The sense of smell in systemic lupus erythematosus

Abstract

Objective

To assess the olfactory functions in systemic lupus erythematosus (SLE) patients compared with age- and sex-matched healthy controls, and to examine the association between the sense of smell and disease activity and central nervous system (CNS) involvement.

Methods

Olfactory functions in 50 SLE patients and 50 age- and sex-matched controls were evaluated using the Sniffin' Sticks test, the 3 stages of which are threshold, discrimination, and identification (TDI) of different odors. TDI scores were analyzed according to SLE disease activity and CNS involvement.

Results

In both the SLE and control groups, smell deficit correlated with male sex and older age. A decrease in the sense of smell was observed in SLE patients (46%) and controls (25%) (P ≤ 0.02), while loss of smell (anosmia) was documented only in SLE patients (10%). Total TDI scores and individual stages of smell correlated with SLE Disease Activity Index (P < 0.001) and CNS manifestations (P < 0.03).

Conclusion

Our findings suggest that there is a decrease in the sense of smell in SLE patients compared with healthy subjects and that the decrease in the sense of smell among SLE patients correlates with disease activity and CNS involvement.

The sense of smell is vital and has an important role in environment recognition. Olfaction is a complex process of the central nervous system (CNS) involving specific areas of the brain (e.g., the limbic system). Olfactory dysfunction has been confirmed in various CNS diseases (1), as well as in multiple sclerosis, a neurologic autoimmune disease (2).

Systemic lupus erythematosus (SLE) is a systemic autoimmune disease, and CNS involvement in the disease (neuropsychiatric SLE [NPSLE]) has been observed in 12–95% of SLE patients (3). Numerous autoantibodies, including anti–ribosomal P (anti-P) antibodies, have been linked to NPSLE manifestations (4). Recently, our group studied the effects of anti-P antibodies in a mouse model of NPSLE, via direct injection of antibodies into mice brains (5). The treated mice developed depressive-like behavior and a decreased ability to smell menthol (5, 6). Histologic examination of the mouse brains revealed that anti-P antibodies had bound and penetrated neuronal cells of the limbic system, including cells in the amygdala and hippocampus (5, 7). It was also previously demonstrated that anti-P antibodies are capable of blocking protein synthesis and inducing apoptosis (7).

The above-described results indicated that NPSLE-like manifestations and olfactory impairment might be induced in mice by autoimmune mechanisms, which target specific areas of the brain. Since the olfactory functions of SLE patients had not been previously assessed, we designed the current study to evaluate the sense of smell in SLE patients and its correlation with disease activity and CNS manifestations.

PATIENTS AND METHODS

Patients.

Fifty SLE outpatients and 50 age- and sex-matched controls were examined. All patients fulfilled the American College of Rheumatology criteria for SLE (8). On the day of examination, subjects were evaluated for olfactory abilities and SLE Disease Activity Index (SLEDAI) score (9), and completed a questionnaire concerning their smoking habits, past surgeries (including plastic surgeries), allergies, asthma, and current therapies. The patients' clinical files were further analyzed for documented CNS manifestations. All study procedures were approved by the local ethics committee.

Method of olfactory assessment.

The levels of olfactory functions were evaluated using the 3 stages of the Sniffin' Sticks kit (Burghart Medizintechnik, Wedel, Germany), which is based on a pen-like odor-dispensing device (10). In stage 1, threshold was assessed using n-butanol as a single odorant. Three sticks were presented to each subject in a randomized order, 2 contained solvent and the third contained the odorant at a certain dilution. The subject was then asked to identify the stick with the odorant. In stage 2, discrimination between 2 different odorants was assessed. The subject was presented with 3 pens; 2 contained the same odorant and 1 contained a different one. The subject was then asked to differentiate between pens. In stage 3, the ability of the subject to identify an odorant out of 4 options was assessed.

The maximum score in each stage was 16 points, with a maximum possible total score of 48 points for the stages of threshold, discrimination, and identification (TDI) combined. Patients with a TDI score of >30 are considered to have normal olfaction (normosmia), patients with a score of 15–30 are considered to have decreased olfaction (hyposmia), and patients with a score of <15 are considered to have a loss of olfaction (anosmia).

Statistical analysis.

All the statistical calculations were performed using SAS statistical software (SAS Institute, Cary, NC). Normally distributed variables were summarized using the mean ± SD, and non–normally distributed variables by the median and range. Univariate comparisons between nominal variables were performed by chi-square test. For assessment of the correlation between 2 continuous variables, we used Pearson's and Spearman's correlation coefficient for normal and non-normal variables, respectively. P values less than or equal to 0.05 were considered significant.

RESULTS

We examined 2 groups of subjects: 50 SLE patients and 50 matched controls (Table 1). No histories of head injuries, nasofacial operations, or active nasal-sinus or allergic diseases were reported in either group.

Table 1. Characteristics and TDI scores of the SLE patients and controls*

	Patients (n = 50)	Controls (n = 50)	P
* Except where indicated otherwise, values are the mean ± SD. TDI = threshold, discrimination, and identification; SLE = systemic lupus erythematosus; NS = not significant.
Sex, no. (%) female	41 (82)	39 (78)	NS
Age, years	41.3 ± 15	39.6 ± 12	NS
Smoking, no. (%)	14 (28)	10 (20)	NS
Total TDI score	28.4 ± 8.7	33.2 ± 3.2	<0.009
Threshold score	5.3 ± 2.4	6.4 ± 1.4	0.04
Discrimination score	10.8 ± 3.6	13.3 ± 1.65	<0.0001
Identification score	12.4 ± 3.5	13.4 ± 1.3	NS

In the SLE patient and control groups combined, 80% of subjects were women. Ages ranged from 19 years to 74 years (Table 1). In both groups, the TDI score was negatively correlated with older age (P ≤ 0.003) and male sex (P = 0.05).

Smell deficit was also associated with SLE; patients had significantly lower TDI scores compared with healthy controls (P < 0.009) (Table 1). Furthermore, anosmia was present exclusively in the SLE group (10% of patients versus 0% of controls) (P = 0.05) (Figure 1).

Figure 1. Threshold, discrimination, and identification (TDI) scores in systemic lupus erythematosus (SLE) patients versus controls. The TDI score was significantly decreased in patients with SLE compared with healthy matched controls (P < 0.009). Anosmia (complete loss of smell) was prevalent in SLE patients exclusively. Hyposmia (decreased sense of smell) was more prevalent in SLE patients than in controls. Normal sense of smell was more prevalent in controls than in SLE patients.

A significantly lower score was observed among SLE patients compared with controls for both the threshold (P = 0.04) and discrimination (P < 0.0001) stages of the test. Smell identification (stage 3) was also decreased in SLE patients compared with controls, but this difference did not reach statistical significance (Table 1).

Our results also suggest that the loss of olfaction was associated with SLE activity. The mean duration of the disease was 8.6 ± 7.4 years (range 1–30 years), and the mean SLEDAI score was 5.9 ± 4.7 (range 0–21). When comparing SLE patients whose SLEDAI scores were high (score of >10) with SLE patients whose scores were very low (score of 0), we found that the total TDI score was decreased in the former group (P < 0.001) (Figure 2A). Moreover, SLEDAI score was significantly associated with a decrease in each of the 3 stages of smell (P = 0.026 for threshold; and P < 0.001 for discrimination and identification).

Figure 2. TDI score and SLE severity. A, TDI scores were significantly decreased in SLE patients with active disease (SLE Disease Activity Index [SLEDAI] score of >10) compared with patients with inactive disease (SLEDAI score of 0) (P < 0.001). B, TDI scores were significantly decreased in patients with central nervous system (CNS) involvement compared with SLE patients without CNS involvement (P < 0.03). Values are the mean and SD. See Figure 1 for other definitions.

Prior to the current study, 24% of patients in the SLE group had experienced NPSLE manifestations (e.g., cerebrovascular disease, transient ischemic attack, epilepsy, transverse myelitis, and optic neuritis). These patients had a significantly lower mean TDI score compared with SLE patients without CNS involvement (P < 0.03) (Figure 2B). CNS involvement was associated with a decrease in both smell discrimination (P = 0.03) and smell identification (P < 0.01).

DISCUSSION

The sense of smell is an ancient sensory modality, used by primitive organisms, animals, and humans for sampling the chemical composition of their surrounding environments and for eliciting information on mates, sensual pleasures, and other choices to be made. Within the last 2 decades, the genetic and anatomic organization of the olfactory system and the odorant receptors has been elucidated (11); 1,000 murine genes and 350 human genes, which encode for odor receptors, expressed in the olfactory epithelium, were identified by Nobel Prize winners Linda Buck and Richard Axel (11). Axons expressing the same odorant converge in the olfactory bulb, of which newer-odor neurons emerge to the olfactory cortex (12). The incidence of olfactory dysfunction in the general population is a matter of debate; ∼1% of the general population have anosmia (loss of smell), and ∼5–8% have hyposmia (reduced olfactory function) (13). In concordance with the findings of previous studies, our findings suggest that a decrease in olfactory functions is associated with male sex and with older age, in both patient and control groups.

As described by the Nobel Prize winner Peter Doherty, these findings might encourage speculation about the correlation between weakness of the immune system (e.g., the immune systems of men and of people of older age) and decreased olfactory function (14). Recently, links between the olfactory system, the immune system, and various diseases have been identified. Decreased olfaction was observed in patients with several CNS diseases in which immune-mediated mechanisms have been implicated (e.g., Parkinson's disease, schizophrenia, Alzheimer's disease, and multiple sclerosis) (2, 15).

To the best of our knowledge, the sense of smell in patients with systemic autoimmune disease has not been evaluated previously; thus, we assessed the olfactory abilities of patients with SLE. The results of our study revealed significant olfactory deficits in SLE patients compared with age- and sex-matched healthy controls. Furthermore, upon examining the interaction between disease activity and olfaction, we found that patients with more active disease (SLEDAI score of >10) and patients with past NPSLE manifestations both had significantly lower TDI scores. In an NPSLE mouse model, our group has previously shown that injection of anti-P antibodies into mouse brains induced depression and olfactory deficit, which suggested humoral-mediated mechanisms in the pathogenesis of these phenomena (5, 7). Moreover, these antibodies are capable of binding and penetrating neuronal cells of the limbic areas that are associated with olfaction (16). Involvement of the limbic system (e.g., atrophy of the hippocampus and amygdala) in SLE patients was recently documented utilizing magnetic resonance imaging (16). These findings suggest that olfactory deficit is associated with SLE disease activity and CNS manifestations in humans and in mice. Although the exact mechanism of olfactory impairment has yet to be elucidated, the possibility of an immune-mediated mechanism is intriguing.

The assessment of the olfactory system utilized for the current study had 3 stages for evaluating a subject's sense of smell: threshold, discrimination, and identification of odors. Upon analysis of the individual stages, we found differences between them. A decrease in olfactory function was observed in SLE patients during both the threshold and discrimination stages.

The importance of a smell threshold can be easily understood, since the inability to smell odorants in normal dilutions (e.g., food that has spoiled or leaks in a gas line) could seriously compromise a person's health. As observed in the current study and by others (17), the smell function is the first to be impaired with age. The threshold stage of the test is the most complex to perform. The second stage, odor discrimination, is easy to test and was found to be the most sensitive. In the final stage, identification, loss of olfaction was observed only in SLE patients with CNS involvement or in those patients with very active disease. This might be explained by the different role or complexity of each stage in the assessment, as well as by study design. The identification stage, which has been excluded in some studies, requires comprehension and verbal abilities, making it the most sophisticated of the 3 stages. The correlation between smell identification and CNS disease (e.g., Parkinson's disease) is supported by the results of previous studies, in which it was found to enable the best discrimination between patients and controls (18).

Therefore, for full assessment of olfaction, all 3 stages of the TDI test are important. Each stage is directed at a different ability and impairment, and may help identify specific problems among patients and diseases.

In conclusion, after comparing SLE patients and their matched controls, we have found a significant decrease in the olfactory abilities of SLE patients, which correlated with disease activity and CNS manifestations. Olfaction is hardly a familiar subject for rheumatologists and immunologists. However, based on the literature on smell deficits, data from our mouse model, and results of the current study, a possible role of an autoimmune mechanism in the pathogenesis of olfactory impairment may exist. Moreover, smell decrement has been found to be an early and predictive sign in several CNS diseases, and therefore, might be a useful and easy tool for the physician in early diagnosis of CNS involvement in autoimmune diseases.

AUTHOR CONTRIBUTIONS

Dr. Y. Shoenfeld had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. N. Shoenfeld, Agmon-Levin, Paran, Katz, Kivity, Langevitz, Zandman-Goddard, Y. Shoenfeld.

Acquisition of data. N. Shoenfeld, Agmon-Levin, Flitman-Katzevman, Paran, Langevitz, Zandman-Goddard, Y. Shoenfeld.

Analysis and interpretation of data. N. Shoenfeld, Agmon-Levin, Flitman-Katzevman, Paran, Katz, Kivity, Langevitz, Zandman-Goddard, Y. Shoenfeld.

Manuscript preparation. N. Shoenfeld, Agmon-Levin, Katz, Langevitz, Zandman-Goddard, Y. Shoenfeld.

Statistical analysis. N. Shoenfeld, Agmon-Levin, Katz.