Olfactory sensitivity through the course of psychosis: Relationships to olfactory identification, symptomatology and the schizophrenia odour

Abstract

There is some evidence for an unusual body odour in schizophrenia that has been linked to a hexenoic acid derivative (trans-3-methyl-2-hexenoic acid; MHA). Poor body odour has been linked to increased negative symptoms and reduced olfactory identification ability. However, the relationship between these findings and MHA, including olfactory sensitivity for MHA, has not been examined. Olfactory sensitivity thresholds were assessed for MHA and n-butyl-alcohol (NBA), in normal controls (CTL; n = 24), patients with chronic schizophrenia (CHR; n = 32) and a first-episode psychosis cohort (FE; n = 31). In addition, forced choice detection of the pheromonal steroids 5-α-androst-16-en-3-one, androsterone-sulphate and estrone-3-sulphate was performed along with a measure of olfactory identification. CHR patients had significantly reduced sensitivity to MHA, but not NBA, compared to FE and CTL subjects. While sensitivity to pheromones was not different between the groups, CHR patients who could not detect them also showed poorer sensitivity to MHA. Further, the CHR group showed a significant association between reduced MHA sensitivity and greater levels of disorganised and negative symptoms. No relationships between identification and sensitivity for any substance were found. Our findings are the first to report reduced sensitivity for MHA in chronic schizophrenia patients, in the absence of similar impairment for more traditionally used substances. This may be linked to olfactory habituation effects, abnormal chemical processing or a genetic predisposition.

Introduction

Clinical evidence suggests that patients with schizophrenia exhibit poor self-care and bad body odour (Brewer et al., 1996), and it has been suggested that a component of this odour may be particular to the disorder and represent a specific metabolic abnormality (Smith and Sines, 1960, Liddell, 1976). This component was isolated from the sweat of patients by gas chromatography, mass spectrometry, and nuclear magnetic resonance spectroscopy, and identified as trans-3-methyl-2-hexenoic acid (MHA: Smith et al., 1969). In that study, analysis of the acidic components of patient's sweat relative to control subjects revealed MHA as peculiar and specific to the schizophrenia specimens. Although Perry et al. (1970) failed to verify these findings, their study was flawed in that more effective methods of preparing and processing MHA for analysis were available (Smith and Leong, 1972). Gordon et al. (1973) also failed to demonstrate increased amounts of MHA in the sweat of schizophrenia patients, but they only studied seven patients, three of their twelve controls had very highly elevated MHA levels, and reanalysis of their data using more appropriate non-parametric statistics (Kolmogorov–Smirnoff Z) reveals a significant elevation of total MHA in the schizophrenia group. Furthermore, Roberts (1980) suggested that the evidence by Perry et al. (1970) against MHA as a candidate for the culprit metabolite was faulty, and he then demonstrated various pathways for synthesis of the same. Since then, these findings have been largely ignored. Rather, it has been argued that the abnormal odour secreted from patients' skin is steroidal in nature (perhaps because the odour of MHA is similar to the steroidal odour of stale urine). Two such compounds, 5-α-androstenone and 5-α-androst-16-en-3-one (androstenol), also have a urinous odour (Kloek, 1961, Bird and Gower, 1982). Habituation effects could explain reduced olfactory sensitivity to these substances — specifically, the threshold at which these substances can be detected (Wiener, 1966, Wiener, 1967, Bradley, 1984). Bradley (1984) demonstrated that males with schizophrenia had an increased olfactory sensitivity for these steroids compared with controls, consistent with the finding that male patients have significantly reduced secretion of androstenol (Brooksbank and Pryse-Phillips, 1964), which would tend to make them more sensitive to the odour. However, Bradley's study failed to take into account the prevalence of androstenone agnosia in the general population of around 30% (Wysocki and Gilbert, 1989). Further, it is unclear to what extent there is a general impairment of olfactory sensitivity in schizophrenia and how it is related to the symptoms of the disorder. A general reduction in sensitivity would not be a result of habituation, but rather indicate an impairment of olfactory processing, which could be central or peripheral.

In order to explore this further, Geddes et al. (1991) studied olfactory sensitivity for a musk ketone in schizophrenia with the aim of investigating associations with positive and negative clinical subtypes. Their findings demonstrated significantly reduced olfactory sensitivity in the negative symptom group compared to the positive symptom group, although neither differed significantly from controls. Complementing this study, Brewer et al. (1996) found that increased levels of negative symptoms were also related to olfactory identification deficits, and that both were associated with an unawareness of the patient's body odour. In this and other similar studies of olfactory identification suprathreshold stimuli are used, so it was not possible to examine the relationship between olfactory sensitivity versus identification. However, previous reports have suggested that identification deficits occur in the presence of intact sensitivity, particularly in established illness (Kopala et al., 1989, Kopala et al., 1992, Kopala et al., 1998). Clarifying the nature of olfactory sensitivity in schizophrenia requires administration of a broad range of odourants to determine general and specific impairments. These can then be subsequently related to identification deficits, symptom profiles and phase of illness. Demonstrating the importance of this, Sirota et al. (1999) found increased sensitivity to isoamyl acetate in a first-episode psychosis group relative to healthy control subjects. Treatment with neuroleptics was associated with a threefold reduction in sensitivity, leading Sirota et al. (1999) to suggest that antipsychotic medication has a negative impact on olfactory function.

The study of olfactory sensitivity is confounded by a number of other factors. First, assessment of sensitivity is less reliable than assessment of identification (Doty et al., 1995, Martzke et al., 1997, Moberg et al., 1999). It is also unclear whether body odour confounds the ability to detect some odours used in olfactory sensitivity research due to habituation. Further, there are sex differences in olfactory sensitivity, where women have a superior detection ability (Koelega and Koster, 1974, Stoddart, 1976, Wysocki and Gilbert, 1989, Dalton et al., 2002). Moreover, there are asymmetric effects of menstrual cycle, such that the right nostril is more sensitive during menstruation and the left nostril more sensitive during ovulation (Purdon et al., 2001).

Another confound for the study of olfactory sensitivity is variation in performance across individuals (Clark and Ball, 1987, Albone and Shirley, 1984), which may be partly genetically based (Hubert et al., 1980, Hubert et al., 1981, Wysocki and Beauchamp, 1984). Finally, the use of odours that do not trigger a 5th cranial (trigeminal) nerve response is required; otherwise, a normal trigeminal response may be erroneously interpreted as intact 1st cranial (olfactory) nerve function (Martzke et al., 1997).

In this study, we examined olfactory sensitivity for steroidal (pheromonal) substances and for MHA in patients with schizophrenia compared to first-episode psychosis and normal controls, and related these measures to olfactory identification ability and symptom profiles. We also addressed laterality effects, the presence of sex differences and menstruation in sensitivity, and potential for a trigeminal response. In particular, we hypothesised that patients with chronic schizophrenia would have reduced sensitivity for MHA compared to patients with first-episode psychosis and controls.