• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Assessment of Upper Respiratory Tract and Ocular Irritative Effects of Volatile Chemicals in Humans

Posted on: Thursday, 22 April 2004, 06:00 CDT

ABSTRACT: Accurate assessment of upper respiratory tract and ocular irritation is critical for identifying and remedying problems related to overexposure to volatile chemicals, as well as for establishing parameters of irritation useful for regulatory purposes. This article (a) describes the basic anatomy and physiology of the human upper respiratory tract and ocular mucosae, (b) discusses how airborne chemicals induce irritative sensations, and (c) reviews practical means employed for assessing such phenomena, including psychophysical (e.g., threshold and suprathreshold perceptual measures), physiological (e.g., cardiovascular responses), electrophysiological (e.g., event- related potentials), and imaging (e.g., magnetic resonance imaging) techniques. Although traditionally animal models have been used as the first step in assessing such irritation, they are not addressed here since (a) there are numerous reviews available on this topic and (b) many rodents and rabbits are obligate nose breathers whose nasal passages differ considerably from those of humans, potentially limiting generalization of animal-based data to humans. A major goal of this compendium is to inform the reader of procedures for assessing irritation in humans and to provide information of value in the continued interpretation and development of empirical databases upon which future reasoned regulatory health decisions can be made.

KEYWORDS: cilia, electrophysiology, glossopharyngeal nerve, irritation, magnetic resonance imaging, psychophysics, rhinomanometry, thresholds, trigeminal nerve, vagus nerve.

TABLE OF CONTENTS

TABLE OF CONTENTS

I. INTRODUCTION

The manufacture and use of volatile materials can result in exposures sufficient to cause sensory irritation in community and occupational environments. Public awareness of indoor and outdoor air pollution, including malodor arising from waste disposal plants, farms, pulp mills, and various industrial facilities, contributes to such concerns, and highlights the public consciousness of the potential adverse role of environmental chemicals on physical and mental health. Such concerns continue to increase as urban areas become more congested and the population increases. Modern regulatory agencies, in turn, have been hard pressed to set standards for acceptable levels of exposure to volatile agents within workplace or community air without, in many instances, the benefit of adequate empirically based information regarding the nature of the hazards that are posed or the levels at which the involved chemicals produce adverse health effects. While such standards are obviously necessary, they should be based on empirical data, as they have farreaching consequences for industrial enterprises and governmental agencies, and for consumers and taxpayers who must bear the burden of the cost of meeting the standards that are mandated.

The identification and remediation of sensory irritancy problems have commanded limited attention in the toxicology literature. Although histological procedures are available for assessing the effects of acute or chronic overexposure to volatile chemicals on the nasal epithelia of rodents and other small animals (Doty and Hastings, 2001), these procedures are relatively difficult to perform and lack standardization. The degree to which concentrationresponse relationships from animal data are applicable to humans is also questionable, not only in terms of biosynthetic pathways within the olfactory mucosa (Gervasi et al., 1991), but because, unlike humans, most mice, rats, and rabbits are obligate nose breathers and have more complex nasal passages (Kimbell et al., 1993; Subramaniam et al., 1999). For example, the ethmoidal turbintes arc greater in number and more complex in rodents than in humans, appearing in double, rather than single, rows. These differences are important, as the turbintes, in large measure, determine the pattern and nature of deposition of inhaled chemicals within the nasal passages (Negus, 1958).

The shortage of reliable data for use by health professionals seeking to set exposure standards or assess chemical hazards is due, in part, to the perception by many that sensory irritation or smell sensations cannot be accurately quantified in humans. This is further complicated by the realization that a number of community complaints of sensory irritation or malodor reflect psychosocial, as well as sensory, factors. However, methods employing humans do exist that can aid in (1) establishing concentration-response information for use in hazard and risk evaluations, (2) explaining discrepancies in irritancy and odor databases, (3) separating true adverse health effects from psychosocial factors, and (4) providing insight into exposure-related factors that affect irritancy or odor responses.

In this article we review the basic physiology of how airborne chemicals induce irritative sensations, and provide a state-of-the- art review of common and practical means employed to assess nasal and ocular irritative properties of chemicals in humans. The focus of the article is mainly on irritative effects of chemicals, although some of the techniques that are described can be used to assess responsivity to odorants with few irritative properties, as well as to airborne particulates. While apparent strengths and weaknesses of various techniques are discussed, we have purposely refrained from providing or recommending specific tests at this time for practical applications, given the diversity of chemicals for which such information needs to be gleaned and the variability in efficacy that can arise across different applications within various fields, including toxicology. This review does, however, provide a relatively comprehensive account of the practical techniques available to the toxicologist and others to assess irritative effects of chemicals in humans.*

Organizationally, this article is divided into the four following sections: first, a section defining how the term irritation is used in this article; sec ond, a basic review of the anatomy and physiology of the neural systems within the nose, mouth, throat, and ocular regions that are responsive to volatile agents; third, a description of procedures for presenting irritants to subjects for assessment; and fourth, the presentation of techniques to quantitatively assess the irritative effects of airborne chemicals in the upper airways and eyes. The latter section is divided into four general categories of measurement: psychophysical, physiological, imaging (structural and functional), and psychological (e.g., assessment of community responses by questionnaires).

II. WHAT IS CHEMICAL SENSORY IRRITATION?

The word "irritation" can have different meanings. In the present article this word always refers to chemical sensory irritation-that is, the broad range of physiological responses (including sensory, secretory, respiratory, cellular, and biochemical) produced when airborne chemicals stimulate unspecialized free nerve endings (see Table 1 ). This is distinct from the chemical stimulation of the specialized olfactory or taste receptor cells. The free nerve endings from exposed mucosae, such as the ocular, nasal, oral, and upper respiratory tract mucosae, are very susceptible to being stimulated by chemicals, given that these sites show high accessibility and permeability.

Several terms have been employed that subsume the sensations evoked by chemicals that are typically viewed as irritative. For example, Parker (1912) introduced the concept of the "common chemical sense" to describe general mucosal sensitivity to chemicals (Keele, 1962). More recently, the terms "chcmesthesis" and "pungency" have been used to describe sensations evoked bv chemicals that are not properly odors or tastes. Chemesthesis encompasses mucosal sensations, including those arising from the dermis (Green and Lawless, 1991; Green et al., 1990), whereas pungency refers to nasal and oral chemosensory responses that are mediated principally via the trigeminal nerve (cranial nerve V or CN V) (Cometto-Muniz and Noriega,1985). Among the pungent sensations are those of stinging, piquancy, burning, tingling, freshness, prickling, irritation, and the like. As these sensations grow in intensity, they all can be ultimately defined as irritative or painful. Note that chemesthesis and pungency include sensations beyond those simply viewed as irritative, the latter of which almost always carries a negative connotation as being unpleasant or unwanted.

In this article, nasal or ocular irritation is defined as localized and often unpleasant or annoying chemosensations such as burning, itching, and stinging, as well as associated physiologic (e.g., secretory) phenomena arising from selected mucosa or surrounding tissues (e.g., eyelid) of the target areas involved. Although, in many instances, its sensory referents are equivalent to those implied in the terms pungency and chemesthesis (e.g., at low stimulus levels the actual sensation being mediated by free nerve endings may be subtle and not articulated as irritating), sensory irritation is the preferred term of this review because of its universal general usage by toxicologists, air pollution researchers, and indo\or air scientists and engineers.

III. ANATOMY AND PHYSIOLOGY OF SYSTEMS THAT MEDIATE CHEMICAL IRRITATION AND OTHER CHEMORESPONSES WITHIN THE NOSE, MOUTH, THROAT, AND EYES

Several sensory systems are responsive to airborne chemicals and mediate irritative responses: most importantly, those of the trigeminal, glossopharyngeal, and vagus nerves (cranial nerves V, IX, and X, respectively).* Branches of CN IX and CN X, along with branches of the facial nerve (CN VII), also innervate the specialized taste buds, which convey largely sweet, sour, bitter, and salty sensations induced by liquid-borne tastants. The sensory experiences elicited by the olfactory nerve or CN I (e.g., chocolate, smoke, strawberry, coffee, and lemon) are qualitatively distinct from the somatosensory sensations elicited by activation of the other nerves just noted, which include such sensations as irritation, coolness, warmth, and sharpness.

TABLE 1

Characteristics of the Term "Chemical Sensory Irritation" as Used in This Review

Although the nose, naso-pharynx, and larynx are often described as separate sites of origin for respiratory tract reflexes and sensations, the boundaries of sensory innervation between these areas are rather diffuse (see Figure 1) (Henkin, 1967). This redundancy or overlap complicates our ability to predict the irritancy of a chemical from the activity or response of a single afferent pathway, as many, if not most, vapor phase stimuli will act upon several sites and on several neural pathways in the upper respiratory tract. Even though chemical solubility, together with How rate, determines the dominant patterns of chemical deposition in the upper airways, most inhaled irritant vapors have the potential of contacting multiple sites of mucosal tissue and thereby directly eliciting sensory irritation via the trigeminal, glossopharyngeal or vagal nerves. Reflex responses to inhaled irritants can be elicited indirectly as well. For example, stimulation of the trigeminal fibers in the nose can elicit reflexes from the nasopharynx, while nasal mucus that reaches the pharynx or larynx can elicit cough (even in the absence of direct stimulation by irritant vapors) (Korpas and Tomori, 1979; Widdicombe, 1986a).

Despite the significant potential for contributions from the glossopharyngeal and vagus nerves to the sensation of upper airway irritation in humans, most of the studies of chemosensitivity of these regions have used animals and have focused on measurement of reflexes, not sensation (Widdicombe, 1986b). Thus, little is known about the integrated sensory responses to irritants from these nerves and their relative sensitivity to volatile irritants, especially when compared with thresholds for trigeminally-mediated sensory irritation. In this section, the primary nerves that mediate irritative responses are described, following by a brief description of the olfactory nerve.

FIGURE 1. An artist's representation of the regions within the nasal and oral cavities innervated by each of several cranial nerves. CN I = olfactory nerve; CN V = trigeminal nerve; CN IX = glossopharyngeal nerve; CN X = vagal nerve. CN VII (facial nerve) innervates the taste buds in the anterior tongue and is not shown in this diagram. The cross hatched area represents regions of overlap between CN IX and X. CN I may extend farther down onto the middle turbinate than depicted here. Copyright 2002, Richard L. Doty.

FIGURE 2. Schematic diagram of the branches of the trigeminal nerve that innervate the nasal, oral and ocular epithelia. From B. Bryant and W.L. Silver (2000) Copyright 2000, Wiley-Liss.

A. Trigeminal Nerve (CN V)

The trigeminal nerve (CN V) is the largest of the cranial nerves, being comprised of three major branches, as noted in detail below. One or more of these branches innervate the epilhelia of the nose, forehead and face, nasal sinuses, oral cavity, teeth, eyelids, cornea, temporomandibular joint, the muscles of mastication, and large sectors of the cranial dura (Figure 2), and mediate physically or chemically evoked somatosensory sensations. Such sensations include pain, deep pressure, irritation, coolness, warmth, and sharpness, among others. Irritative sensations are most prominent within the CN V free nerve endings of the mucous membranes.

1. Divisions and Subdivisions of CNV

The three divisions of the trigeminal nerve are the ophthalmic nerve, the maxillary nerve, and the mandibular nerve (Table 2; Figure 2). The ophthalmic division is purely sensory, whereas the other two divisions contain both sensory and motor fibers. The cell bodies of all three divisions are found within the trigeminal ganglion (also termed the semilunar or Gasserian ganglion).

2. Fiber Classes and Receptor Mechanisms of CN V

In general, irritative and other chcmesthetic responses arise, as noted later in this review from activation of polymodal nociceptors within free nerve endings (Silver and Finger, 1991). A number of types of fibers have been found within CN V branches. For example, within the rat's infraorbital nerve, both myelinated and unmyelinated axons are present, with myelinated ones ranging from 0.8 to 14.9 m in diameter and unmyelinated ones ranging from 0.3 to 1.5 m in diameter (Jacquin et al., 1984). More unmyelinated than myelinated axons are contained within the ethmoidal nerve (Biedenbach et al., 1975).

TABLE 2

Divisions and Subdivisions of the Trigeminal Nerve (Cranial Nerve V)

The fine unmyelinated C-fibers that innervate the nasal cavities contain substance P (SP) and, in many cases, associated calcitonin gene-related peptide (CGRP) (Finger et al., 1990). The unmyel inated C-fibers are most likely responsible for irritative reactions in the nasal and respiratory passages, as well as those from the epithelium in general, although small myelinated ?-delta fibers may also be involved (Jansco et al, 1967; Lunblad et al., 1983). Chronic administration of capsaicin, which depletes SP from fine unmyelinated afferents, eliminates or severely reduces trigeminal nerve responses in rats, suggesting that the small unmyelinated and possibly some myelinated fibers subserve trigeminal pain reactions (Silver et al., 1985). Polymodal nociceptors within the free nerve endings of axons belonging to C- and A-delta fibers have been proposed as the mediators of irritation (Martin and Jessell, 1991).

Although a few nonolfactory nerve fibers have been found that extend to the surface of the nasal epithelium (Lunblad et al., 1983), electron microscopic studies suggest that the vast majority of CN V free nerve endings terminate within the lamina propria. Nevertheless, a few trigeminal fibers do terminate within 1 m of the epithelial surface, just below the tight junctions (Finger et al., 1990). Foivolatile chemicals to stimulate these nerve endings, they must (1) pass into the nasal cavity, (2) partition into and diffuse through the mucus, and (3) cross the epithelial membranes and/or intercellular tight junctions. Since many trigeminal stimulants are lipid soluble, such access is not difficult.

Several mechanisms have been proposed to explain how irritative chemicals initiate transduction at the surface of cell membranes (Nielsen, 1991), although the nature of these processes is still poorly understood. Compounds that are chemically reactive (for discussion of reactive/nonreactive, see Alarie et al, 1995, 1996, 1998a, 1998b) can produce irritation directly by reacting with a receptor or indirectly by mucosal tissue damage via chemical reaction without the need to interact with any particular receptor (Nielsen, 1991). In the latter case, damaged cells would release endogenous chemicals such as ATP, H+, and bradykinin which, in turn, could act specifically upon ion channels to produce the neural response (Cesare and McNaughton, 1997; McCleskey and Gold, 1999; Steranka et al., 1987).

Other compounds are likely to act on specific receptors. Eccles et al. (1990), for example, suggest that menthol alters directly the calcium conductance of the trigeminal free nerve ending membranes. As suggested by Jancso and associates (1967; Jansco, 1960), it has been shown that a subset of sensory C-fibers expresses a receptor particularly sensitive to capsaicin, the pungent principle in hot peppers, and to structurally related molecules known as vanilloids (Szallasi, 1994). Interestingly, it has also been shown that this receptor can also be activated by noxious heat (Caterina et al., 1997). Results from recent electrophysiological studies in rats suggest that the irritant nicotine binds to a specific receptor on nasal trigeminal nerve endings (Alimohammadi and Silver, 2000). In fact, electrophysiological studies in rats (Walker et al., 1996), as well as psychophysical and electrophysiological studies in humans (Thurauf et al., 1999), suggest the existence of a dose-dependent stereoselective activation of the trigeminal sensory system by S(-)- and R(+)nicotine. Stereoscleclivity has been similarly noted for other agents. For example, rat studies employing a decrease in respiratory rate as an index of sensory irritation have observed marked differences in potency between various pinene enantiomers (Kasanen et al., 1998).

That being said, however, the great majority of volatile substances found in indoor and outdoor air are common hydrocarbons with varied chemical functionalities such as alcohols, esters, ketones, carboxylic acids, aldehydes, and the like, including linear and branched, saturated and unsaturated, aliphatic and aromatic molecules (Brown et al., 1994; Wolkoff and Wilkins, 1993). Most of these compounds, at high enough concentrations, can trigger trigeminal sensory irritation (Cometto-Muniz, 2001). Given their wide variety in chemical structure, it could be expected that their trigeminal impact rests heavily on general physicochemical parameters that govern the transfer of the irritant from the vapor phase to the trigeminalbiophase where reception takes place. The applicability of a chemical model based on up to five such general physicochemical parameters to describe and predict human thresholds for nasal pungency (Abraham et al., 1998a) and eye irritation (Abraham et al., 1998b) is in accord with this expectation, and is described in detail later in this review. In addition, previous studies have shown the likely existence of a size restriction for molecules of potential irritants to be able to actually evoke irritation (Cometto-Muniz et al., 1998a). The stimulus-size restriction manifested itself in the appearance of a "cutoff" point along homologous chemical series whereby members larger than a certain size would fail to evoke trigeminal sensory irritation in the nose or the eyes. This suggests a need to incorporate a size parameter in chemical models, such as the model just mentioned, to better account for the irritative effects of chemicals (Abraham et al., 2002).

B. Glossopharyngeal Nerve (CN IX)

The glossopharyngeal nerve (CN IX) is named for the main anatomical regions it innervates (glosso, tongue; pharyngeal, beginning of the alimentary canal). It possesses chemosensitive nerve endings within the mucosal lining of the pharynx, except in the anterior portion of the nasopharynx (also termed the epipharynx, a part of the upper respiratory tract behind the soft palate), which is mostly innervated by CN V. This nerve also supplies the taste buds of the posterior tongue (which can respond to some volatile chemicals), and serves both visceral and general sensory functions.

The glossopharyngeal nerve supplies most of the sensory innervation to the nasopharyngeal area, and both mechanical and chemical irritation of the nasopharyngeal mucosa can elicit the aspiration ("gag") reflex, repeated inspiratory efforts, and associated vagal reflexes (Korpas and Tomori, 1979; Tomori and Widdicombe, 1969; Widdicombe, 1996a). Despite anecdotal evidence that sensations of pain, rawness, and irritation from the pharyngeal region follow chemical stimulation, there have been only a few studies that have determined thresholds or otherwise quantified irritant sensations in this area in response to vapor stimuli (Cain et al., 1986, 1987a; Walker et al., 1997).

C. Vagus Nerve (CN X)

Like the glossopharyngeal nerve, the free nerve endings of the vagus nerve also respond to some inhaled vapors. The internal branch of the superior laryngeal nerve supplies sensory fibers to the supraglottic region-the area encompassing all laryngeal regions above the vocal cords. Various types of nerve endings have been identified in and under the laryngeal epithelium. Most are free nerve endings in the mucosa and submucosa that, in animal studies (Hatakeyama, 1969; Lewis and Prentice, 1980), have been shown to respond to a wide variety of gases and aerosols (e.g., ammonia, SO2, cigarette smoke, and CO2) (Andrew, 1956; Boushey et al., 1974). Although little systematic human research on vagal irritation from inhaled vapors has been conducted, recent studies examining oral exposure to liquid ibuprofen demonstrated that pharyngeal irritation can be quantified and that the pharyngeal area appears to be highly sensitive to certain chemical stimuli (Breslin et al., 2001; Rentmeister-Bryant and Green, 1997). However, chemical stimuli can elicit vagal activity through indirect pathways as well. Irritation in the nose elicits vagal reflexes in the lower respiratory tract (Eccles, 1982; Korpas and Tomori, 1979; Widdicombe, 1986b) and is important to understand, as it may lead to laryngeal constriction or secretion of mucus in the lower respiratory tract (Phipps, 1981; Richardson and Phipps, 1978; Widdicombe and Wells, 1982).

D. Olfactory Nerve (CN I)

The olfactory nerve is comprised of 6 million or so receptor cells whose cell bodies and dendritic extensions are located within the olfactory neuroepithelium at the roof of the nasal chambers. The axons of these cells extend from the nasal cavity into the brain (Figure 3). While it is generally believed that this nerve does not produce irritative sensations, per se, there is some evidence that stimulation of this nerve may influence irritative responses of other nerves, most notably the trigeminal nerve. Moreover, most irritants produce olfactory sensations. The olfactory neuroepithelium, which contains a number of cell types in addition to the bipolar receptor cells (e.g., basal cells, microvillar cells, sustentantacular or supporting cells), also harbors trigeminal free nerve endings. This epithelium is found within the region of the cribriform plate, as well as on the superior turbinate, superior septum, and sectors of the middle turbinate (Widdicombe, 1986b). It is noteworthy that the olfactory epithelium loses its homogeneity postnatally, and as early as the first few weeks of life metaplastic islands of respiratory-like epithelia begin to appear, presumably as a result of insults from environmental agents such as viruses, bacteria, and toxins (Nakashima et al., 1984). Such islands increase in extent and number throughout life. Surprisingly, the exact size of the olfactory neuroepithelium in humans is still not well established, and there is recent evidence that it may extend further onto the middle turbinate, at least in some individuals, than commonly believed (Leopold et al., 2000).

The CN I receptor cells are unique, in that they serve not only as a highly specialized receptor cell, but as the first-order neuron, synapsing for the first time within the olfactory bulb of the CNS, just above the cribriform plate. The cilia of the olfactory receptor cells project into the overlying mucus, and differ from the cilia of the cells within the respiratory epithelium in being much longer and lacking dynein arms (hence, intrinsic motility). Odorant transport through the mucus to the cilia is aided by "odorant binding proteins." Approximately 1000 classes of odorant receptors are now believed to exist (Buck and Axel, 1991), reflecting the expression of the largest known vertebrate gene family-a family accounting for ~1% of all expressed genes. However, a large proportion of the odorant receptor genes are, in fact, pseudogenes. In general, the olfactory receptors are linked to the stimulatory guanine nucleotide-binding protein G^sub olf^ (Jones and Reed, 1989). When stimulated, they activate the enzyme adenylate cyclase to produce the second messenger cyclic adenosine monophosphate (cAMP) and subsequent events related to depolarization of the cell membrane and signal propagation (Lowe et al., 1989). Although a given receptor cell seems to express only one type of receptor derived from a single allele (Chess et al., 1994), each cell is electrophysiologically responsive to a wide, but circumscribed, range of stimuli (Holley et al., 1974). This implies that a single receptor accepts a range of molecular entities, and odor coding occurs via a complex cross-fiber patterning of responses. The reader is referred elsewhere for more specific details of the anatomy and physiology of the olfactory system (Doty, 2003).

FIGURE 3. Low-power electron micrograph (670) of a longitudinal section through a biopsy specimen of human olfactory mucosa taken from the nasal septum. Four cell types are indicated: ciliated olfactory receptors (c), microvillar cells (m), supporting or sustentacular cells (s), and basal cells (b). The arrows point to ciliated olfactory knobs of the bipolar receptor cells. d = degenerating cells; bs = base of the supporting cells; lp = lamina propria; n = nerve bundle; bg = Bowman's gland. Photograph courtesy of David T. Moran.

IV. PROCEDURES FOR PRESENTING CHEMICAL IRRITANTS FOR ASSESSMENT

A prerequisite for accurately assessing irritative responses to different concentrations of chemicals is having a means for quantitatively measuring and metering the stimuli that are presented. Stimulus generation and presentation procedures vary considerably, ranging from rather simple squeeze or "sniff" bottles to elaborate devices employing computerized mass flow controllers that allow for generating and presenting mixtures of chemicals in known concentrations and ratios (Figure 4). The stimuli can be presented directly to the nares or eyes, or presented within a chamber or room where "whole-body exposure" occurs. In cases where extended exposure is to be made, stimulus presentation can be done in the subjects' homes or offices using atomizers or other similar devices (Dalton and Wysocki, 1996).

FIGURE 4. The Burghard OM4/B air-dilution oifactometer, a device that presents odorants or irritants to the nasal chambers at well- defined quantities and durations. Left: Subject being presented with nasal stimulants and performing a computerized visual attention task. Right: Data collection module. Center: olfactometer body. Photo courtesy of the University of Pennsylvania Smell and Taste Center, Philadelphia, PA.

Stimulus generation devices for volatile agents are divided by some into two classes: "static," where the stimulus concentration arises from dilutions made in solvents (also termed diluents) (Amoore and Ollman, 1983; Doty, 2000), and "dynamic," where dilutions are derived from active mixing of an airstream containing the irritative substance with a non-odorized carrier airstream (Prah et al., 1995). In some cases, the initial concentration of a stimulus is formed through a static dilution process and subsequent dilutions are performed dynamically. Purely static dilution techniques, however, are the most widely used, largely because of their practicality. In most static systems, a dilution series for a substance of interest is prepared in closed containers using a solvent with little or no odor. The containers can vary in volume from a few hundred milliliters to several liters, and the stimulus is either sniffed directly from each container after it is opened or is ejected or puffed from the cont\ainers into the nose (Cometto- Muniz and Cain, 1990; Doty et al., 1986; Doty, 2000) or into the eye (Cometto-Muniz and Cain, 1991). Some investigators provide ocular exposure using goggles through which the stimulus flows (Hempel- Jorgensen et al., 1999). A number of solvents have been used to produce the variations in concentration, depending upon the solubility characteristics of the stimuli involved, and include distilled/deionized water, USP grade light mineral oil (paraffin oil), diethyl phthalate, and purified propylene glycol. The dilution factor is usually logarithmic, with most workers using binary volume dilution steps. The containers employed are usually glass, although some plastics have been employed [e.g., Teflon, high-density polyethylene (HDPE), and polypropylene (PP)]. Many plastics must be "cured" by lengthy preheating or chemical treatment to eliminate their odor or the odor left from molding oils before use, and care must be taken to assure that the plastics do not react with the stimuli to be employed.

In static systems, an equilibrium is ideally established between the liquid and vapor phases, although the time required for complete vapor saturation can be many minutes, depending upon the substance. At equilibrium, the concentration in the headspace (the actual stimulus) is proportional to that in the liquid. This factor varies among chemical stimuli, solvents, and stimulus-solvent pairs, and often deviates from Raoult's Law (Haring, 1974). In general, the best assurance for an accurate delineation of the concentration of the vapor-phase stimulus is its direct measurement via an analytical instrument such as the gas chromatograph, although quantification of low concentrations may require collection procedures that extend for considerable periods of time. It is important to note that while a given concentration of agent may be presented near the nares or surface of the eye, the actual concentration reaching the epithelia can vary as a function of such idiosyncratic factors as the thickness or composition of the mucus or tear layer, and the shape and size of elements of the nasal chambers (e.g., turbinates), which influence airflow and sorption patterns. Furthermore, the final stimulus becomes diluted to varying degrees with surrounding air. The latter problem is more germane to nasal stimulation, since ocular stimulation is more passive and stimuli can be presented directly to the corneal or epithelial surfaces via goggles or other means. In the case of nasal stimulation, this problem can be mitigated to some degree by placing the orifice of the sniff or squeeze bottle inside of or over the naris. However, because sniff volumes can be several liters, sniffs that outpace the restoration of saturated stimulus to the vessel become diluted with surrounding air. Thus, numerous empirical studies indicate that the volume of the vessel from which a stimulus is presented can influence, over a given range of volumes, sensitivity to volatiles, with larger vessels being associated with greater sensitivity (Cometto-Muniz et al., 2000; Doty et al., 1986).

Dynamic dilution techniques are generally be lieved to provide a more accurate stimulus concentration than static procedures, although they require more complex equipment and also depend upon a stimulus airstream that is assumed to be saturated. Hence, the final concentration should be verified analytically. The reader is referred elsewhere for more specific information on dynamic stimulus presentation (Cain et al., 1992; Dravnieks, 1975; Hempel-Jorgensen et al., 1999; Kendal-Reed et al., 1998; Prah et al., 1995).

While the output of most static and dynamic systems is directed to the proximity of the nares or ocular areas, in some cases the output is sent more generally into a room or environmental chamber, usually in an effort to mimic real-life exposures. In such situations, the subject's nose, respiratory tract, eye, and uncovered skin are concomitantly exposed to the chemical stimulus. Exposures in such situations can continue for hours while the subject rests comfortably and engages in such activities as reading or playing games, making them more amenable to evaluating the build- up of irritative or other responses, that is, the "time" factor (Cain et al., 1986, 1987; Hudnell et al., 1992; Kjaergaard et al., 1991; Molhave et al., 1986; Otto et al., 1990). However, largely because of stimulus control issues (e.g., purging a stimulus before presenting another), experiments in rooms or environmental chambers cannot proceed at the pace of experiments in which the stimulus is directed more locally into the region of the nose or eyes.

V. QUANTITATIVE TECHNIQUES FOR ASSESSING IRRITATION

In this section, we review techniques employed to quantitatively assess the irritative effects of airborne chemicals in the upper airways and eyes. Four general categories of measurement procedures arc presented: psychophysical, physiological, imaging (structural and functional), and psychological (e.g., assessment of community responses by questionnaires). A summary of these procedures is presented in Table 3, along with a listing of general strengths and weaknesses of each approach.

A. Psychophysical Measures

The science of psychophysics-the study of the relationship between perceptual responses and physical stimuli-arose in the mid- 19th century and formed the backbone for much of 20th-century experimental psychology, audiology, and visual science. Today, psychophysical methods are commonly employed to assess chemosensory function in humans in academic, clinical, and industrial settings. As discussed below, psychophysical procedures can be divided into two categories, threshold procedures (where the goal is to detect barely discernible stimuli) and suprathreshold procedures (where the employed stimuli are clearly discernible and such indices as perceived intensity are assessed).

1. Threshold Procedures for Assessing Irritative Responses to Chemicals

Generally speaking, the lowest concentration of an irritant that can be discerned by sniffing or by ocular exposure is considered to be the threshold for irritation or, more simply, the irritation thresh-old. Such a threshold can vary considerably among individuals, and depends not only upon subject factors and the stimuli evaluated, but also upon the specific psychophysical procedure employed for its measurement.

There are numerous paradigms for operationally determining a threshold value (for reviews, see (Doty and Laing, 2003; Woodworm and Schlosberg, 1965). The "classic procedures" were formally developed by Fechner (1860), as outlined in his treatise, Elemente der Psychophysik (Fechner, 1860). More modern techniques employ fewer trials and forced-choice responses. Those that have received the most use in recent years are the ascending method of limits (AML) procedure and the single-staircase (SS) procedure. In the AML procedure, chemicals are presented sequentially from low to high concentrations and the point of transition between detection and no detection is estimated. In the SS method, the concentration of the stimulus is increased following trials on which a subject fails to detect the stimulus, and decreased following trials where correct detection occurs. An average of a number of the up-down transitions ("reversals") is used to estimate the threshold value. The SS procedure is typically more reliable than the single series AML procedure, since a more thorough sampling of the perithreshold region is made (Doty and Laing, 2003). On the other hand, the SS procedure is much more time-consuming and for extremely irritating substances can be trying on the subject. In both the AML and SS procedures, the direction of initial stimulus presentation is made from weak to strong in an effort to reduce potential adaptation effects of prior stimulation. In most cases, a blank is paired with the stimulus at each stimulus concentration level, and the blank and the stimulus are successively presented in a counter-balanced order. The subject is required to report which one seems strongest, the first or the second. This "forced-choice" procedure produces a more stable threshold value than one obtained by simply asking a subject whether something is perceived or not, as it controls to a large degree subject response biases (e.g., liberalism or conservatism in reporting the presence or absence of a sensation in an uncertain situation). The reader is referred elsewhere for more detailed information about forced-choice testing (Blackwell, 1952; Doty and Laing, 2003).

As a general rule, most volatile chemicals that are capable of eliciting irritative sensations (e.g., via the trigeminal nerve) can also elicit an odor (via CN I); furthermore, the odor is evoked at concentrations one or more orders of magnitude below those that evoke irritation (Cometto-Muniz, 2001; Cometto-Muniz and Cain, 1994a, 1996). Thus, when one wishes to establish the lowest concentration of a vapor that can be detected via non-CN I afferents, confusion can arise since the stimulus is already discernible by odor. This is problematic when one wishes to use forced-choice responses against a blank, since the stimulus, whether producing irritative sensations or not, will be apparent to the subject via its odor. To avoid this problem, and still allow for the use of forced-choice procedures, three strategies for assessing irritation thresholds have been devised: (a) to test subjects lacking a functional sense of smell (i.e., anosmics) (Cometto-Muniz and Cain, 1990; Doty, 1995), (b) to test for ocular irritation (which is equivalent in sensitivity to nasal irritation for most volatiles (Cometto-Muniz and Cain, 1995, 1998; Cometto-Muniz et al., 1998c) and (c) to test for nasal localization or lateralization (Cometto-Muniz and Cain, 1998; Dalton et al., 2000; Kobal et al., 1989; von Skramlik, 1925; Wysocki et al., 1997). The latter st\rategy can be employed because, as noted later, irritative, but not olfactory, sensations can be localized to one or the other side of the nose.

TABLE 3

Examples of Classes of General Measurement Procedures with Strongths and Limitations. Selected References, as well as Location within Review, are also Indicated

TABLE 3

Examples of Classes of General Measurement Procedures with Strongths and Limitations. Selected References, as well as Location within Review, are also Indicated

TABLE 3

Examples of Classes of General Measurement Procedures with Strongths and Limitations. Selected References, as well as Location within Review, are also Indicated

a. Nasal Irritation Thresholds Determined in Anosmic Subjects

Anosmics detect many volatile chemicals intranasally via CN V (Cometto-Muniz and Cain, 1993, 1994b, 1995; Doty, 1975). Although anosmia can be due to a number of causes, cognitively normal individuals whose clinically verified anosmia is due to head trauma or to the congenital lack of olfactory bulbs or tracts are preferred for studies of intranasal CN V function since the anosmia is typically complete and permanent. Because anosmics cannot perceive any odor background, they can be tested for nasal detection of chemicals using forced-choice procedures employing blanks. Thus, nasal detection thresholds in anosmics may represent relatively unbiased CN V thresholds that are independent of olfactory input.*

b. Eye Irritation Thresholds

As noted in Section IIIA, the ocular mucosa, as well as the nasal mucosa, is innervated by CN V. Trigeminal chemosensitivity in the eyes can easily be measured in both normosmics (without olfactory interference) and anosmics. Numerous studies employing homologous n- alcohols, n-2-ketones, and alkylbenzenes, selected terpenes, butyl acetate, and toluene have reported intranasal and ocular irritation thresholds to be of equivalent magnitude, and that stimulus- response functions within the perithreshold region are essentially equivalent for most volatiles (Cometto-Muniz and Cain, 1995, 1998; Cometto-Muniz et al., 1998c, 1999, 2001, 2002). Importantly, such studies have found that eye irritation thresholds do not meaningfully differ between anosmic and normal subjects, further validating the use of anosmics in establishing CN V-mediated irritation thresholds. Figure 5 illustrates the comparable irritation sensitivity shown by the ocular and nasal mucosae toward various vapor compounds.

c. Nasal Localization (i.e., Lateralization) Thresholds

It is well established that when blank air is presented to one side of the nose and an irritating chemical to the other, most persons can readily identify the side of the presentation of the irritating chemical. Odorants that have no irritating or other somatosensory effects cannot be so localized (Kobal et al., 1989; Schneider and Schmidt, 1967; von Skramlik, 1925), contrary to what had been reported by von Bekesy, who employed odorants at concentrations that most likely had CN V activity (von Bekesy, 1964). This localization phenomenon provides the opportunity to test directly the chemosensitivity of the nasal trigeminal system in normosmics irrespective of the presence of background odorous sensations. Under this paradigm, two streams of air are directed into the nose, each entering one of the nostrils. One of these streams contains the chemical of interest, and the other not. The task of the subject is to decide which nostril experienced the stronger sensation, not to determine whether something was present or not-in other words, to localize the side of stimulus presentation.

FIGURE 5. Illustration of the similarity of nasal pungency (squares) and eye irritation (triangles) thresholds in humans towards a variety of vapor compounds. Bars, sometimes hidden by the symbol, indicate standard deviations.

Similar lateralization thresholds have not been performed for the eye, although theoretically establishing such thresholds would seem possible. Hempel-Jorgensen and colleagues have demonstrated that humans can distinguish which eye is most irritated when different concentrations of an irritating agent are presented separately to each eye simultaneously (Hempel-Jorgensen et al., 1999). Moreover, they can match the degree of irritation produced in one eye to a reference concentration of CO2 presented to the other eye (see later section on intensity matching procedures). Thus, it would seem straightforward to establish ocular lateralization threshold values.

2. Suprathreshold Procedures for Assessing Irritative Responses to Chemicals

In order to obtain suprathreshold irritation ratings where odor plays no role, the same strategies just described for separating trigeminal from olfactory input at the threshold level need to be considered. As an alternative, normosmics can be instructed to either rate total nasal sensation (Cometto-Muniz et al., 1989), or to rate separately the odor and irritative sensations (ComettoMuniz and Hernandez, 1990; Doty et al., 1978; Kendal-Reed et al., 1998). This last option can have merit in applied studies where the interest is not to study the functional characteristics of the trigeminal chemosensory system, but to assess the overall adverse sensory impact of a chemical stimulus, including one composed of numerous unknown elements. In theory, psychophysical measurements of chemosensation are guided by the same principles as for any other sensory input (Stevens, 1951, 1960). Nevertheless, the specific characteristics of a chemical vapor and of a chemosensory system directly tuned to chemicals introduce practical limitations that are absent in such sensory systems as vision or hearing (Cain, 1978; Cain and Moskowitz, 1974; Doty, 1991).

a. Rating Scales

Since the intensity of an irritant is typically a function of its concentration, ratings or other measures of perceived intensity have been used to evaluate the degree of perceived irritation. Such measures have the advantage of being relatively brief, easy to administer, and less susceptible than threshold tests to subtle stimulus contamination. In chemosensory assessment, two types of rating scales are popular: category scales, where the relative amount of irritation is signified by indicating which of a series of discrete categories best describes the magnitude of the sensation, and line scales (also termed visual analog or graphic scales), where the strength of the sensation is indicated by placing a mark along a line that has descriptors (termed anchors) located at its extremes (e.g., very weak-very strong) and/or midpoint (Figure 6, a and b). Due to their simplicity and ease of use, such scales are common in practical applications.

Numerous studies have employed rating scales in studies of irritation. For example, visual analog scales have been used in studies of nasal (Anton et al., 1992) and ocular irritation to CO2 (Chen et al., 1995; Hempel-Jorgensen et al., 1997) and to other agents (Doty, 1975; Doty et al., 1984), as well as in studies of ocular irritation to n-butanol, 1-octene, and various irritative mixtures (Hempel-Jorgensen et al., 1999). In one study, an annotated 26-cm line served to explore the psychophysical properties of two candidates, pyridine and cis-3-hexen-1-ol, for possible odorization of inert gases in occupational settings (Cain et al., 1987b). A peculiarity of this scale was that it had a mark placed 5 cm from its zero end (i.e., the no odor end), representing the perceived intensity of a comparison reference stimulus: the odor of a 57 ppm 1- butanol vapor presented via a squeeze bottle.

An example of a hybrid between a category and a visual analog scale is one that was used in studies of the odor and irritation of formaldehyde (Cain et al., 1986) and of tobacco smoke (Cain et al., 1987a). In these studies, the scale had six categories. The upper boundary was labeled "None" and, below at equal intervals, the labels read "Slight,""Moderate,""Strong,""Very Strong," and "Over- powering" (at the lowest boundary). Participants used this scale to rate eye irritation, nose irritation, throat irritation, and odor and did so by marking the line at any point deemed appropriate (including between labels). The measurement of interest was the length, in centimeters, from the boundary labeled "None" to the place where the mark was made.

FIGURE 6. Examples of four rating scales. From left to right: (a) A standard category scale in which the subject provides answers in discrete categories; (b) a visual analog or graphic scale with anchors (descriptors) at each end; (c) a category scale with logarithmic visual density referents to denote non-linear increasing magnitudes of sensation, with verbal anchors at each end; (d) a labeled magnitude scale with labels or anchors positioned in logarithmic fashion. In these examples the scales are oriented in a vertical position; in many cases, such scales are presented in a horizontal (left:right) configuration. Copyright 2002, Richard L. Doty.

Several scales have been developed in which logarithmic elements have been incorporated into their design (Figure 6, c and d) in an effort to overcome ceiling effects and to more closely mimic ratio- like properties of magnitude estimation, which is discussed in detail later (Green et al., 1996; Neely et al., 1992). A currently popular scale is the labeled magnitude scale (LMS), which was initially employed to rate oral sensations, such as taste, chemesthesis, and temperature (Green et al., 1993), and was subsequently applied to taste and smell stimuli (Green et al., 1996). The LMS consists of six verbal labels arranged in a roughly logarithmic manner (Figure 6d).

Psychophysical functions produced by the LMS and by magnitude estimation do not differ statistically for a number of chemosensory stimuli, suggesting that the LMS mimics the ratio-like properties of magnitude estimation scaling (Green et al., 1993; Intranuovo and Powers, 1998; Kurtz an\d White, 1998; Lucchina et al., 1998), although it might also mimic the contextual effects that influence magnitude estimation (Lawless et al., 2000). The LMS has been successfully employed in a number of nasal and oral irritation studies (Breslin et al., 2001; Dalton et al., 1997, 2000; Wysocki et al., 1997). The reader is referred elsewhere for discussions of the properties of rating scales, including the influence of category number on their psychometric properties (Anderson, 1970; Doty, 1991; Guilford, 1954).

b. Intensity Matching Procedures

Intensity matching procedures have been used to assess how suprathreshold irritation increases as a function of stimulus concentration, with cross-modal matching procedures (e.g., magnitude estimation) being the most popular. In cross-modal matching, the relative magnitude of each member of a stimulus set is estimated by using some other sensory modality or cognitive domain. A key difference between this procedure and most rating-scale procedures is that the ratio relations among the intensities of the different stimuli are sought, and the subjects' responses are not confined to categories or a short response line. Continua commonly used in the cross-modal matching task termed magnitude estimation include number (e.g., assigning numbers proportionate to the degree of perceived nasal irritation) and distance (e.g., pulling a tape measure a distance proportional to the degree of such irritation) (Berglund et al., 1971; Stevens, 1960). When intensities of sensations from two or more modalities are judged on a single common scale, the procedure is termed the method of magnitude matching (Stevens and Marks, 1980). Magnitude estimation and magnitude matching are among the most commonly used cross-modal matching procedures.

In the prototypical magnitude estimation paradigm, the subject assigns numbers relative to the magnitude of the sensations. For example, if the number 20 is used to indicate the intensity of an irritative response from one concentration of a stimulus, a concentration that seems four times as intense would be assigned the number 80. If another concentration is perceived to be half as strong as the initial stimulus, it would be assigned the value 10. The examinee can assign any range of numbers Io the stimuli, so long as they reflect the relative magnitudes of the perceived intensities. In some cases, a standard for which a number has been preassigned (often the middle stimulus of the series) is presented to the subject in an effort to make his or her responses more reliable. In other cases, the individual is free to choose any number system he or she wishes, so long as the numbers are made proportional to the magnitude of the attribute (the "free modulus method"). For example, one subject may choose to assign the first stimulus the number 25, whereas another may choose to assign this same stimulus the number 5. If a second stimulus is perceived to be 10 times stronger than the first by each of these individuals, the first one would assign the number 250, whereas the second one would assign the number 50. The important point is that the absolute values of the numbers are not important; only the ratios between them are relevant.

It should be noted that procedural and subject factors can systematically influence or bias magnitude estimation measures, perhaps more so than measures from most other suprathreshold sensory procedures (Doty, 1991; Marks, 1974). Magnitude estimation is a relatively complex task, in that accurate responses to a stimulus require a good memory for the prior stimulus. If too much time lapses between the presentation of stimuli, the memory of the prior stimulus fades. However, if the trials are spaced too closely together, adaptation can distort the relationship. Not all subjects can consistently provide ratio estimates of stimuli, and many do not understand the concept of producing ratios (Baird, 1970; Moskowitz et al., 1976).

The degree to which these and other potential shortcomings hinder the use of magnitude estimation procedures in applied settings is unknown; however, presumably such problems can be minimized to a large degree by ensuring that the instructions, stimuli, and test procedures are carefully standardized and monitored. Comparative assessments of nine-point rating scales, line scales, magnitude estimation scales, and a hybrid of category and line scales suggest that, for untrained or mathematically unsophisticated subjects, category scales and line scales may be superior to magnitude estimation when such factors as variability, reliability, and ease of use are considered (Lawless and Malone, 1986a, 1986b). The labeled magnitude scale (LMS) appears to have similarly comparative utilitarian attributes as simple line and category scales.

Because the magnitude estimation function's intercept and distance above the origin depend to a large degree on idiosyncratic differences in the use of numbers and the specific magnitude estimation method employed (e.g., fixed vs. free modulus), only its slope has traditionally been used as an index of sensory function. In an attempt to gain additional information from the function's ordinale position, investigators have employed the method of crossmodal magnitude matching, which provides, at least theoretically, information about the perceived intensity of stimuli from the absolute position of the magnitude estimation function and corrects, to some degree, for differences among subjects in number usage (for a detailed discussion of this procedure, see Marks, 1988). In the most common application of this method, judgments of the intensity of sensations from two modalities (e.g., loudness and the perceived degree of nasal irritation) are made on a common magnitude estimation scale (ASTM, 1988). Under the assumption that subjects experience stimuli on one of the continua (i.e., loudness) in a similar manner, differences among their loudness ratings would be expected to reflect differences in number usage. The irritation intensity continuum can then be adjusted accordingly. Such normalization allows, theoretically, for a direct comparison of scale values across subjects; thus, if the adjusted nasal irritation magnitude value for one subject is 10 and for another subject is 20 at the same concentration level, the second subject is presumed to experience twice the nasal irritation as the first subject.

One way of expressing suprathreshold intensities of chemical sensations is to use a matching procedure employing a scale of "concrete chemical references." In this way, the perceptual scale can be reproduced by other investigators and, over time, retain its intrinsical meaning. An example of this type of scale is that provided by Dravnieks (1975) for the expression of odor intensities. In this paradigm, a participant has to match the odor intensity of a presented stimulus to one of eight butanol concentrations, extending from 16 to 2160 ppm by volume, delivered by an olfactometer. This procedure has been recommended as a standard method for referencing suprathreshold odor intensities (ASTM, 1988).

3. Modeling Perceived Chemical Sensory Irritation

In summary, the success of the general solvation Eq. (3) to describe and predict nasal and ocular irritation thresholds toward a broad range of nonreactive airborne chemicals suggests that transport processes, as denned earlier, are key components of the mechanism through which these compounds exert their chemoesthetic effect. For other substances, for example, nicotine, the key component of such mechanism seems to rest on binding to a very specific receptor (Alimohammadi and Silver, 2000; Thurauf et al., 1999).

4. Psychophysical Responses to Chemical Mixtures

The topic of the toxicology of chemical mixtures is drawing considerable attention nowadays, particularly in the field of risk assessment, given that many exposures in environmentally realistic situations involve the simultaneous presence of a number of substances (Cassee et al., 1998; Korpi et al., 1999; Schaper et al, 1995). Although most human studies addressing fundamental issues concerning function and mechanisms for the sensory irritation potential of volatiles have employed exposures to single chemicals, a few studies addressing issues of direct practical significance have employed complex mixtures. In the latter studies, the identity, number, and concentration of many individual components have remained unknown. Such mixtures have included, for example, tobacco smoke (Cain et al., 1983, 1987a; Walker et al., 1997; Clausen et al., 1984), body odor (Cain et al., 1983; Clausen et al., 1986), carpet emissions (Dietert and Hedge, 1996), and building products emissions (Knudson et al., 1999; Wolkoff, 1999). Studies on sensory reactions to indoor air have employed, in some instances, a model mixture of as many as 22 components (Hudnell et al., 1992; Kjaergaard et al., 1991; Molhave et al., 1991, 1993) believed to be representative of indoor exposures (Kostiainen, 1994; Molhave and Nielsen, 1992; Rottweiler and Schlatter, 1993).

The bulk of the literature on chemosensory detection of chemical mixtures by humans has focused on olfaction (Berglund and Olsson, 1993; Laska and Hudson, 1991; Olsson, 1994; Patterson et al., 1993). Until the various techniques for separating olfactory from trigeminal input were implemented (see section VA2, a, b, and c), studies on the detection of sensory irritation from mixtures relied on asking participants to ignore odor and focus on nasal pungency (Cometto-Muniz and Hernandez, 1990), a procedure that, as discussed in section VA1, cannot control optimally for response biases. A study employing anosmics and including measurements of nasal pungency and eye irritation thresholds for mixtures having three, six, and nine components found various degrees of stimulus agonism that increased with number of components and with the lipoph\ilicity of such components (ComettoMuniz et al., 1997). This work did not include complete detectability (i.e., concentration-response, also called psychometric) (Cometto-Muniz et al., 2002) functions and, thus, only allowed a restricted interpretation of the results. Later studies with binary mixtures included such functions and found support, as a first approximation, to the notion of chemosensory agonism, in the sense of additivity between the components of the mixtures presented at perithreshold levels (Cometto-Muniz et al., 1999, 2001). There are indications that the degree of sensory agonism decreases as the detectability of the mixtures approaches high values (ComettoMuniz et al, 2001). It would be a breakthrough to be able to predict the degree of sensory irritation agonism in mixtures based on the physicochemical and structural properties of the components via, for example, a model such as the solvation equation described in section VA3. Nevertheless, such possibility awaits the availability of results from additional and more complex mixtures where the components cover a wide range of properties and structures.

5. Factors That Influence Psychophysical Measures of Irritation

a. Time of Exposure

Repetitive or chronic exposure to volatile irritants can result in either increases or decreases in perceived sensory irritation, depending upon the stimulus, time course, and nature of the exposure. Increases related to exposure are commonly termed sensitization, although in some cases a gradual increase in the accumulation of the chemical at the target site may explain the enhanced sensitivity or reactivity. Sensitization should not be confused with immunological sensitization, although it is conceivable that, in rare instances, immunological processes might become involved. Decreases in sensation reflect either sensory adaptation, which is often peripheral, or habituation, which can involve more central circuits. In general, habituation is more amenable to modulation from higher order central nervous system processes, such as arousal or cognitive processes, than is adaptation (Thompson and Spencer, 1996).

There are numerous examples of apparent sensitization to airborne irritants (Cain et al., 1986, 1987a; Hudnell et al., 1992). For example, Hudnell and others (1992) found, in a 2.75-hour-long chamber exposure, the intensity of nose, throat, and eye sensory irritation increased as a function of the duration of exposure to volatile organic chemicals, with the perceived eye irritation being concentration related. More recently, Hempel-Jorgensen and colleagues (1999) examined the time course of sensory eye irritation to n-butanol and 1-octene in 16 subjects, demonstrating consistent 10-fold increases in perceived irritation following 20 to 40 min of exposure, which thereafter remained relatively constant. In the case of 1-octene, but not n-butanol, sensitization remained for some time after the removal of the stimulant.

While exposure-induced adaptation can produce dramatic reductions in nasal irritant sensations elicited by a volatile chemical, the time course of such reduction appears to be longer than for a primarily olfactory stimulus (Cain et al., 1986). As in the case of olfaction, adaptation can be relatively specific to the compound to which an individual is exposed. For example, repetitive occupational exposure of textile workers to acetone elevated the nasal irritation threshold and decreased the perceived magnitude of irritancy for acetone. These changes were not observed for butanol (Dalton et al., 1997; Wysocki et al., 1997). Similarly, the isopropanol irritation thresholds of phlebotomists who were regularly exposed to isopropanol in the workplace were elevated, but their irritant thresholds for butanol did not differ from unexposed, naive controls (Smeets and Dalton, 2002).

b. Subject Characteristics

i. Smoking Behavior

A few studies have looked into the issue of whether smoking affects the perception of sensory irritation (Cometto-Muniz and Cain, 1991b). Measurement of a transitory reflex apnea induced by the nasal pungency produced by CO2 has shown that smokers are less sensitive, that is, present thresholds for the reflex that are 29% higher than nonsmokers (Dunn et al., 1982). Nasal detection thresholds for CO2 can be 44% higher in smokers than in nonsmokers (Shusterman and Balmes, 1997a). Even immediately after short periods of smoking (6 to 10 min) smokers showed a further 12% decrease in sensitivity to this reflex (Cometto-Muniz and Cain, 1982), indicating that, on top of a chronic reduction in nasal pungency sensitivity, smokers experience an acute desensitization right after smoking a cigarette. A recent study by Millquist and Bende (2001) reports that coughing induced by capsaicin is decreased in smokers, in accord with this general notion. Use of the virtually odorless stimulus CO2 in measurement of the transitory apnea or in a forced- choice detection task has been quite common for assessing nasal irritant sensitivity (Shusterman and Balmes, 1997b).

ii. Gender

Again using CO2 as the stimulus and reflex apnea as the outcome, it has been shown that females are 14 to 30% more sensitive than males to nasal pungency whether evoked unilaterally or bilaterally (Dunn et al., 1982; Garcia-Medina and Cain, 1982). Nasal detection thresholds for CO2 have also been found to be lower for females than for males (Shusterman et al., 2001). Further experiments employing a magnitude matching technique (see Section VA3b) revealed that females produced steeper stimulus-response functions for nasally evoked CO2 pungency and that they were actually experiencing between 50 and 67% more nasal pungency from the same range of CO2 concentrations than their male counterparts (ComettoMuniz and Noriega, 1985). Interestingly, no di

More News in this Category