr, & Doty, 1998 for review). Despite this extensive literature, studies have not always yielded uniform results (Liberini & Parola, 2001; Martzke et al., 1997, Mesholam et al., 1998; Parola & Liberini, 1999). For example, olfactory threshold has been reported to be preserved in AD by some authors (Koss et al., 1988, Serby et al., 1991) but not others (Knupfer & Spiegel, 1986; Morgan, Nordin, & Murphy, 1995; Rezek, 1987). Some studies have reported an olfactory perceptual impairment (Doty, Reyes, & Gregor, 1987), whereas others have not (Koss et al., 1988, Royet et al., 2001).

In contrast to AD, there have been no studies of olfaction in SD and FTD, two clinical syndromes of frontotemporal lobar degeneration (Neary et al., 1998). SD is a multimodal disorder of meaning, in which patients have difficulty recognising the significance of words, objects, faces, and nonverbal sounds and tastes, despite normal perception of those stimuli (Bozeat, Gregory, Lambon Ralph, & Hodges, 2000; Coccia, Bartolini, Luzzi, Provinciali, & Lambon Ralph, 2004; Snowden, Neary, & Mann, 1996; Snowden, Thompson, & Neary, 2004). It would be anticipated that there would be a parallel loss of meaning for odours but no impairment in their detection and discrimination. This would be predicted, not only on cognitive grounds based on the findings in other modalities, but also on the basis of the anatomical distributions of pathology. In SD (Snowden et al., 1996) there is severe atrophy of the amygdala and middle and inferior temporal gyri with relative sparing of the hippocampal formation (hippocampus, entorhinal cortex, and parahippocampal gyrus). Thus, impairment in odour recognition ought to occur in SD because of the underlying degradation of the meaning associated with each odour (arising from the anterior, inferiolateral atrophy of this disorder).

FTD is a disorder of behaviour, characterised by striking personality change, breakdown in social conduct and executive impairments (Gustafson, Brun, & Risberg, 1987; Neary et al., 1998). Atrophy is most prominent in the frontal lobes, although it invariably involves the anterior temporal cortex to some degree (Neary et al., 1998, Snowden et al., 1996). One might predict, therefore, some of the deficits of SD, albeit to a lesser degree. However, in view of the effortful search involved in odour identity, performance might also potentially be compromised by patients’ executive impairments. In addition, orbitofrontal atrophy might also impact directly upon smell perception.

The study of odour perception and identification has additional relevance in SD and FTD in that it might potentially shed light on some of the characteristic behavioural changes of these disorders, namely alterations in dietary habits (Bathgate, Snowden, Varma, Blackshaw, & Neary, 2001; Bozeat, Gregory et al., 2000; Ikeda, Brown, Holland, Fukuhara, & Hodges, 2002; Miller, Darby, Swartz, Yener, & Mena, 1995; Snowden et al., 2001). SD patients commonly exhibit food fads and are over-selective in what they eat (Bathgate et al., 2001, Snowden et al., 2001), although in the later stages of illness they may attempt to eat inanimate objects. This behaviour, recorded more frequently in SD than in FTD or AD (Bathgate et al., 2001, Snowden et al., 2001) is consistent with the Klüver-Bucy syndrome, which has been associated with temporal lobe pathology. Evidence for the loss of semantic knowledge of odours would lend support to the view that alterations in oral behaviour in SD are crucially linked to patients’ semantic impairment. That is, patients are wary of what they eat as they no longer have olfactory cues by which to judge edibility or freshness of foodstuffs and are prone to errors in differentiating edible from inedible.

FTD patients exhibit a different pattern of eating behaviour. They commonly eat excessively and indiscriminately (Neary et al., 1998, Snowden et al., 1996), yet frequently do so only when food is present and do not actively seek it out. This raises the possibility that the behaviour is related to a general stimulus-boundness and failure of inhibitory control rather than to smell or taste senses per se. The finding of preserved odour perception would lend support to the view that indiscriminate eating in FTD is not a consequence of primary olfactory deficits.

CBD provides a valuable reference disorder to AD, SD and FTD in that it is a disorder predominantly of the basal ganglia and premotor and parietal cortex, and thus spares the temporal cortex. Olfaction might therefore be expected to be preserved. Indeed the relatively few available studies of olfaction in CBD (Müller, Reichmann, Livermore, & Hummel, 2002; Wenning et al., 1995) found no olfactory impairment. It has been suggested (Hawkes, 2003) that preserved olfaction may provide a useful diagnostic aid in distinguishing CBD from other Parkinsonian disorders. Given the low number of studies, however, further data on olfaction in this patient group are merited.