As presented previously in this chapter, the brain mechanisms implicated in the olfactory function have been studied considerably over the past decades. Thus, the peripheral and the brain structures involved in different olfactory abilities (acuity, discrimination, identification, memory, familiarity, intensity, and pleasantness) are now known. Several psychophysical, neurophysiological and neuroimaging studies demonstrated that these peripheral and the brain structures are differentially affected in several psychiatric, neurological and neurodevelopmental disorders. For some of these diseases the olfactory deficits are broadly investigated (e.g. the Parkinson's and Alzheimer's disease, schizophrenia), for others there are very few studies (e.g. different affective disorders, autism and Asperger's syndrome). In this part of the chapter, we describe the olfactory impairment in three diseases: one non lesional: depression, one associated with cerebral lesions: Alzheimer's disease and one neurodevelopmental: autism.
5.1. Olfactory alterations in Alzheimer's disease
Alzheimer's disease (AD) is the most common form of neurodegenerative dementia. Anatomically, AD is characterised by a progressive neuronal loss with the accumulation of abnormal proteins in the brain: the amyloid plaques (composed of extracellular Ab peptide deposits) and the neurofibrillary tangles (composed of intracytoplasmic aggregates of a microtubule associated protein). Precise mechanisms are still widely unknown.
With the progression of the disease, lesions affect the whole brain. Nevertheless, the development of the lesions follows a particular way during time (H Braak et E Braak 1997). Firstly, temporal regions (hippocampus and peri-hippocampic cortices) are involved, with precocious consequences on declarative memory. Later, limbic and associative cortices (frontal and parietal lobes) are involved with consequences on emotional and cognitive abilities (aphasia, apraxia, agnosia) while sensory motor regions are relatively spare. Then, a global involvement is observed, with alterations of vision (visual hallucinations for example) or motor abilities (bedridden state).
Olfactory system is early invaded by AD pathological hallmarks, both at peripheral and central levels (Price et al. 1991; RI Mesholam et al. 1998). Despite a high prevalence of olfactory impairment in older people (Murphy et al., 2002) independently of any disease, olfactory processes seem to be severely impaired in AD. Consequences on patients are unrecognized but can affect abilities to avoid danger but also alimentation and social interactions.
' Olfactory function in Alzheimer's disease at peripheral level
At an early stage of the disease, the nasal sensory epithelium (Talamo et al. 1989; Jafek et al. 1992), the anterior olfactory nuclei (Averback 1983; Esiri et Wilcock 1984) and the olfactory bulbs (Esiri et Wilcock 1984; Ohm et H Braak 1987; Kov'cs, Cairns, et Lantos 1999; Kov'cs, Cairns, et Lantos 2001; Huesa et Ferrer 2010) are the site of AD hallmarks lesions. Moreover, olfactory tracts show a severe axonal loss, demonstrated on post-mortem studies (Davies, Brooks, et Lewis 1993). These modifications strongly suggest that the most peripheral level of the olfactory system is vulnerable to the disease. Thus, the olfactory deficits of olfactory sensitivity were studied in order to aid in diagnosis and understanding the disease.
Despite few investigations that failed to exhibit detection impairment in AD (Koss et al. 1988; Serby, P Larson, et Kalkstein 1991; Larsson et al. 1999), a majority of the studies show a significant difference of olfactory acuity in AD compared to controls, with an increased threshold for odour detection (Knupfer et Spiegel 1986; RL Doty, PF Reyes, et T Gregor 1987; Rezek 1987; C Murphy et al. 1990; R L Doty 1991; Morgan, S Nordin, et C Murphy 1995). As possible reasons for this inconsistency in findings are suggested: (1) the diversity of procedures and/or odorants used in measuring threshold in different studies and the lower reliability of some threshold test (Djordjevic et al. 2008); (2) the patient populations' variation in the degree of dementia; (3) and the fact that in some cases demented Alzheimer's patients
are unable to perform a threshold task (C Murphy et al. 1990). This procedure requires patients to carefully focus on the detection of a very weak and ephemeral stimulus, and the task is repetitive and boring. Nordin et al. (S Nordin et al. 1997) reported concerns about the performance of patients with dementia on odor detection threshold tasks, and significantly modified their procedures in light of these concerns. Nevertheless, thresholds could not be determined in some patients. In order to avoid this bias, it is preferable to use the olfactory tests more adapted for demented patients as for example the Alcohol Sniff Test (Davidson et C Murphy 1997) or the Sniff Magnitude Test (RA Frank, Dulay, et Gesteland 2003) although their principle and/or measured parameter are different aas compared the classic procedure of threshold evaluation (The Smell Threshold TestTM, Sensonics; Sniffin'Sticks Threshold Test, Burghart).
The studies exploring the relationship between olfactory thresholds impairment and cognitive performances are not always in agreement. Thus, Murphy et al. (C Murphy et al. 1990) found that the threshold scores were significantly correlated with the scores on the IMC (Information-Memory-Concentration test) and the MMS (Mini-Mental-State examination). Nevertheless, in a recent study, using a battery of 15 neuropsychological tests, Djordjevic et al. (Djordjevic et al. 2008) reported that detection threshold impairment is not correlated with the cognitive performances of patients and this was observed from the precocious stages of AD such as mild cognitive impairment. These last findings are in agreement with those observed by other researchers (Serby, P Larson, et Kalkstein 1991; Larsson et al. 1999).
For some authors, odor detection deficit is observed at the late stages of the disease (Serby, P Larson, et Kalkstein 1991; Larsson et al. 1999) whereas for others, it is a precocious marker
(S Nordin et C Murphy 1996; Bacon et al. 1998; Djordjevic et al. 2008) Peters et al., 2007. Nevertheless, detection threshold tends to be less impaired than odor identification in AD (RI Mesholam et al. 1998) even if it is related with the severity of dementia, i.e. the importance of pathological charge in patient's brain (C Murphy et al. 1990).
' Olfactory function in Alzheimer's disease at central level
Central olfactory structures are also involved by AD lesions. It concerns amygdalar, hippocampal, perihippocampal and orbitofrontal cortices (Hopper et Vogel 1976; RC Pearson et al. 1985; Harrison 1986; PF Reyes et al. 1987; Ferreyra-Moyano et Barragan 1989; R.C.A. Pearson et T.P.S. Powell 1989; Jellinger et Bancher 1998). The general retrograde degeneration of these structures in Alzheimer's patients provides the existence of several olfactory alterations.
In order to evaluate the central processes of olfaction in Alzheimer's patients, the three principal psychophysical tests are used: odor identification test, discrimination test and odor memory test. Evaluation of the odors' pleasantness, intensity and familiarity are occasional. Neurophysiological and neuroimaging studies are also carried out to investigate the central processes of olfaction in AD (Peters et al. 2003; C Murphy et al. 2005; C Murphy et al. 2009).
Among the three different olfactory functions, the odor identification is the more used. Several studies reported a severely odor identification impairment in AD (Waldton 1974; Peabody et Tinklenberg 1985; Serby et al. 1985; Warner et al. 1986; RL Doty, P Reyes, et T Gregor 1986; Knupfer et Spiegel 1986; Koss 1986; Koss et al. 1987; R.L. Doty 2003; Luzzi et al. 2007). Some authors evocate possible differences in the clinical profile of patients with severe odor identification impairments compared with less severe ones (Westervelt, Carvalho, et K. Duff 2007), but differences are spares and the overall cognitive profile is globally the same.
Odor identification is also altered precociously and probably before the first clinical symptoms of AD (prodromal stage) (Devanand et al. 2000; Devanand et al. 2008; Djordjevic et al. 2008). For this reason, odor identification impairment may represent a clinical marker of evolution toward AD (Serby, P Larson, et Kalkstein 1991; Morgan, S Nordin, et C Murphy 1995; Djordjevic et al. 2008). When the disease is clinically recognizable, odor identification is impaired from the earliest stages of AD (Kov'cs, Cairns, et Lantos 2001). It is related with the severity of dementia, as evaluated with global efficiency scores (the Mini Mental State Evaluation) (Waldton 1974; Knupfer et Spiegel 1986; C Murphy et al. 1990), and may represent a risk factor of a pejorative evolution (Graves et al. 1999). Indeed, on the one hand, it has been shown that people presenting high risk factors for developing AD dementia (for example ApoE gentotype) exhibit olfactory identification deficits compared with subjects who do not (Serby et al. 1996; Bacon et al. 1998; C Murphy et al. 1998; Handley et al. 2006; Djordjevic et al. 2008). On the other hand, people who present olfactory dysfunction exhibit a high risk to develop dementia during the next few years (Graves et al. 1999): risk of dementia is x1.92 if complete anosmia.
A recent post-mortem study investigated the relationship between odor identification performances and the accumulation of Alzheimer's disease pathology (Wilson et al. 2007). The authors reported that density of ? tangles was inversely related to odour identification score and that the association with odor identification was robust for tangles in the entorhinal cortex and CA1/subiculum area of the hippocampus, but not for tangles in other cortical sites. They concluded that difficulty in identifying familiar odors in old age is partly due to the accumulation of neurofibrillar pathology in central olfactory regions (Wilson et al. 2007).
The odor discrimination performances are rarely investigated in Alzheimer's patients. When this task is carried out, it is evaluates simultaneously with the identification task, because the two tests investigate different impairments of the olfactory function. Thus, identification requires high level cognitive functions such as long term memory (declarative or semantic component), attentional processes, naming and is thus dependant on cognitive performances (Larsson et al. 1999; Larsson, Finkel, et Pedersen 2000; Larsson, Oberg, et B'ckman 2006; Djordjevic et al. 2008). At the opposite, discrimination requires more basic cognitive functions and seems to be less related to cognitive performance, and to the severity of dementia. Djordjevic and colleagues (Djordjevic et al. 2008) showed that patients with pre-dementia states (mild cognitive impairment) have intermediary scores at a discrimination and identification tasks between normal elderly controls and Alzheimer's patients, suggesting a linear progression the olfactory impairment while disease evolution. However, deficits in olfactory detection tasks influence discrimination but not odour identification.
A number of studies have shown significant deficits in odor-recognition memory (semantic and episodic memory) in AD (PJ Moberg et al. 1987; S Nordin et C Murphy 1998; Niccoli-Waller et al. 1999; PE Gilbert et C Murphy 2004; Razani et al. 2010). The hypothesis that odor memory may be impaired even in preclinical cases of AD was also been tested by carried out a studies with the individuals showing impairment in cognitive functions similar to patients with probable AD. The results suggested a larger impairment in recognition memory for odors than for faces and symbols in these preclinical cases of AD (S Nordin et C Murphy 1996). It has also been showed that remote recognition memory for odors, but not faces, as measured by familiarity rating is impaired in probable AD (Niccoli-Waller et al. 1999). A number of studies have reported significant deficits in odor-recognition memory and/or odor identification (based on semantic memory) and odor detection in elderly individuals with genetically risk for developing AD (individuals positive for the apolipoprotein E e4 allele) compared to those not at risk (SS Schiffman et al. 2002; PE Gilbert et C Murphy 2004).
In order to study the central processes of olfaction in AD, neurophysiological and neuroimaging techniques are also used.
Several studies confirm the olfactory deficits in AD observed with subjective psychophysical tests by studying the olfactory event-related potentians (ERPs) (Morgan et C Murphy 2002; Peters et al. 2003). Thus, Morgan and Murphy observed that the odor evoked response of the brain is significantly reduced in amplitude and delayed in its latency in normally aging persons and dramatically more delayed in Alzheimer's patients (Morgan et C Murphy 2002). More recent study investigated olfactory and cognitive processing deficits in individuals with genetically risk for AD (e4+) by using a cross-modal recognition memory task and the event-related potential (ERP) (C Murphy et al. 2009). The findings showed that nondemented e4+ individuals will expend greater effort in cognitive processing or engage in alternative strategies and therefore require greater activation of neural tissue or recruitment of different neural populations. The authors suggest that cross-modal ERP studies of recognition odor memory discriminate early neurocognitive changes in individuals with genetically risk and individuals not at risk and may contribute to identifying the phenotype of persons who will develop Alzheimer's disease (C Murphy et al. 2009).
The results of neuroimaging studies on AD patients were used in order to understand and explain the olfactory impairment observed with psychophysical tests (Buchsbaum et al. 1991; C. Murphy, Jernigan, et Fennema-Notestine 2003; C Murphy et al. 2005; Devanand et al. 2008; Chou et Bohnen 2009). Moreover an fMRI technique was used in order to study the correlations in fMRI activation among specific regions of interest (ROIs) for olfactory tasks in three groups of subjects: young adults, older adults and patients with AD (C Murphy et al. 2005). The authors observed that in older adults there is a breakdown of connectivity between orbito-frontal cortex and mesial temporal lobe and this was especially true in AD. Patients showed lower overall activation, particularly in mesial temporal lobe.
As a summary, olfaction is precociously impaired in AD at every stage of olfactory processing. Olfactory impairment is associated with an increased risk of evolution toward AD in cognitively unimpaired subjects, and is a marker of severity of bad evolution in AD patients. The cognitive aspects as recognition and memory are less studied but appear to be significantly impaired. Consequences on patients' quality of life and environmental adaptation are important to consider in clinical practice.
