Caccappolo, Elise PhD; Kipen, Howard MD, MPH; Kelly-McNeil, Kathie BA; Knasko, Susan PhD; Hamer, Robert M. PhD; Natelson, Benjamin MD; Fiedler, Nancy PhD
Patients with multiple chemical sensitivities (MCS) report a varied array of symptoms involving the central nervous system as well as other organ systems in response to low-level chemical exposures. The symptoms of MCS include muscle/joint pain, fatigue, cognitive difficulties, and gastrointestinal problems. A number of MCS patients have reported multiple adverse food reactions and acquired intolerance for various drugs and alcohol. 1 In addition, a heightened sense of smell, or lower odor threshold, is often reported. 2–4 This hypersensitivity to odors has been investigated in previous research for clues to the etiology of MCS symptoms. 3,4
In their investigation of olfactory sensitivity associated with MCS, Doty et al 3 examined odor detection thresholds for two substances, phenyl ethyl alcohol (PEA) and methyl ethyl ketone, in MCS subjects and healthy controls. In contrast to clinical reports, results showed that MCS subjects did not have lower olfactory thresholds than healthy controls. However, the MCS group exhibited higher nasal resistance and respiration rate both before and after testing. These results are limited by the fact that subjects were selected without specific criteria for a definition of MCS. Subjects were diagnosed with MCS after their completion of a questionnaire and case reviews to determine adverse responses to environmental chemicals in the absence of other medical conditions that could explain these symptoms. The present study attempts to replicate and extend Doty et al’s 3 findings by comparing MCS subjects selected on the basis of Cullen’s 5 more explicit case definition with medical control groups who, respectively, share different features of MCS (ie, asthmatic patients and chronic fatigue syndrome [CFS] patients), and with matched healthy controls. Olfactory thresholds for PEA and pyridine (PYR), a test of odor identification (University of Pennsylvania Smell Identification Test [UPSIT]), 6 and ratings of trigeminal symptoms and esthetic qualities in response to suprathreshold levels of PEA and PYR were administered to assess olfactory responses.
Rationale for Comparison Groups
Despite self-reports of MCS patients that suggest heightened sensitivity and attribution of illness to chemical exposure, no objective tests have been developed to diagnose MCS or to differentiate MCS from other illnesses with similar features. For example, some asthmatic patients report respiratory symptoms in response to odors. 7–11 They resemble MCS patients in their experience of environmentally induced symptoms and in their tendency to avoid certain environmental stimuli. 12 The physiologic basis for this complaint may be in the bronchi, the central nervous system, or both. 7 However, the symptoms reported by asthmatic patients are limited to the respiratory system and do not generally result in the severe restrictions in lifestyle reported by MCS patients. Furthermore, no reports of a heightened sense of smell in asthmatic patients have been documented. Because of these similarities and differences, MCS subjects were compared with asthmatic control subjects for the effect of known bronchial hyperreactivity on odor identifications and thresholds.
CFS patients represent another optimal comparison group because they share a number of similar features with MCS patients. Demographically, both disorders occur predominantly in well-educated women 13,14 and lack a demonstrated organic explanation for their symptoms. MCS and CFS patients also share many symptoms, including the complaint of impaired cognitive function, psychiatric symptoms such as anxiety and depression, and multiple physical symptoms. Finally, although sensitivity to chemicals is the hallmark of MCS, a subset of CFS patients also report symptoms in response to chemical exposures. 2,15,16 However, unlike MCS patients, CFS patients are not known to report a heightened sense of smell despite the overlap of symptom presentation.
Assessing Odor Perception, Sensation, and Symptoms
According to symptom presentation and anecdotal reports, MCS patients describe having a heightened sense of smell, 4 or lower odor detection threshold, as well as sensitivity to chemicals in products that have traditionally been considered safe and/or pleasant, such as household cleaners and perfumes. 17 To accurately assess the overall olfactory functioning and perception of MCS patients, the present study attempted to validate their symptoms in response to two substances: PYR, an unpleasant odorant known as a trigeminal irritant that elicits sensations by means of the common chemical sense served by cranial nerve V, 18 and PEA, a non-irritating chemical that is the major component in rose oil, a common constituent of fragrances.
Odor detection threshold is measured by defining the minimum concentration necessary for perception to occur. The UPSIT, developed by Doty et al, 19 provides a measure of odor identification ability. Sensory irritation in response to different supratheshold levels of odorants can be assessed through subjective ratings of particular sensations such as stinging/pricking, burning, numbness, and temperature. Esthetic perception can be obtained through ratings of intensity, safety, and pleasantness/hedonics. 18,20 The present study examines: (1) the ability of MCS patients to detect and identify odors, and (2) the trigeminal symptom reports and esthetic ratings in response to suprathreshold levels of both irritating and non-irritating substances. If MCS subjects do in fact have a heightened sense of smell, they should exhibit lower thresholds and make fewer errors in identification than asthmatic, CFS, and healthy subjects.
Neither asthmatic nor CFS patients have reported a heightened sense of smell and were therefore not expected to demonstrate a lower threshold or to make fewer errors in odor identification than normal controls. However, because a subset of CFS patients report symptoms in response to chemicals and some asthmatic subjects report respiratory symptoms in response to chemicals, both groups were expected to respond with more trigeminal symptoms and negative ratings of suprathreshold levels of PYR, a known irritant, than normal controls.
MCS subjects were recruited from patients referred to the Environmental and Occupational Health Sciences Institute at Robert Wood Johnson Medical School on the basis of the criteria derived from Cullen. 5 The seven main diagnostic components were: (1) initial symptoms occur following an identifiable environmental exposure(s) such as pesticide poisoning, respiratory tract irritation, or solvent intoxication; (2) symptoms involve more than one organ system and almost always include the central nervous system; (3) symptoms recur and abate in response to predictable stimuli, particularly a perceived environmental exposure; (4) symptoms occur in response to low-level exposures to multiple agents of varying structural classes; (5) the doses of these agents are at least two orders of magnitude lower than the established thresholds for acute health effects; (6) tests of physiologic function are unable to explain the symptoms; and (7) the pattern of symptoms cannot be explained by any other organic disorder. Also, the subjects were not under the care of a clinical ecologist at the time of their participation and had no current or previous medical or psychiatric diagnosis that would explain their symptoms. Twenty-seven subjects met all of the above criteria, and six met all criteria except for an identifiable exposure.
CFS patients were referred to the study through the CFS Center at the New Jersey Medical School. They were required to have two major criteria (ie, new onset with greater than 50% reduction in activity and no other preexisting medical condition) along with at least 6 of 11 minor criteria (ie, mild fever, sore throat, anterior or posterior cervical/axillary nodes, general muscle weakness, myalgias, prolonged fatigue after exercise, generalized headaches, migratory arthralgias without joint swelling or redness, neuropsychologic complaints, sleep disturbances, and description of the main symptom complex as initially developing over a few hours to a few days) and two of three physical criteria (ie, low-grade fever, non-exudative pharyngitis, anterior or posterior cervical or axillary nodes). Patients also met the criteria if they had two major criteria and a total of eight or more criteria from the minor and physical items. 21,22
Asthmatic subjects were recruited by letters sent to local pulmonologists who had previously participated in University studies of asthma, as well as by advertisements in local newspapers. These subjects were required to have received a diagnosis of asthma by a pulmonary or allergy physician and to be taking daily asthma medication, at a minimum, inhaled corticosteroids, beta agonists, or cromolyn sodium. Five subjects had taken oral steroids within 6 months before testing. Healthy controls were recruited from the community in response to an advertisement posted in area newspapers similar to that used for recruiting asthmatic subjects.
All potential subjects gave informed consent in compliance with standard Institutional Review Board procedures. Each subject completed a confidential medical history questionnaire and received a physical examination, a spirometric test, and a routine blood chemistry test including measurement of immunoglobulin and complement levels.
On the basis of the medical questionnaire and physical examination, no subject from any group had a diagnosis of serious concomitant medical disorders, including neurologic disease or brain injury, stroke or cardiovascular disease, serious pulmonary disease, hypertension, liver or kidney disease, serious gastrointestinal disorders (eg, colitis), diabetes, Lyme disease, significant toxic exposure, or major psychiatric conditions (ie, psychoses, organic mental disorder, manic depression, substance use disorders). Subjects were not eligible if they were currently smoking, had sinus problems, or had a history of nasal polyps or allergies.
All subjects were female, between the ages of 23 and 60, and non-smokers. MCS subjects were significantly different in age from CFS subjects but not from asthmatic or control subjects. MCS and CFS subjects had more years of education (Table 1). Although statistically significant, these differences were not regarded as meaningful to the study.
A total of 83 subjects were excluded from participation in the study for the following reasons: 36 had medical disorders or psychiatric conditions outlined above; 24 declined participation (12 had scheduling problems, 8 moved or lived too far away, 4 gave no reason); 10 did not meet age requirements; 8 were smokers; and 5 did not use English as their first language. Thus, 33 MCS patients, 13 CFS patients, 16 asthmatic subjects, and 27 healthy controls participated in the study.
Odor threshold tests.
PEA and PYR were used to determine odor detection thresholds. Following standard procedures used at the Monell Chemical Senses Center (Philadelphia, PA), both substances were presented as a set of dilutions in plastic squeeze bottles fitted with flip-up caps with a 3.2-mm orifice. The plastic smell was removed from the bottles and closed caps by boiling for 3 hours in deionized/distilled water and then air-dried.
The PEA (Eastman Kodak Co, Rochester, NY) dilution with glycerol (no odor) (Fisher, Springfield, NJ) is a 20-step series starting from step 0 (undiluted) to step 19. This is a half-log dilution series (ie, each step is a one-half log of the preceding one), and the amount of liquid in each bottle is always 18 mL when the series is complete.
The PYR (Fisher, Springfield, NJ) series is prepared according to the following protocol, taken from the Monell Chemical Senses Center. The dilution with formulary grade mineral oil (Spectrum Chemical, Gardena, CA) is based on several serial dilutions from step 0 (undiluted) to step 29. The first serial dilution (steps 0 to 7) begins with a stock solution of 100% pure PYR, which is serially diluted with mineral oil to create standards containing concentrations of PYR between 50 and 0.8%. For steps 8 to 15, a new stock solution of 0.4% PYR is created to minimize errors associated with repeated serial dilution of a standard. This series is diluted 7 times to create a new stock solution ranging between 0.4% and 0.003% PYR (step 15). To continue, a new stock solution of 0.0015% PYR is created and serially diluted 7 times (step 16 to 23) to continue the series. This series consists of PYR ranging from 0.0015% to 1.17 × 10-5 (0.01 parts per thousand). For the final steps (24 to 29), a new stock solution of 0.000005859% PYR is created and serially diluted 5 times to step 29, which is equal to 1.83 × 10-7% PYR (0.1 ppm).
To assess the trigeminal (ie, burning, stinging/pricking, numbness, temperature) and esthetic (ie, intensity, safety, pleasantness/hedonic) properties of PEA and PYR, suprathreshold testing was conducted. Dilutions of 100%, 1%, 0.01%, and glycerol (blank) were used for PEA; dilutions of 0.4%, 0.025%, 0.0015%, and mineral oil (blank) were used for PYR. Subjects were instructed to squeeze the bottle to sniff the contents and then rate the following properties of the odor: burning (can occur with or without the sensation of temperature, not synonymous with hot), stinging/pricking (sharp irritation that could be either constant or intermittent), numbness (not the absence of sensation but the feeling experienced during the onset and offset of local anesthesia), and temperature (cool to warm) on a Likert scale from low (1) to high (9). These four properties constitute what are traditionally referred to as trigeminal qualities. 15,20 Subjects were also asked to rate the odors in terms of strength from very weak (1) to very strong (9), safety from very safe (1) to very unsafe (9), and pleasantness from very pleasant (1) to very unpleasant (9). These three qualities are considered esthetic. 15,20
To reduce type I error due to multiple comparisons of suprathreshold rating scales, trigeminal (burning, stinging/pricking, numbness, and warmth ratings) and esthetic (intensity, safety, and pleasantness ratings) scales were formed a priori by summing ratings for each variable at each suprathreshold concentration. The adequacy of these constructs, however, was also assessed by means of factor analyses. These were conducted separately for both PEA and PYR to assess the possibility that the relationships were different under the two exposure conditions and to validate the factor analysis with different data. A maximum likelihood factor analysis was run on correlations among the seven variables after partialing out effects due to group or concentration, using squared multiple correlations as communality estimates, and using a varimax rotation. The rotated factor pattern matrix “loadings” for PEA and PYR can be found in Table 2. Using 0.30 as a minimum value, factor 1 for PEA and PYR consists of three variables: burning, stinging, and numbness; factor 2 for PEA and PYR consists of three variables: strength, safety, and intensity. The variability of warmth did not load on either variable. These results provide a reasonable confirmation of the a priori construction of the trigeminal and esthetic constructs for both PEA and PYR.
The UPSIT 6 is a standardized test of odor identification designed to be self-administered once the proper instructions have been provided by an examiner. It is comprised of four booklets of 10 odors that are each microencapsulated in a small strip of sandpaper. The subject is provided with a pencil and instructed to make a few firm scrapes on the sandpaper strip to release the odor, bring the booklet close to the nose, and sniff. Four different odor alternatives are listed on the same page, and the subject must choose the one that most corresponds to his or her smell experience.
Subjects were assessed with a psychiatric interview that incorporated the Diagnostic Interview Survey from the Diagnostic and Statistical Manual of the American Psychiatric Association, 3rd Revision 23 and with tests of memory, mood, and personality. These results are reported separately.
Odor threshold ratings.
Subjects then completed the assessment of their ability to identify odors. Participation was rescheduled if a subject had had a cold or active allergies within 1 month before testing. On the day of testing, subjects were instructed not to wear perfume or lipstick and not to eat anything for 11/2 hours before their appointment time. Threshold testing was conducted in a room with an open window with a fan pushing air out of the room. A screen was placed so that the subject could not see the bottle arrangement or coding sheet. Because of the trigeminal nature of PYR, the PEA threshold and suprathreshold tests were administered first, followed by a 15-minute break. Another 15 minutes elapsed before the UPSIT was administered.
The procedure to determine threshold used the two-alternative, forced choice, staircase procedure. 24 Two bottles were presented simultaneously and the subject was instructed to report which one had an odor. For PEA, one bottle contained 100% glycerol (blank) and the other contained a dilution of PEA. For PYR, one bottle contained 100% mineral oil (blank) and the other a PYR dilution. The order of bottle presentation (ie, odorless or dilution, handed right to left or vice versa) was determined using a random numbers table. 25 Subjects sniffed the contents by bringing the bottle close to their nose and squeezing it to release any odor through the open cap orifice. They could sniff each bottle as often as they liked before choosing which of the bottles had an odor. If they thought neither had an odor or that both had an odor, they were required to guess one or the other bottle (ie, forced choice).
Following threshold testing, one squeeze bottle of each dilution previously described or blank was presented twice in random order (thus allowing us to determine a mean value) so that any one dilution or blank did not immediately follow its own presentation, and subjects rated the trigeminal and esthetic qualities of each.
Two separate repeated measures analyses of the esthetic and trigeminal scales were conducted for PEA, and the same repeated measures analyses were conducted for the esthetic and trigeminal responses to PYR. The repeated measures model was a 4 × 4 two-factor treatment analysis of variance (ANOVA), with one factor (group) the between-subjects factor and the other (concentration) the within-subjects factor. To investigate the individual group contributions to the differences between the groups, all possible pairwise comparisons were conducted with Tukey’s Studentized range (honest significance difference) post-hoc tests.
A preliminary multivariate repeated measures ANOVA, followed by a univariate mixed growth curve model, was used to model the polynomial functions that best fit the observed pattern of esthetic and trigeminal responses to PEA and PYR. The between-subjects factor was modeled as a fixed effect, and the within-subjects factor was modeled as a numeric regression variable. A restricted-residual maximum likelihood estimation was made by using the SAS PROC MIXED (SAS Institute, Cary, NC). The same method was used to develop a model for both the esthetic and trigeminal variables.
Odor Identification and Odor Threshold
No statistically significant differences were observed between any of the groups for odor detection thresholds for either PEA or PYR or for odor identification (UPSIT) (Table 3)
Esthetic and Trigeminal Ratings of Suprathreshold Concentrations
Separate repeated measures ANOVAs analyzing esthetic and trigeminal responses to suprathreshold concentrations of PEA and PYR were conducted. The ANOVA analyzing esthetic ratings of PEA revealed significant main effects for group (F = 6.06, degrees of freedom [df] = 3,85, P < 0.0009) and concentration (F = 66.74, df = 3,253, P < 0.0001) and a significant group-by-concentration interaction effect (F = 3.61, df = 9,253, P < 0.0003). The ANOVA analyzing trigeminal responses to PEA revealed a significant main effect for concentration (F = 20.17, df = 3,253, P < 0.0001), a main effect for group that approached significance (F = 2.49, df = 3,85, P < 0.07), and a significant interaction effect for group-by-concentration (F = 2.47, df = 3,253, P < 0.02). The analysis of esthetic and trigeminal responses to PYR, however, showed no significant main or interaction effects. Therefore, no further analyses were conducted for PYR.
Because there was a significant interaction for esthetic and trigeminal response to PEA, simple effects were examined at each of the four concentrations of PEA. For esthetic ratings of PEA, the groups failed to differ at zero concentration but differed significantly at the three other concentrations (Table 4). This pattern is consistent with a dose–response effect, with the groups differing increasingly as the concentration increased. Tukey’s (honest significance difference) tests revealed that MCS subjects reported significantly higher esthetic ratings than the control group at 1% and 0.01% concentrations of PEA. At 1% the CFS group also reported significantly higher ratings than the control group. At 100% concentration the MCS group reported significantly higher ratings than both the control and asthmatic groups, but not the CFS group.
For trigeminal symptoms, the groups failed to differ at the zero or 0.01% concentrations of PEA, but they differed at the 1% and 100% concentrations, consistent with a weaker dose–response effect for this variable (Table 5). Tukey’s honest significance difference tests revealed that MCS subjects reported significantly more trigeminal symptoms than control subjects at 1% PEA. No other differences were found.
Growth Curve Modeling
For esthetic ratings, a preliminary multivariate repeated measures ANOVA revealed that the groups differed significantly with respect to the linear term (F = 2.61, df = 3,83, P < 0.06), the quadratic term (F = 3.69, df = 3,83, P < 0.02), and the cubic term (F = 4.01, df = 3,83, P < 0.02), indicating that the groups differed not only in magnitude of response but also in the slopes of the lines between the concentrations (Fig. 1).
Therefore the utility of linear, quadratic, and cubic terms was tested sequentially with a mixed linear growth curve model, in which the within-subjects factor (concentration) was used as a numeric regression predictor. Preliminary examination of the covariance structure indicated that an autoregressive lag 1 covariance was appropriate, with correlations between concentrations inversely related to the differences in concentrations. This model assessed the significance of the contribution of each increasingly complex term to a model that already includes the previous simpler terms.
These tests indicated that there was a significant linear, quadratic, and cubic trend averaged over the groups, and the coefficients for the terms differed among the groups. Because there were different coefficients for the linear, quadratic, and cubic terms in each group, the model was refit without the average terms (linear, quadratic, cubic) and with a separate intercept for each group. This growth curve model was used to estimate the lines in each group (Fig. 2).
Figure 2 highlights the differences between the esthetic ratings of the groups as the concentration of PEA increases. All three comparison groups (CFS, asthma, and control) responded in a comparable manner to all of the concentrations, with either no change in response or a slight increase in response from 0% concentration to the lowest concentration (0.01%), followed by a greater response to the next two concentrations, and ending with either a slight or moderate increase in response when presented with 100% PEA. The MCS group response, however, was linear, increasing steadily with higher concentrations of PEA.
To model trigeminal ratings, a multivariate repeated measures analysis revealed that the groups did not seem to differ with respect to the linear or quadratic trends, but they differed marginally with respect to the cubic trend (F = 3.07, df = 3,83, P < 0.04). Figure 3 suggests that the only group showing a response to 0.01% PEA was the MCS group. In fact, the control group showed little trigeminal response until reaching 100% PEA. In contrast, the asthma, CFS, and MCS groups reported the largest increase in response from 0.01% to 1% PEA. MCS responses continued to increase over the concentrations in what appears to be a linear function, whereas the asthma and CFS groups showed no appreciable increase in trigeminal response from 1% to 100% PEA.
The same autoregressive lag 1 covariance structure was fit to the trigeminal variable to test the above observations. Using a mixed linear model, the terms in the model were tested sequentially, by using type I tests, which indicated that there was a linear term and that the coefficients for the linear term differed among the groups. Because the quadratic term was non-significant, model-building was discontinued with the linear term.
The model was refit with separate intercepts and slopes for each group (Fig. 4). As indicated, the MCS group had higher scores at all concentrations except for 0%, at which the asthma group was higher. Furthermore, the slope for the asthma group was flatter than those of the three other groups, thus reflecting less of a linear dose–response for the asthmatic subjects.
The current results regarding odor detection thresholds demonstrate that contrary to symptom reports, MCS patients demonstrated no greater ability to identify odors or to detect odors at lower concentrations than did age and gender-matched healthy controls or other patient groups with overlapping features. As expected, neither CFS patients nor asthmatic subjects differed from normal controls in their ability to identify or detect odors.
MCS subjects did not differ from the three other subject groups in their ratings of trigeminal and esthetic symptoms in response to suprathreshold levels of PYR, a known unpleasant trigeminal stimulant. However, they did report significantly more negative esthetic ratings and trigeminal symptoms than the asthma and control groups in response to suprathreshold concentrations of PEA, thus validating clinical observations of reactions to odors regarded as pleasant. The finding that CFS subjects responded differently from the asthmatic and control subjects at 1% and 100% PEA is noteworthy primarily because only a subset of CFS have been documented as reporting sensitivity to chemicals. It is evident that the higher symptom reports of the MCS group, and, to a lesser extent, the CFS group, in response to suprathreshold levels of PEA was not due to a capacity to detect odors at lower levels. In addition, because neither MCS subjects nor CFS subjects reported a greater number of trigeminal symptoms or lower esthetic ratings than the other groups in response to suprathreshold levels of PYR, the possibility of a trigeminal nerve mechanism underlying the difference in these patients seems unlikely. Instead, alternative mechanisms must be involved.
The MCS group, and to some extent, the CFS group, differed from the other groups only in their symptom reports to suprathreshold levels of the more pleasant, non-irritating odor, PEA. This finding is not surprising for the MCS subjects given that the reporting of MCS symptoms is often associated with perfumes and fragrances that are usually at doses below harmful levels and innocuous to most individuals 26 and PEA is, in fact, a common constituent of perfume. The similar response style demonstrated by the CFS group in response to two concentration levels of PEA is significant in that it reflects another similarity between CFS and MCS. The challenge lies in determining the cause of this type of response.
The tendency to report symptoms in response to odors with very low irritancy properties may result from various psychological causes. 27 For example, MCS and CFS patients may demonstrate situational response specificity, in which a specific task, such as smelling particular odors, elicits physiologic responses. 28 For example, the study by Doty et al 3 documented increased heart rate and airflow in MCS patients in response to the perception of an unsafe odor.
The similarities between the CFS and MCS groups are limited in this study in that the presentation of suprathreshold levels of PEA led to a different dose–response pattern in the MCS group for both esthetic properties and trigeminal symptoms. As the suprathreshold concentration of PEA increased, all subject groups reported higher levels of negative esthetic properties and trigeminal symptoms. When presented with 0% concentration of PEA, the MCS group did not report the lowest esthetic ratings. However, when presented with 0.01% PEA, the MCS group demonstrated a steep increase in response. This increase in negative esthetic responses from 0% to 0.01% concentration caused the slope of the curve for the MCS group to appear linear as opposed to the non-linear slopes of the three other groups. This same phenomenon occurred with trigeminal symptoms in response to PEA, whereby the MCS group again did not respond when no PEA was present but instead demonstrated a steep increase in their response pattern from a concentration of 0% to 0.01%, thus forming a linear slope (Fig. 3). Once again, the slopes of the other three groups were non-linear in shape. The MCS group reported a higher magnitude of trigeminal symptoms across all concentrations except 0%, suggesting that once the odor was perceived it was judged to be irritating.
The difference between the slope of the line representing MCS subjects’ responses to PEA and the slopes of the three other groups highlights an important distinction. In contrast to the other groups, the MCS group reacted similarly to each increase in concentration. In other words, the MCS subjects reacted as much to the increase in concentration from 0 to 0.01% as they did from 0.01 to 1%, as well as from 1 to 100%. This difference suggests that even at the lowest concentration of PEA, MCS subjects perceived the odor as irritating and this perception remained consistently higher (when compared with that of the other groups) across increasing concentrations of the odor. Unfortunately, the present study did not assess responses to odors such as judgments or the perceived valences of the odors. Such information may have helped to interpret the differences between groups in their suprathreshold ratings of PEA versus PYR. Even without this information, however, it is clear that for the MCS group, PEA was associated with symptoms. What remains unclear is why PEA, along with fragrances and perfumes, 28 seem to be perceived as irritating, whereas a traditional irritant such as PYR is not.
The present results are particularly meaningful because the definition of MCS used to select subjects was more rigorous than that used in earlier studies (eg, Doty et al 3); specifically, MCS subjects were recruited on the basis of uniform criteria, including Cullen’s 5 definition of chemical sensitivity. The use of such precise subject selection criteria served to create a more homogeneous group of subjects. Furthermore, MCS subjects were compared with two other medically ill groups: asthmatic patients, who have been found to report respiratory symptoms in response to odors, and CFS patients, who have previously been found to report symptoms in response to odors.
Future studies investigating the symptom–response styles of MCS patients to odors may benefit from using an alternative means of measuring odor thresholds, such as signal detection theory, and by measuring physiologic responses to odors. Furthermore, assessing the meaning of odors might shed light on the symptom-reporting differences in MCS subjects.
This research was supported in part by National Institute of Environmental Health Sciences Center Grant ES05022 and NIEHS Superfund Grant ES05955 (parent grant no. U01-AI 32247).
1. Miller CS, Mitzel HC. Chemical sensitivity attributed to pesticide exposure versus remodeling. Arch Environ Health. 1995; 50:119–127.
2. Fiedler N, Kipen H, DeLuca J, Kelly-McNeil K, Natelson B. A controlled comparison of multiple chemical sensitivity and chronic fatigue syndrome. Psychosom Med. 1996; 58:38–49.
3. Doty RL, Deems DA, Frye RE, Pelberg R, Shapiro A. Olfactory sensitivity, nasal resistance, and autonomic function in patients with multiple chemical sensitivities. Arch Otolaryngol Head Neck Surg. 1988; 114:1422–1427.
4. Randolph T, Moss R. An Alternative Approach to Allergies. New York, Lippincott & Crowell; 1980:.
5. Cullen MR. Workers with multiple chemical sensitivities. Philadelphia, Hanley & Belfus, Inc.; 1987: 655–662.
6. Doty RL. The Smell Identification Test. Sensonics, Inc. Haddon Heights, NJ; 1983.
7. Shim C, Williams M. Effect of odors in Asthma. Am J Med. 1986; 80:18–22.
8. Brown EA, Colombo NJ. The asthmogenic effect of odors, smells, and fumes. Ann Allergy. 1954; 12:14–24.
9. Herbert M, Glick R, Black H. Olfactory precipitants of bronchial asthma. J Psychosom Res. 1967; 11:195–202.
10. Brooks SM, Bernstein IL, Raghuprasad PK, Maccia CA, Mieczkowski L. Assessment of airway hyperresponsiveness in chronic stable asthma. J Allergy Clin Immunol. 1990; 85:17–26.
11. Kipen HM, Fiedler N, Lehrer P. Multiple chemical sensitivity: a primer for pulmonologists. Clin Pulmon Med. 1997; 4:76–84.
12. Juniper EF, Guyatt GH, Epstein RS, Ferrie PJ, Jaeschke R, Hiller TK. Evaluation of impairment of health related quality of life in asthma: development of a questionnaire for use in clinical trials. Thorax. 1992; 47:76–83.
13. Shafran SD. The chronic fatigue syndrome. Am J Med. 1991; 90:730–739.
14. Fiedler N, Maccia C, Kipen H. Evaluation of chemically sensitive patients. JOM. 1992; 34:529–538.
15. Buckwald D, Garrity D. Comparison of patients with chronic fatigue syndrome, fibromyalgia, and multiple chemical sensitivities. Arch Int Med. 1994; 154:2049–2053.
16. Kipen HM, Hallman W, Kang H, Fiedler N, Natelson B. Prevalence of chronic fatigue and chemical sensitivities in Gulf registry veterans. Arch Environ Health. 1999;.
17. Fiedler N, Kipen HM. Experimental approaches to chemical sensitivity. Environ Health Perspect. 1997; 105:409–547.
18. Cometto-Muniz JE, Cain WS. Thresholds for odor and nasal pungency. Physiol Behav. 1990; 48:719–725.
19. Doty RL, Shaman P, Dann M. Development of the University of Pennsylvania Smell Identification Test: a standardized microencapsulated test of olfactory function. Physiol Behav. 1984; 32:489–502.
20. Doty RL, Brugger WE, Jurs PC, Orndorff MA, Snyder PJ, Lowry LD. Intranasal trigeminal stimulation from odorous volatiles: psychometric responses from anosmic and normal humans. Physiol Behav. 1978; 20:175–185.
21. Holmes GP, Kaplan JE, Gantz NM, et al. Chronic fatigue syndrome: a working case definition. Ann Intern Med. 1988; 108:387–389.
22. Schluederberg A, Straus S, Peterson P, et al. Chronic fatigue syndrome research: definition and medical outcome assessment. Ann Intern Med. 1992; 117:325–331.
23. Robins LN, Helzer JE. Diagnostic interview schedule (DIS) version III-A. Washington University School of Medicine. 1985;St:Louis, MO.
24. Wetherill GB, Levitt H. Sequential estimation of points on psychometric function. J Math Stat Psych. 1965; 18:1–10.
25. Minium E. W. Statistical Reasoning in Psychology & Education. New York, John Wiley & Sons, Inc.; 1970.
26. Royet JP, Koenig O, Gregorie MC, Cinotti L, et al. Functional anatomy of perceptual and semantic processing of odors. Journal of Cognitive Neuroscience. 1999; 11:94–109.
27. Dalton P, Wysocki CJ, Brody MJ, Lawley HJ. Perceived odor, irritation, and health symptoms following short-term exposure to acetone. Am J Ind Med. 1997; 31:558–569.
28. Lehrer PM. Psychophysiological hypotheses regarding multiple chemical sensitivity syndrome. Environ Health Perspect. 1997; 105:479–483.
© 2000 Lippincott Williams & Wilkins, Inc.