Introduction

Intolerance to odorous and pungent substances includes the clinical diagnoses multiple chemical sensitivity (Multiple chemical sensitivity: a 1999 consensus 1999), idiopathic environmental intolerance attributed to chemicals (IPCS 1996), and sensory hyperreactivity (SHR; Millqvist 2008). Persons with these intolerances react to common volatile substances, such as perfume and cleaning agents that are tolerated by a majority of the general population. Apart from olfaction, the so-called “common chemical sense” plays an important role for these conditions by mediating pungent sensations of irritation, tickling, burning, warming, cooling, and stinging from, predominantly, the nasal cavity and cornea that are mediated by nociceptors and the trigeminal nerve (Doty and Cometto-Muniz 2003). Moreover, it is possible that transient receptor potential (TRP) channels, such as TRPA1 and TRPV1, contribute to chemical hypersensitivity, as well as to chronic cough, airway inflammation in asthma, and chronic obstructive pulmonary disease (Bessac and Jordt 2008).

The symptoms in intolerance to odorous and pungent substances cover a wide range, including airway, affective, cognitive, head-related, cardiac, neuromuscular, musculoskeletal, and gastrointestinal symptoms (Andersson et al. 2009a). Airway symptoms are particularly common among patients with SHR when exposed to pungent substances. Depending on definition, the prevalence of intolerance to odorous and pungent substances in the general population ranges from about 6 and 33 % (Johansson et al. 2005, 2006). The impact on quality of life may be considerable in some of these persons (Ternesten-Hasséus et al. 2007; Larsson and Mårtensson 2009; Söderholm et al. 2011).

The chemical sensitivity scale (CSS) was developed to be analogous to the noise sensitivity scale (NSS; Weinstein 1978) and to quantify affective reactions to and behavioral disruptions by odorous/pungent substances in the environment. The CSS consists of 21 statements (e.g., “At movies, other persons’ perfume and aftershave disturb me”) for the individual to respond to on a Likert scale (Liker 1932). The CSS is easily administrated and inexpensive, and has been shown to have good test–retest reliability, internal consistency, and validity (Nordin et al. 2003) and has normative data (Nordin et al. 2004a). From the 21-item CSS, eleven items were subsequently identified that showed particular sensitivity and specificity for identifying odor intolerant individuals from non-intolerant. Since the intolerant subjects were patients with SHR, the 11-item scale was referred to as the chemical sensitivity scale for sensory hyperreactivity (CSS-SHR). As for the CSS, the CSS-SHR was found to have good test–retest reliability, internal consistency, and validity (Nordin et al. 2004b).

A suggested marker for SHR is increased sensitivity in terms of heavy coughing in response to inhaled capsaicin, indicative of a C-fiber-mediated hypersensitivity of the sensory nerves (Millqvist et al. 1998; Millqvist 2000). In addition to increased sensitivity to capsaicin provocation, diagnostic criteria for SHR include self-reported sensitivity to odorous/pungent substances based on the CSS-SHR. The diagnostic cutoff for a positive CSS-SHR score is ≥43, with a correct classification rate of 92 % (Nordin et al. 2004b). As would be expected, CSS-SHR score is positively related to capsaicin sensitivity (Johansson et al. 2006). Apart from use of the CSS-SHR for investigating patients with suspected SHR, this short scale has proven useful for assessing general forms of chemical annoyance/intolerance (Andersson et al. 2009b, 2009c) as well as pregnancy-induced odor intolerance (Nordin et al. 2005, 2007).

A limitation of the CSS-SHR has been lack of normative data. Based on data from the general population, the objective of the present work was to (1) describe the frequency distribution of CSS-SHR scores, (2) provide a measure of the internal consistency of the CSS-SHR, and (3) establish normative data for the CSS-SHR. The normative data consist of means, standard deviations (SDs), confidence intervals (CIs), and proportions of individuals who met the diagnostic cutoff for a positive CSS-SHR score. Apart from such information for a general population, the objective was to provide reference data for combinations of specific age groups (young, middle-aged and elderly adults) and gender. This was conducted with data from a population-based study, the Västerbotten environmental health study.

Methods

Population and sample

The Västerbotten environmental health study is an embracing name for different investigations on the same general population regarding various forms of environmental intolerance in Sweden. The study population, inhabitants in the county of Västerbotten in Northern Sweden, has an age and gender distribution that is very similar to that of Sweden in general (Statistics Sweden 2011). A random sample, drawn from the municipal register, of 8,600 individuals aged 18–79 years was invited to participate. The sample was stratified for age and gender according to the following age strata: 18–29, 30–39, 40–49, 50–59, 60–69, and 70–79 years. Of the 8,600 individuals, 8,520 could be reached, among whom 3,406 (40.0 %) volunteered to participate. Age and gender distributions are given in Table 1 for both responders and non-responders. The lowest response rate is found among men aged 18–29 years. The participants are further described in Table 2 with respect to demographic and health issues of relevance to intolerance to odorous/pungent substances.

Table 1 Numbers of responders (and response rates) across age and gender strata
Full size table
Table 2 Description of the respondents (n = 3,406)
Full size table

Questionnaire

A questionnaire was used that included the CSS-SHR as well as questions on background information pertaining to the demographic and health issues (Table 2). The items and rating scales of the CSS-SHR are presented in Table 3. The unweighted sum of the eleven CSS-SHR items (after reversed coding of seven items) makes up the individual’s total CSS-SHR score, ranging from 0 to 54 (high score representing high intolerance; Nordin et al. 2004b).

Table 3 Items and rating scales of the chemical sensitivity scale for sensory hyperreactivity
Full size table

Procedure

The participants were mailed the questionnaire, to be returned by mail with prepaid postage. Non-responders received up to two reminders. All participants responded to the questionnaire during the period March–April 2010, before the onset of the pollen season in Västerbotten. The study was conducted in accordance with the Helsinki Declaration and approved by the Umeå Regional Ethics Board. All participants gave their informed consent to participate. IBM SPSS Statistics 19 (IBM Corporation, New York) was used for statistical analysis.

Results

Frequency distribution and normative data

Descriptive statistics for the CSS-SHR are presented in Table 4 for combinations of the different age groups and gender, for the three age groups separately, for gender separately, and for the total sample. The means, SDs, 95 % CIs, and proportions of participants who met the diagnostic cutoff for a positive CSS-SHR score (≥43) constitute normative data for the CSS-SHR. The measures of skewness and kurtosis are close to zero, suggesting approximately symmetrical and mesokurtic distributions.

Table 4 Descriptive statistics for scores on the chemical sensitivity scale for sensory hyperreactivity (CSS-SHR), constituting normative data, and the scale’s reliability (Cronbach’s α) for various sub-groups and for the total sample
Full size table

Internal consistency

Cronbach’s α coefficients of internal consistency for the CSS-SHR are given in Table 4, for combinations of age groups and gender, for the three age groups separately, for gender separately, and for the total sample. These coefficients range from 0.78 to 0.83, suggesting satisfying reliability of the CSS-SHR for all groups.

Discussion

The population-based data from this large-scale study (n = 3,406) suggest that the CSS-SHR has satisfying reliability (Cronbach’s α = 0.78–0.83) and that it generates scores with an approximately normal distribution from young, middle-aged, and elderly adults, from women and men, from various combinations of age groups and gender, and from the general population. Prior research has demonstrated good predictive and concurrent validity of the CSS-SHR and has established a diagnostic cutoff criterion (score ≥43; Nordin et al. 2004b). In addition to investigating the distribution form and reliability, the present study established normative data for the CSS-SHR, including means, SDs, CIs, and proportion of individuals who meet the diagnostic cutoff score. Such data make it possible to quantify individuals or groups with respect to affective reactions to and behavioral disruptions by odorous/pungent environmental substances. This may be valuable in both health-care and research settings. The quantification can be expressed in terms of, for example, percentile, z and T score, and stanines. As would be expected (e.g., Kipen et al. 1995; Nordin et al. 2003), the mean CSS-SHR score obtained in this study was higher in women than in men and tended to increase over the adult life span (cf. Table 4).

The present study population has an age and gender distribution that is very similar to that of Sweden in general (Statistics Sweden 2011), which enhances the generalization of the findings for use of the CSS-SHR on the general Swedish population. Among the randomly selected individuals for this study, 40 % chose to volunteer. This is a somewhat low percentage, which may have consequences for the representativeness of the study. According to research ethical regulations for conducting research in Sweden, it is not allowed to ask the selected individuals why they chose not to participate or about certain characteristics they may possess (Proposition 2007/08:44). However, information on age and gender is available for those who declined participation in this study, and the largest proportion of non-responders was found among young men. Assuming that young men to a relatively low degree are bothered by odorous/pungent environmental substances, the normative data may have higher mean scores and larger proportions of individuals who meet the diagnostic cutoff score than otherwise would have been the case. We can only speculate on the representativeness of the sample concerning other variables. An important variable in this context is intolerance to odorous/pungent environmental substances. It is quite possible that individuals who are high on this variable, compared with those who are low, are more likely to agree to participate, since they may be more inclined to find this type of investigation to be important. If so, this may also have contributed to slightly higher mean scores and larger proportion of individuals who met the diagnostic cutoff score. When using the normative data established from this study, one should therefore keep in mind that the comparison of a CSS-SHR score with the normative data may be a slight underrepresentation of the individual’s affective reactions to and behavioral disruptions by odorous/pungent environmental substances. However, this is not an issue when comparing an individual’s score with the diagnostic cutoff score for the CSS-SHR. With this in mind, the CSS-SHR can be recommended for reliable and valid quantification of affective reactions to and behavioral disruptions by odorous/pungent environmental substances, and with the advantage of being able to compare the score with normality.