Perceived odor, irritation, and health symptoms following short-term exposure to acetone

Perceived Odor, Irritation, and Health SymptomsFollowing Short-Term Exposure to Acetone

Pamela Dalton, PhD,* Charles J. Wysocki, PhD, Michael J. Brody, BA, and Henry J. Lawley, BS

The subjectivity of irritancy judgments can bias attempts to establish exposure guidelines thatprotect individuals from the sensory irritation produced by volatile chemicals. At low tomoderate chemical concentrations, naive and occupationally exposed individuals often showconsiderable variation in the reported levels of perceived irritation. Such variation couldresult from differences in exposure history, differences in the perceived odor of a chemical, ordifferences in generalized response tendencies to report irritation, or response bias. Thus,experimental evaluation of sensory irritancy must dissociate sensory irritation from responsebias. To this end, judgments of perceived irritation from 800 ppm acetone were obtained fromacetone-exposed workers and age- and gender-matched naive controls. To assess the role ofresponse bias during exposure to odorants, subjects were also exposed to phenylethyl alcohol(PEA), an odorant that does not produce sensory irritation. Following exposure, subjectscompleted a subjective symptom survey that included symptoms that have been associatedwith long-term solvent exposures and symptoms that have not.

Acetone-exposed workers and naive controls reported large differences in the perceivedintensity of odor and irritation from acetone, yet no differences in the perception of PEA.However, for both groups, the most significant factors mediating reported irritancy and healthsymptoms from acetone were the perceived intensity of its odor and an individual’s bias toreport irritation from PEA. The perception of odor intensity and degree of response bias willdiffer between and within groups of exposed and naive individuals; hence, an assessment ofthe influence of these factors in experimental and workplace studies of chemical irritancy iswarranted.Am. J. Ind. Med. 31:558–569, 1997.r 1997 Wiley-Liss, Inc.

KEY WORDS: sensory irritation; odor; acetone; response bias; occupational exposure;health

INTRODUCTION

The highly subjective nature of perceived irritation cancause difficulties in establishing exposure guidelines toprotect workers from sensory irritation produced by volatilechemicals. At sufficiently high concentrations, many com-pounds unequivocally produce irritation in the eyes and

upper respiratory airways (chiefly, although not exclusively,through mediation of the trigeminal nerve [Alarie, 1966]).The intent of the guidelines is to restrict exposures to levelsbelow which irritation occurs; hence, the exposure limitstypically target low to moderate ranges of chemical concen-trations. At these levels, naive and exposed individuals oftenshow considerable variation in the levels of perceivedirritation they report.

Interindividual variation in the subjective irritancy of achemical could result from differences in exposure history ifpeople adapt to a compound’s irritancy following repetitiveexposures, just as they adapt to its odor [Cain, 1970;Cometto-Mun˜iz and Cain, 1995; Chang and Barrow, 1984;Dalton and Wysocki, 1996]. Alternatively, variation couldresult from the many nonsensory factors that are thought to

Monell Chemical Senses Center, Philadelphia, PA.Contract Grant sponsor: Chemical Manufacturer’s Association; Contract Grant

sponsor: National Institutes of Health; Contract Grant numbers: F32 DC-00197,R01 DC-00298, and P50 DC-00214.*Correspondence to: Pamela Dalton, Ph.D., Monell Chemical Senses Center,

3500 Market Street, Philadelphia, PA 19104.

Accepted for publication 30 July 1996

AMERICAN JOURNAL OF INDUSTRIAL MEDICINE 31:558–569 (1997)

r 1997 Wiley-Liss, Inc.

influence perception [Dalton, 1996; Swets et al., 1961;Williams and Lees-Haley, 1993]. Although the irritancy thatis attributed to a volatile chemical can result from stimula-tion of chemoreceptors in the eyes or airways, judgments ofirritation can also result from the psychological discomfortincurred by smelling the odor of an unfamiliar chemical, oreven from misattributing the symptoms of an unrelatedillness or stress that coincide with exposure. Thus, asignificant concern in the experimental evaluation of sensoryirritancy is the degree to which a study can separate truesensory irritation from generalized response tendencies toreport irritation, orresponse bias[Swets et al., 1961].

The role of response bias surfaces frequently in themedical and psychological assessment of chronic pain.Individuals can differ both in their sensitivity to painfulstimuli and in the amount of pain they will experience beforereporting the sensation as ‘‘painful.’’ An individual’s crite-rion (response bias) for reporting the presence or intensity ofpain can be shifted by social factors or expectations aboutthe source of the pain [Mechanic, 1972]. Irritation from achemical could be equally susceptible to variables thatinfluence the perception of pain because chemical irritationis thought to reside at the lower end of the pain continuum[Jones, 1954].

One approach to addressing subjectivity found in hu-man studies of chemical irritancy has been to use animalmodels to assess the irritancy of airborne chemicals [Alarie,1966]. However, some authors have raised concerns regard-ing the reliability and utility of this technique for establish-ing human exposure guidelines [Bos et al., 1992]. A secondapproach to separating irritation from response tendencieshas been to use a clean air control. Previous studies ofirritancy to airborne chemicals have contrasted responsesduring exposure to the chemical with responses made duringexposures to clean air [Cain et al., 1986; Hudnell et al.,1992; Otto et al., 1992; Seeber et al., 1992]. However, mostvolatile chemicals also have an odor. For this reason,individuals may adopt very different criteria for reportingirritation in the presence of an odor than in the presence ofrelatively odorless air.

This paper reports a study investigating the degree towhich repetitive exposure to a volatile chemical can mediatethe perceived odor and irritation from that chemical during ashort-term, chamber exposure by comparing the responsesfrom occupationally exposed workers and occupationallynonexposed controls. Recognizing that responses from thesegroups can differ based on nonsensory factors (e.g., bias toreport irritation), as well as sensory factors (e.g., exposure-induced adaptation), the study design includes controls thatmake it possible to separate sensory irritation from responsebias.

Exposure to weak to moderate concentrations of ac-etone (2-propanone) has produced considerable variation in

reports of perceived irritation and discomfort. Response biasmay be a contributing factor to this variation. Amongindustrial solvents, the toxicity and irritancy of acetone havebeen established as relatively low [Kane et al., 1980;Morgott, 1993]. Eight-hour occupational exposure limits inthe United States generally range from 750 to 1,000 ppm,depending on the organization setting the limits [AmericanConference of Governmental Industrial Hygienists, 1993;OSHA, 1989] Nevertheless, reports of human exposures toacetone have noted that individuals experience irritation ofthe upper respiratory airways or mucosal membranes atconcentrations below occupational exposure levels [Dick etal., 1989; Kiesswetter et al., 1994; Matsushita et al., 1969a].However, the concentration at which irritancy is said tooccur differs widely for exposed workers and nonexposedcontrols [Nelson et al., 1943; Raleigh and McGee, 1972].Unfortunately, studies evaluating the irritancy of acetone inindividuals with different exposure histories have not concur-rently compared their responses to a control compound, towhich they were not previously exposed Thus, it is notknown whether the perceptual changes following long-termexposure to acetone represent specific adaptation to the odorand irritation of acetone or indicate more general impair-ment of olfactory and irritation perception.

To assess the level of perceived odor, irritation, andsubjective health symptoms from a brief exposure to acetoneand also to investigate whether those responses can bemediated by specific experience with the chemical, acetone-exposed workers and age- and gender-matched nonworkercontrols were given short-term, chamber exposures to bothacetone and a novel control odorant–phenylethyl alcohol(PEA). PEA is considered to be a pure olfactory stimulus—its odor can be detected by individuals with intact olfactoryabilities, but in vapor phase it does not produce sensoryirritation at any concentration [Cometto-Mun˜iz and Cain,1990; Doty et al., 1978; Kobal, 1982] and cannot be detectedby people who lack olfaction [Doty et al., 1978]. Thus,reports of irritation during exposure to PEA can be inter-preted asfalse positives[Swets et al., 1961] and can serve asa measure of the individual’s response bias to reportirritation. Finally, ratings of regional irritation and a varietyof subjective health symptoms were obtained in order toexamine the extent to which responses on subjective symp-tom questionnaires are due to response bias.

MATERIALS AND METHODS

Subjects

Twenty-seven workers (23 females, 14 smokers) in acellulose fiber production plant volunteered to participate inthe study. Subjects were recruited from a possible pool of778 workers (612 females, median age 43, range 19–77),

559Perceived Symptoms Following Short-Term Exposure to Acetone

by means of postings throughout the plant one month priorto the study. The median age of the worker subjects was 38years (range 23–65). All were full-time employees and hadworked in their present position for a minimum of 12months (median 10 years). Acetone was the primarysolvent to which they were occupationally exposed. Mostworkers were employed as ‘‘doffers’’ (Table I), whoseprimary task is to remove and replace full bobbins ofcellulose acetate yarn in the spinning process. Theremaining workers were employed in other cellulose fiberprocess areas. Workers were screened for respiratoryproblems or allergies as a condition for participation. Thehealth status of the participants did not differ from thegeneral population of workers from whom they wererecruited.

Thirty-two, individually age- and gender-matchedcontrol subjects (14 smokers) were recruited inPhiladelphia via newspaper advertisements and paid fortheir participation. The volunteers were screened for res-piratory problems or allergies and any history of exposureto solvents or chemicals. An additional five control subjectshad to be recruited because the exposure concentrationvaried from 4% to 10% above or below the target level forfive subjects, necessitating that their data be discarded. Thisleft 27 subjects for analysis in each group. Sixteen controlsubjects were unemployed at the time of the study, 7worked in offices and the remainder were full or part-timestudents. None of the control subjects had any history ofoccupational exposure to chemicals.

Experimental Procedure

All subjects were tested individually in a single sessionthat lasted approximately 2 hr. Workers assumed theirregular duties for at least 1 hr (and no more than 5 hr) priorto the test. The session consisted of multiple tasks. Odordetection thresholds were obtained for two compounds,acetone and 1-butanol, immediately before and after thechamber exposure to acetone. The logic behind obtainingmeasures of sensitivity to these two odors was to observe

whether responses to suprathreshold concentrations of ac-etone in the chamber were paralleled by changes in thedetection threshold following exposure. Butanol served asthe control substance; exposure to acetone should not affectsensitivity to butanol. Each subject spent 20 min in each oftwo chambers in which they rated the odor and irritationintensity of 800 ppm acetone or 200 ppm PEA.

Immediately after each chamber session, subjects com-pleted a questionnaire in which they rated the occurrenceand/or intensity of a variety of health symptoms. Followingthe last chamber exposure, subjects were debriefed and thendismissed.

Chamber Exposures

Exposures were performed in portable chambers, bothat the factory and at Monell. The chambers measured 53 m3

and were constructed of PVC pipe covered with polypropyl-ene sheeting. Ventilation was accomplished by an exhaustfan and a blower installed inside the chamber which allowedcomplete air changes at the rate of 20 times/hr. Odorizing thechamber was accomplished by bubbling air through theliquid compound (i.e., acetone or PEA) and mixing theoutput with the chamber’s influent air stream. Certified ACSgrade acetone (Fisher, Milwaukee, WI) with a stated purityof .99.5%, and PEA (Aldrich, Pittsburgh, PA) were usedfor the exposures. Temperature and humidity were main-tained at 25C (62°C) and 50% (67%), respectively. Theconcentrations inside the chamber were monitored with anon-line photo-ionization detector (Rae Systems, Inc, Sun-nyvale, CA) and the verification of vapor concentration wasaccomplished by gas chromatography. Samples of thechamber air were collected in Tedlar bags and analyzed oneither a Hewlett-Packard 5890 or 5890 II GC equipped witha 128 3 1/89 stainless steel column (Supelco, Bellefonte, PA;3% SP-1500 on 80/120 Carbopack B), a flame ionizationdetector and a gas-sampling valve. The actual exposureconcentration of acetone did not vary more than 3% from thetarget of 800 ppm for any of the subjects included in thisstudy. The exposure concentration of PEAdid not vary morethan 10% from the target level of 200 ppm for any of thesubjects.

Before entering the chamber, subjects had been in-formed that the odorant concentration in the air could varyduring the session or stay the same. These instructions weredeemed necessary to reduce the influence of experimentaldemand characteristics and to encourage reports based onthe subject’sperceptual experience, rather than reportsbased on their own hypotheses concerning the experiment.In fact, odorant concentration was maintained at a steady-state throughout the exposure session. Subjects were in-structed to rate the odor intensity and the irritation intensityof the room air by moving a cursor to the appropriateposition of a computerized line scale [Green et al., 1993]. At

TABLE I. Occupational Acetone Exposure (ppm)

n Job title Average 1995 Average (1982–1994)

18 Doffer 568 585

2 Cone Examiner 291 216

2 Bobbin Stores Operator 415 387

1 Supervisor (Extrusion) 414 414

1 Physical Analysis Lab 648 643

1 Extrusion Mechanic 330 330

1 Engineer 114 114

1 TG Operator/Cone 350 216

560 Dalton et al.

subsequent 1-min intervals (for a total of 20 times persession), subjects rated the perceived odor and the perceivedirritation from the room air.

The computerized scale used for the ratings of odor andirritation, the Labeled Magnitude Scale (LMS), is depictedin Figure 1. The scale has several significant advantagesover other popular scaling methods: it yields ratio level dataand, because it allows subjects to use natural languagedescriptors to scale perceived experience, it produces abso-lute intensity estimates as well. Its use has recently beenvalidated for scaling odorous chemicals that produce irrita-tion [Green et al., 1996].

Odor Detection Thresholds

A 20-step binary dilution series was used for acetone. A26-step tertiary dilution series was used for butanol. High-performance liquid chromatography (HPLC) grade acetoneand butanol (Fisher), with a stated purity of.99.5% and.99.4%, respectively, were used as solutes in the dilutionseries. The solvent for acetone dilution was propylene glycol(Sigma, St. Louis, MO); the solvent for butanol wasodorless, light, white, mineral oil (Sigma), which wasfiltered through silica gel before use. The odorant dilutionsteps were presented in cleaned, 280-ml, glass bottles fittedwith a flip-top cap into which a Teflon nosepiece wasinserted. The acetone series consisted of the undilutedsubstance (100% v/v: dilution step 0) and subsequent binaryliquid dilutions (e.g., 50%, 25%, 12.5% v/v), representing

dilution steps 1-20. The dilution series for butanol consistedof the undiluted substance (100% v/v: dilution step 0) andsubsequent tertiary liquid dilutions (e.g., 33%, 11%, 3.7%,v/v), representing dilution steps 1–26. Each bottle contained10 ml of solution. Blanks consisted of 10 ml of theappropriate solvent.

Two identical odorant series were prepared fresh foreach day of testing from the same stock dilution. One serieswas used for threshold tests while the second series was usedto analyze the headspace concentration in each dilution step.The vapor concentration in the headspace of each bottle wasmeasured with a gas chromatograph (HP 5890A or HP5890II with a flame ionization detector). The GC wasequipped with a gas sampling valve that permitted directheadspace sampling. Repeated samples were taken fromevery other dilution step (even or odd numbers on alternat-ing days except for step 0—undiluted compound—whichwas never presented to subjects). Thus, bottles representingthe entire dilution series were sampled every two days.Ten-point calibration curves for acetone and butanol weregenerated from injection of known amounts and theirassociated chromatographic readings and were used toconvert the readings from the dilution steps into concentra-tion units (ppm by volume). The coefficient of determination(r2) for each calibration curve was typically.0.96, and not,0.92.

Thresholds were estimated by using a two-alternative,forced-choice, up/down, staircase method with a five-reversal criterion (described in more detail in Dalton andWysocki, [1996]). On each trial, the subject was presentedwith two pairs of bottles. Each bottle pair consisted either ofan odorant in solvent and solvent alone (stimulus pair) ortwo bottles of solvent (blank pair). Subjects simultaneouslyinserted the nosepiece from each pair of bottles snugly intheir nostrils and took a normal sniff from the air in theheadspace. After sniffing from both pairs, subjects wereasked to identify which pair contained the odorant. Presenta-tion order and location (left/right nostril) of the pairs wererandomized between trials.

Thresholds were obtained for both odorants in the sametest; acetone and butanol were presented on alternating trialsuntil the threshold criterion was reached for each odorantseparately. In all cases, we used the value representing thefirst time the criterion was obtained for a given odorant. Thestarting concentration used for both odorants on the first test(baseline) was near the midpoint of the series (the 10thdilution step). On the second threshold test, the startingconcentration was one step weaker than the baseline thresh-old test.

SymptomQuestionnaire

To assess both the frequency of occurrence and theperceived intensity of regional irritation and health symp-

FIGURE 1. The Labeled Magnitude Scale used to obtain ratings of odor, irritationand health symptoms. The numbers on the left of the line are arbitrary units,

presented for analytic reference; subjects saw only verbal labels when they used

the scale.


toms as a consequence of exposure to acetone or PEA, acomputerized questionnaire employing the LMS was pre-sented to subjects immediately after each exposure. Thequestionnaire was adapted from that used by Hudnellet al. [1992] and included items that have been reportedfollowing exposure to solvents at high concentrations (i.e.,eye, nose, or throat irritation, lightheadedness, headache,drowsiness andnausea), as well as 18 somatic or control itemsnot associatedwith solvent exposure (e.g., back pain, leg cramps,numbness in extremities, ear ringing) [Agency for Toxic Sub-stances and Disease Registry, 1994; Baelum et al., 1982;Elofsson et al.,1980; Iregren, 1986; Husman, 1980; Morgott,1993].

Data Analysis

Separate repeated-measures analyses of variance(ANOVA) were used to analyze the odor and irritationratings as a function of group (acetone-exposed workers vs.controls), compound (acetone or PEA), and exposure time(20 ratings at 1-min intervals). Differences among thetreatment groups were evaluated by post-hoc tests (Tukey’sHSD). Odor detection thresholds were not normally distrib-uted, hence, the nonparametric Mann-Whitney Rank SumTest was used to examine the effects of group (acetone-exposedworkers vs. controls), compound (acetone or butanol)and short-term exposure to acetone (pre- and postexposuretest). For the questionnaire intensity ratings, a repeated-measuresANOVAevaluated overall differences between thegroups and compounds for each item; significant differenceswere further evaluated by planned comparisons of itemresponse profiles within and between groups and com-pounds. Finally, multiple regression and correlation analysis[Cohen and Cohen, 1983] evaluated whether demographicfactors, such as age, smoking, and acetone-exposure history

predicted odor or irritation perception at threshold orabove-threshold concentrations and the resulting coeffi-cients were tested for significance with ANOVA.

RESULTS

Perceived Odor and Irritation

The summary of statistical results on perceived odorand irritation, as a function of subject group, compound andexposure duration, are presented in Table II. Occupationalexposure to acetone reduced the perceived odor intensity of800 ppm acetone, relative to the perception of thenonexposed control subjects (p, 0.001; see Figure 2). Onaverage, the workers rated the odor of acetone as weak tomoderate, whereas control subjects rated the odor as strongto very strong.

Both acetone-exposed workers and controls rated theodor of PEA as weaker than the odor of acetone (p,0.001;compare Fig. 2, left and right). However, an appreciableinteraction between group and compound suggested thatdifferences existed among the groups in the rated odorintensity of the two compounds. Post hoc comparisons ofthe means revealed that workers and controls differeddramatically in their perception of acetone odor (p,0.001)but did not differ in their perception of the odor intensity ofPEA, suggesting that occupational exposure to acetoneproduces a specific decrement in the odor intensity ofacetone.

Both workers and controls showed a general effect ofexposure duration on the perceived odor intensity ofacetone and PEA (p5 0.001). A 20-min exposure to theodor of 800 ppm acetone produced a 46% reduction inaverage perceived intensity for the controls, and a 28%reduction for the workers. Similar degrees of adaptation

TABLE II. Summary Statistics for Chamber Ratings of Odor and Irritation

Source

Odor Irritation

dfa MSeb Fc pd dfa MSeb Fc pd

Group (workers vs. controls) 1, 52 3.229 11.70 0.001 1, 52 7.185 17.51 0.000

Compound (acetone vs. PEAe) 1, 52 1.677 47.90 0.000 1, 52 1.545 71.89 0.000

Test exposure duration 19, 988 0.079 61.41 0.000 19, 988 0.097 1.16 0.280

Group 3 compound 19, 988 1.677 12.18 0.001 19, 988 1.545 26.80 0.000

Group 3 exposure duration 19, 988 0.079 1.17 0.276 19, 988 0.097 1.97 0.008

Compound 3 exposure duration 19, 988 0.073 1.28 0.183 19, 988 0.083 0.85 0.646

Group 3 compound 3 duration 19, 988 0.929 0.55 0.475 19, 988 0.083 1.13 0.318

aDegrees of freedom.bMean square error.cF-test.dProbability.ePhenylethyl alcohol.

562 Dalton et al.

were seen for PEA (30% and 36%, for controls andworkers, respectively).

Responses to the perceived irritation from acetone andPEA mirrored the results for odor (Fig. 3). Post-hoccomparisons on the means for these groups revealed thatoccupationally exposed workers perceived the irritationfrom acetone as significantly lower than did thenonoccupationally exposed controls (p,0.001) consistentwith their perception of the odor intensity. There was also asignificant difference in perceived irritation from the twocompounds (p,0.001) with less irritation reported fromexposure to the nonirritant control, PEA, than fromexposure to acetone. However, post-hoc comparisons of themeans showed that the difference in perceived irritationfrom acetone and PEA was restricted to the non-occupationally exposed controls (p,0.001). By comparingthe workers’ ratings of irritation from acetone and PEA inthe two panels of Figure 3, it is obvious that, on average,workers who were occupationally exposed to acetonereported very low levels of irritation from exposure toeither acetone or PEA.

There was no main effect of exposure duration on theirritation ratings, however, the reduction in perceivedacetone irritancy across the session for control subjects, butnot workers, produced a significant interaction (p,0.001)between group and duration.

When the average acetone odor and irritation ratingsprovided by each individual were simultaneously regressedon the variables of age or smoking status, neither factorproved to be significant a predictor of perceived odor orirritation. Variation in the amount of occupational exposureto acetone also did not influence the perceived odor orirritation from acetone. Two measures of acetone exposurewere calculated for each worker: the average level of daily

exposure in their current position and a cumulativeexposure level, calculated by summing across the averageexposure level for each six-month period of theiremployment. Neither index of occupational exposure toacetone was predictive of the perceived odor or irritationfrom acetone.

Thresholds for Odor Detection

Pre- and postexposure thresholds were tested for (1)differences in sensitivity to acetone or butanol betweenacetone-exposed workers and nonexposed controls, and (2)effects of short-term exposure to acetone in the chamber onthreshold sensitivity to acetone or butanol.

As shown in Table III, neither the mean nor the mediandetection thresholds for either acetone or butanol differedbetween the workers and the control subjects (p5 0.24).Short-term exposure to acetone produced a greater shift insensitivity to acetone for the nonexposed controls than it didfor the exposed workers; however this difference failed toreach significance (p5 0.11). The regression analysisshowed that neither smoking status, age, gender or exposurehistory were related to threshold sensitivity for acetone orbutanol.

SymptomQuestionnaires

Group average intensity ratings for the 22 items areshown in Table IV, organized by compound (acetone andPEA), by group (workers and control), and by symptom type(solvent-related and somatic). Current irritation was re-ported separately, as it referred to any persistent irritation thesubject was experiencing when answering the questionnaire,which was typically 2–4 min after exiting the chamber. Asshown in Table V, responses to the symptom questionnaire

FIGURE 2. Average intensity ratings of the perceived odor of acetone and PEAfor acetone-exposed workers and non-exposed controls during a 20-min chamber

exposure to each compound. The semantic descriptors that subjects used to rate

their experience are shown on the right axis. VS, very strong; S, strong; M,

moderate; W, weak; BD, barely detectable.

FIGURE 3. Average intensity ratings of the perceived irritation from acetone and

PEA for acetone-exposed workers and non-exposed controls. The semantic

descriptors that subjects used to rate their experience are shown on the right axis.

VS, very strong; S, strong; M, moderate; W, weak; BD, barely detectable.


differed significantly as a function of compound (acetone orPEA), occupational exposure history, and item. The interac-tion between group and compound revealed that controlsubjects showed greater differences in their responses to thetwo compounds than did workers. Control subjects alsorated significantly more health symptoms following expo-sure to both compounds than did workers.

Responses to PEA were treated as an irritant baselinefor assessing the intensity of irritant effects from acetone;hence we compared ratings for the same symptoms follow-ing each type of exposure. For workers, the frequency andintensity of symptoms reported following chamber exposureto acetone did not differ from the levels reported followingexposure to PEA. By contrast, the control subjects experi-enced higher ratings of throat and nasal irritation andlightheadedness following exposure to acetone than follow-ing exposure to PEA.

The responses were also compared across each symp-tom for acetone and PEA, individually. Planned compari-sons of the reported symptoms from acetone by workers andcontrols revealed significant differences for six solvent-related effects but also for four somatic effects (itemsmarked ‘‘c’’ in Table IV). More significantly, workers ratedthe intensity of most symptoms from acetone exposure asbarely detectable, and only one item, nasal irritation, slightlyexceeded a rating of ‘‘weak,’’ whereas the control subjectsgave much higher ratings to both solvent-related and so-matic symptoms.

Although PEA is considered a pure olfactory stimulus,which is not irritating in any concentration, the medianratings of solvent-related symptoms from PEA were oftengreater than zero, particularly among the nonexposed.Workers and controls showed more similarity in their ratingsof symptoms following exposure to PEA: significant differ-ences were observed for only three items.

Predictive Relationships AmongIrritation, Odor, and Response Bias

From the foregoing results, it is clear that nonoccupation-ally exposed control subjects reported significantly moreodor and irritation from exposure to acetone than didworkers with repetitive occupational exposure. Yet, onpostexposure subjective symptom questionnaires, workersand controls reported higher levels of irritation to bothacetone and PEA than they did during exposure. Althoughboth types of assessment are subjective by nature, each typemay be differentially influenced by experimental demandcharacteristics. For example, response bias may influenceretrospective reports more than contemporaneous judg-ments. To identify the factors that can influence both on-lineand retrospective reports of sensory irritation, even at lowlevels of exposure, we used multiple regression analysis[Cohen and Cohen, 1983] to explore patterns of relation-ships among the sensory and cognitive components ofperceived irritation from a volatile compound. In particular,we hoped to identify whether factors that were unrelated tothe compound’s potential for sensory irritancy could nonethe-less have a significant influence on the degree of reportedirritation following exposure to that chemical.

Two factors appeared viable: (1) the odor from achemical, and (2) the subject’s willingness (bias) to reportirritation. The design of the present study allowed separateassessments of both factors from each subject. While in thechamber, subjects rated both the odor intensity from acetone,as well as its irritancy. If intensity of odor is mediating thereported intensity of irritation, we would expect to find thatodor (AC-ODOR) and irritation (AC-IRR) are significantlycorrelated. Also, if odor alone produces perceived irritation,we would expect to find reports of irritation during exposureto PEA (PEA-IRR).

TABLE III. Descriptive Statistics for Olfactory Detection Thresholds (ppm) for Acetone (AC) and Butanol (BU)

Group Threshold Median 25th percentile 75th percentile Mean SEMc

Workers AC Prea 86.10 10.00 388.00 362.00 157.40

AC Post 89.20 28.00 1853.00 1960.00 1147.10

BU Preb 1.82 0.01 8.60 18.26 10.55

BU Post 1.29 0.00 7.70 8.22 3.55

Controls AC Pre 83.90 8.43 382.30 395.30 131.23

AC Post 278.30 19.70 709.40 552.70 152.57

BU Pre 0.20 0.03 1.19 2.72 1.04

BU Post 0.32 0.02 0.61 5.31 2.94

aAcetone pre- or postexposure threshold.bButanol pre- or postexposure threshold.cStandard error of the mean.

564 Dalton et al.

Additional measures of response bias can be derivedfrom subjects’ ratings of somatic symptoms followingacetone exposure (AC-CON), and from their responses toboth solvent-related and somatic symptoms following expo-sure to PEA (PEA-CON). Of course, demographic factors,such as age and smoking history, as well as subject factors,like acetone exposure history within the worker population,could significantly contribute to the observed variance in thereported irritation to acetone. Accordingly, these factorswere also included in the analysis.

To predict the reported irritation from exposure toacetone for workers and controls, we hierarchically re-gressed average irritation for exposure to acetone onto the

TABLE IV. Summary of Health Symptom Questionnaire*

Acetone PEAa

Workers Controls Workers Controls

Mean SD Mean SD Mean SD Mean SD

Solvent-associated symptoms

Throat irritation 2.43c 0.48 8.59d 0.65 1.77 0.45 3.50 0.59

Eye irritation 3.77 0.64 4.63 0.61 1.70 0.49 2.75 0.51

Nasal irritation 6.74c 0.61 14.43d 0.51 3.08 0.52 4.08 0.60

Lightheadedness 2.07c 0.51 12.57d 0.62 1.21c 0.23 3.72 0.62

Headache 2.14c 0.52 5.09d 0.56 1.48c 0.36 2.88 0.53

Nausea 1.34c 0.38 5.17 0.64 1.63 0.43 2.33 0.42

Drowsiness 2.30c 0.47 5.64 0.69 2.22 0.56 4.18 0.65

Somatic (control) symptoms

Skin irritation 1.73 0.52 2.02 0.46 1.34 0.36 1.76 0.41

Bad taste 2.88 0.58 5.85d 0.60 1.57 0.40 2.59 0.53

Nasal congestion 2.26 0.54 4.04 0.61 1.65 0.39 3.16 0.57

Coughb 1.06c 0.10 3.59 0.63 1.37 0.34 1.94 0.44

Sneezeb 1.14 0.21 1.77 0.39 1.14 0.20 1.85 0.40

Stomachache 1.34c 0.33 2.71 0.55 1.43 0.38 1.96 0.47

Shortness of breath 1.28c 0.32 3.15 0.64 1.23 0.28 1.83 0.44

Heart palpitations 1.06c 0.10 2.37 0.52 1.14 0.18 1.71 0.40

Numbness/tingling 1.12 0.15 1.51 0.38 1.16 0.18 1.49 0.29

Ear ringing 1.02 0.04 1.66 0.43 1.03 0.07 1.79 0.44

Leg cramps 1.00 0.00 1.43 0.29 1.00 0.00 1.37 0.27

Back pain 1.02 0.04 1.51 0.37 1.05 0.08 1.85 0.41

Sweating 1.06 0.10 1.63 0.33 1.05 0.08 1.43 0.27

Itching 1.18 0.18 1.86d 0.45d 1.02 0.04 1.20 0.18

Current irritation 2.54c 0.54 5.17 0.63 1.82c 0.46 4.17 0.60

*Category boundaries of LMS: 0 5 No sensation; 1.37 5 Barely detectable; 5.46 5 Weak; 15.75 5 Moderate; 33.57 5 Strong; 50.47 5 Very strong;90.45 5 Strongest imaginable.aPhenylethyl alcohol.bNo subjects coughed or sneezed during exposure, hence these symptoms were treated as somatic.cResponses to items differed significantly (p , 0.001) between workers and controls.dSubjects’ responses to acetone differed significantly (p , 0.001) from responses to phenylethyl alcohol.

TABLE V. Summary Statistics for Health Questionnaire

Source df MSe F p

Group (workers vs. controls) 1, 52 3.099 13.9 0.000

Compound (acetone vs. PEA) 1, 52 0.279 37.4 0.000

Group 3 compound 1, 52 0.279 6.1 0.017

Symptom 21, 1092 0.168 25.2 0.000

Group 3 symptom 21, 1092 0.168 2.6 0.000

Compound 3 symptom 21, 1092 0.091 5.7 0.000

Group 3 compound 3 symptom 21, 1092 0.091 1.8 0.016


measures of odor and response bias (AC-ODOR and PEA-IRR), and the demographic/subject variables. The reportedirritation from acetone was well predicted by a linearcombination of the perceived odor of acetone (AC-ODOR)and the perceived irritation from PEA (PEA-IRR). Thecombination of these variables accounted for a significantamount of the variance (r2 5 0.80 and r2 5 0.89) for theworkers and controls, respectively; p, 0.001. None of thedemographic or subject variables predicted irritation fromacetone, and the addition of this set to the hierarchicalanalysis did not significantly improve the fit of the model.

To predict the questionnaire intensity ratings of solvent-related effects following exposure to acetone, we hierarchi-cally regressed those symptom ratings onto the somaticratings and measures of response bias (AC-CON, PEA-CON, PEA-IRR) and the demographic/subject variables.The rated intensity of solvent-related effects from acetonefor both groups was well predicted by a linear combinationof the three response bias variables (r2 5 0.51; r2 5 0.73; forworkers and controls, respectively; p,0.001). However, forworkers, the responses to PEA did not significantly increasethe fit of the model. As before, none of the demographic orsubject variables predicted the ratings of symptomaticeffects to acetone and when entered into the analysis did notsignificantly improve the model’s fit.

DISCUSSION

The present study investigated the degree of perceivedirritation and odor from 800 ppm acetone in two groups withdifferent histories of exposure to acetone—occupationallyexposed workers and nonoccupationally exposed controls.The results illustrate that the same concentration of acetonecan evoke very different degrees of irritation and healthsymptoms, depending on exposure history, perceived odorintensity, and the subject’s response bias for reportingirritation from a volatile chemical.

The most significant factor mediating the degree ofperceived irritation from acetone for both workers andcontrols appeared to be the perceived intensity of the odor(Fig. 4). Although it is possible that odor sensitivity andirritation sensitivity are both indices of an underlyinggeneral chemosensitivity trait that can vary among individu-als, two lines of evidence argue against such an interpreta-tion of the data. First, for most volatile compounds, odordetection occurs at concentrations well below irritationdetection [Cometto-Mun˜iz and Cain, 1990]. Second, instudies that obtained objective measures of odor and irrita-tion sensitivity or subjective measures of odor and irritationintensity, the responses to these two attributes have beenlargely independent [Hudnell et al., 1992; Wysocki et al.,1996].

Occupational exposure to acetone appears to produce avery specific change in the perceived intensity of acetone butapparently does not impair the perception of odor orirritation more generally, as shown by the similarity in therated intensity of PEAby the two groups.A concurrent studyon many of these same individuals revealed that detectionand irritation thresholds for another substance, 1-butanol,did not differ as a function of exposure to acetone [Wysockiet al., 1996]. Although exposure to some organic solventshas been correlated with more general olfactory impairments[Ahlstrom et al., 1986; Schwartz et al., 1989], no such effectwas found from long-term exposure to acetone. The lack ofgeneral olfactory dysfunction following occupational expo-sure to acetone is consistent with the failure to observe grossmorphological changes in the olfactory epithelium follow-ing chronic acetone exposure [Morgott, 1993].

Also consistent with acetone’s toxicological profile wasthe failure to find differences in the perceived odor, irritationor threshold sensitivity as a function of exposure amongworkers. The lack of a correlation could have occurredbecause most of the workers we tested had very similarlevels of acetone exposure in their current occupation. Yet,cumulative exposure to acetone differed markedly amongworkers and this factor was also unrelated to any measure ofodor or irritation from acetone. Thus, it remains unresolvedwhether the workers’ reduced perception of odor andirritation from acetone represents an acute or chronicadaptive change.

FIGURE 4. Correlation between perceived intensity of odor and perceived

intensity of irritation from exposure to 800 ppm acetone for both workers and

controls.

566 Dalton et al.

How do the results of the present study conform withthe reported irritant effects from acetone in the literature?Some investigators have reported finding irritation anddiscomfort from exposure to acetone at levels below thoseused in the present study [Cain et al., 1992; Dick et al., 1988,1989; Kiesswetter et al., 1994; Matsushita et al., 1969a,b;Seeber et al., 1992]. Upon review of the design of theprevious studies, however, it is difficult to conclude thatexperimental or occupational exposure to acetone was theprimary determinant of reported irritation. In several studies,investigators assayed the reactions of individuals exposed toacetone without a concurrent control exposure [Cain et al.,1992; Matsushita et al., 1969a,b]. When a control conditionwas used, subjects were exposed to clean air [Kiesswetter etal., 1994; Seeber et al., 1992], which may well have eliciteda different criterion for judging irritation than when exposedto an odorous substance. In studies that examined irritationfrom acetone in occupational exposures, the degree to whichacetone exposure was responsible for irritation or healthcomplaints among exposed workers cannot be determinedbecause exposure to acetone was confounded with the typeof job the worker performed [Kiesswetter et al., 1994;Raleigh and McGee, 1972]. Moreover, studies that haveused questionnaires to assess the health effects from acetonehave often failed to include or analyze control items in orderto permit assessment of the subject’s response bias [Cain etal., 1992; Dick et al., 1988; Matsushita et al., 1969a,b].

The results of the present study clearly demonstrate thatthe perception of irritation from acetone (as well as PEA) istightly correlated with the perception of the odor, and thatresponse bias plays a large role in the evaluation ofperceived irritation from acetone. Increases in concentrationproduce increases in the perceived odor intensity. If odor ismediating the irritancy response, then the reports of strongirritation among naive controls from the exposure concentra-tion is not surprising. However, the responses to thenonirritant control (PEA) strongly suggest these ratings arenot pure measures of irritation independent of odor.

Recently, an alternative approach has been developed toobjectively measure the concentrations at which compoundsproduce sensory irritation [Wysocki et al., 1992]. Thismethod takes advantage of the relationship between irrita-tion and skin localization: inhaled compounds at concentra-tions sufficient to stimulate the trigeminal nerve, whichmediates intranasal (as well as oral and ocular) irritation, canalso be localized to the stimulated nostril. This method hasrecently been employed to objectively determine that theintra-nasal irritancy threshold for acetone occurs well abovecurrent workplace exposure limits [Wysocki et al., 1996].The magnitude of the difference between the objectiveirritancy threshold and the subjective reported irritation inthis study emphasizes the possibility that subjective ratingsof irritation are largely influenced by odor and bias.

Importance of Measuring Bias in Studiesof Perceived Irritation

Response bias appears to play a significant role in thereports and evaluation of sensory irritation, as previousinvestigators have suggested [Pennebaker and Skelton,1981; Skov et al., 1992; Winneke, 1992] and may beendemic to human perception. For example, the informationan individual is given about the consequences of exposure toa chemical can determine the rated odor intensity duringexposure [Dalton, 1996]. Under identical exposure condi-tions, individuals who were told they were being exposed toa potentially harmful chemical rated the odor as being moreintense than did individuals who were told they are beingexposed to a healthy essential oil. Both the judged odor andirritation intensity from 800 ppm acetone depends on thetype of bias (positive, negative, or neutral) the subject hasbeen given about the consequences of exposure [Dalton etal., 1996]. It is conceivable that the cognitive expectationsconcerning exposure to acetone differed dramatically be-tween the two subject populations in this study, thusaugmenting the discrepancy in perceived odor and irritation.The present study helps refine the methodology for studyingchemical irritation by measuring the bias factors that canunderlie subjective irritation.

It is noteworthy that PEA elicited almost as muchirritation as did acetone among the workers, even though theconcentrations of the compounds differed during exposure(800 ppm vs. 200 ppm for acetone and PEA, respectively). Itis likely that exposure to higher concentrations of PEAwould have evoked greater irritation, particularly for thenonexposed controls. That is, acetone may have producedhigher ratings of irritation than did PEA simply becausetheodor of acetonewas perceived as much stronger (at least bythe nonexposed controls). When the odors of acetone andPEA were perceptually equivalent, as they were for theacetone-exposed workers (Fig. 2), exposure to acetoneelicited no more irritation than did exposure to PEA (Fig. 3).Owing to the low volatility of PEA (as compared to acetone)achieving chamber concentrations of PEA to match theperceived intensity of the odor of 800 ppm acetone was notfeasible with our present odor delivery system. Given therelationship between perceived odor and irritation in cham-ber studies, however, future studies employing a nonirritantodor control should attempt to perceptually equate the odorintensity for both compounds. This may have to be done onan individual basis, however, because subjects differ widelyin their choices of perceptual matches [Pierce et al., 1996].

The vast discrepancy between exposed workers andnonexposed controls in their reported irritation from acetoneilluminates the need for further study of the development ofadaptation to the odor and the irritation from a volatile


chemical. Reduced sensitivity and intensity of an odor(adaptation) following chronic exposure has long beenanecdotally reported. Recently, a persistent, but reversible,decrease in the sensitivity and perceived intensity of an odorhas been experimentally induced following a mere two-week daily exposure [Dalton and Wysocki, 1996]. If theirritancy response to acetone derives wholly or in part fromthe response to its odor, individuals who are temporarilyadapted to the odor will report less irritation.

The discrepancy also raises an important issue inmaking recommendations for workplace exposure limits,namely, are nonexposed control subjects a suitable popula-tion for assaying the irritation from volatile chemicals? Onthe basis of the current study, it appears that two importantinfluences on perceived irritation are more likely to exist athigher levels in a nonexposed population: the perceived odorintensity of a chemical and the bias to report irritationfollowing exposure to an unknown chemical. Thus, thevalue of using naive controls to measure the sensoryirritancy of acetone is questionable, when no assessment ofthese factors are included. It should be acknowledged thatexposed individuals may also vary on these dimensions.Perceived odor intensity can vary due to individual differ-ences in sensitivity, or duration or frequency of exposure tothe chemical. Bias to report irritation among exposedworkers can stem from dissatisfaction with nonchemicalaspects of their job, such as personality and psychosocialenvironment [Pennebaker and Skelton, 1981; Skov et al.,1992].

CONCLUSIONS

The determination of the sensory irritancy of volatilechemicals, such as acetone, has traditionally relied onhuman assessments of perceived irritation during exposuresor animal studies of respiratory reflexes during inhalationthat are taken as markers of sensory irritation. Often thereare large discrepancies between the values derived fromeach type of investigation. The results of the present studyshow that the perceived irritation elicited by exposure to 800ppm acetone varies with the individual’s history of exposureto acetone. However, two other factors that co-vary withexposure history were far more significant predictors of therated irritancy of acetone: the perceived odor of acetone andthe bias to report irritation during exposure to an odorouschemical. The perception of odor and bias can vary betweenand within groups who are occupationally exposed tochemicals and those who are not, hence, the influence ofthese factors on perceived irritation and symptom reports inexperiments and in the workplace merits examination.

ACKNOWLEDGMENTS

This work was supported by grants from the ChemicalManufacturer’s Association and the National Institutes of

Health (F32 DC-00197, R01 DC-00298, P50 DC00214). Aportion of these data was presented at the American Indus-trial Hygiene Conference and Exposition, May 1996, Wash-ington, DC. The authors thank Drs. David Morgott and J.W.Whitlow, Jr., for many valuable discussions and commentson the manuscript.

REFERENCES

Agency for Toxic Substances and Disease Registry (1994): ‘‘ToxicologicalProfile for Acetone.’’ Atlanta, GA: Agency for Toxic Substances andDisease Registry.

Ahlstrom R, Berglund B, Berglund U, Lindvall T, Wennberg A (1986):Impaired odor perception in tank cleaners. Scand J Work Environ Health12:574–581.

Alarie Y (1966): Irritating properties of airborne materials to the upperrespiratory tract. Arch Environ Health 13:433–449.

American Conference of Governmental Industrial Hygienists (ed) (1993):‘‘Threshold Limit Values for Chemical Substances and Physical Agents andBiological Exposure Indices’’. 1993–1994. Cincinnati, OH: ACGIH.

Baelum J, Andersen I, Molhave L (1982): Acute and subacute symptomsamong workers in the printing industry. Br J Ind Med 39:70–75.

Bos PMJ, Zwart A, Reuzel PGJ, Bragt PC (1992): Evaluation of the sensoryirritation test for the assessment of occupational health risk. Crit RevToxicol 21:423–450.

Cain WS (1970): Odor intensity after self-adaptation and cross–adaptation.Percept Psychophys 7:271–275.

Cain WS, See LC, Tosun T (1986): Irritation and odor from formaldehyde:Chamber studies. ASHRAE Proc 14:126–137.

Cain WS, Cometto-Mun˜iz JE, de Wijk RA (1992): Techniques in thequantitative study of human olfaction. In Serby MJ, Chobor KL (eds):‘‘Science of Olfaction.’’ New York: Springer-Verlag, pp 279–308.

Chang JCF, Barrow CS (1984): Sensory irritation tolerance and cross–tolerance in F-344 rats exposed to chlorine or formaldehyde gas. ToxicolAppl Pharmacol 76:319–327.

Cohen J, Cohen P (1983): ‘‘Applied Multiple Regression/CorrelationAnalysis for the Behavioral Sciences.’’ 2nd Ed. Hillsdale, NJ: LawrenceErlbaumAssociates.

Cometto–Mun˜iz JE, Cain WS (1990): Thresholds for odor and nasalpungency. Physiol Behav 48:719–725.

Cometto–Mun˜iz JE, Cain WS (1995): Olfactory Adaptation. In Doty RL(ed): ‘‘Handbook of Olfaction and Gustation.’’ New York: Marcel Dekker,pp 257–281.

Dalton P (1996): Odor perception and beliefs about risk. Chem Senses21:447–458.

Dalton P, Wysocki CJ (1996): The nature and duration of adaptationfollowing long–term exposure to odors. Percept Psychophys58(5):781–792.

Dalton P, Wysocki CJ, Brody MJ, Lawley HJ (1996): The influence ofcognitive bias on the perceived odor, irritation and health symptoms fromchemical exposure. Int Arch Occup Environ Health (in press).

Dick RB, Brown WD,Setzer JV, Taylor BJ, Shukla R (1988): Effects of shortduration exposures to acetone andmethyl ethyl ketone. Toxicol Lett 43:31–49.

Dick RB, Setzer JV, Taylor BJ, Shukla R (1989): Neurobehavioural effectsof short duration exposures to acetone and methyl ethyl ketone. Br J IndMed 46:111–121.

568 Dalton et al.

Doty RL, Brugger WE, Jurs PC, Orndorff MA, Snyder PJ, Lowry LD(1978): Intranasal trigeminal stimulation from odorous volatiles: Psychomet-ric responses from anosmic and normal humans. Physiol Behav 20:175–185.

Elofsson S, Gamberale F, Hindmarsh T, Iregren A, Isaksson A, Johnsson I,Knare B, Lydahl E, Mindus P, Persson H, Philipson B, Steby M, Struwe G,Soderman E, Wennberg A, Widen L (1980): Exposure to organic solvents.Scand JWork Environ Health 6:239–273.

Green BG, Shaffer GS, Gilmore M (1993): Derivation and evaluation of asemantic scale for oral sensation magnitude with apparent ratio properties.Chem Senses 18:683–702.

Green BG, Dalton P, Cowart B, Shaffer G, Rankin KR, Higgins J (1996):Evaluating the ‘‘Labeled Magnitude Scale’’ for measuring sensations oftaste and smell. Chem Senses 21:323–334.

Hudnell HK, Otto DA, House DE, Molhave L (1992): Exposure of humansto a volatile organic mixture. II. Sensory Arch Environ Health 47:31–38.

Husman K (1980): Symptoms of car painters with long-term exposure to amixture of organic solvents. Scand JWork Environ Health 6:19–32.

Iregren A (1986): Subjective and objective signs of organic solvent toxicityamong occupationally exposed workers. Scand J Work Environ Health12:469–475.

Jones MH (1954): A study of the common chemical sense. Am J Psychol67:696–699.

Kane LE, Dombroske BS, Alarie Y (1980): Evaluation of sensory irritationfrom some common industrial solvents. Am Ind HygAssoc J 41:451–455.

Kiesswetter E, Blaszkewicz M, Vangala RR, Seeber A (1994): Acuteexposure to acetone in a factory and ratings of well-being. Neurotoxicol15:597–601.

Kobal G (1982): A new method for determination of the olfactory and thetrigeminal nerve’s dysfunction: Olfactory (OEP): and chemical somatosen-sory (CSEP) evoked potentials. In Rothenberger A (ed): ‘‘Event–RelatedPotentials in Children.’’ Amsterdam: Elsevier Biomedical Press, pp 455–461.

Matsushita T, Goshima E, Miyagaki H, Katsuyoshi M, Takeuchi Y, Inoue T(1969a): Experimental studies for determining the MAC value of acetone:Biological reactions in the ‘‘six-day exposure’’ to acetone. Jpn J Ind Health11:507–515.

Matsushita T, Yoshumune A, Inoue T, Yamada S, Suzuki H (1969b):Experimental studies for determining the MAC value of acetone. 1.Biologic reactions in the one-day exposure to acetone. Jpn J Ind Health11:477–485.

Mechanic D (1972): Social psychological factors affecting the presentationof bodily complaints. N Engl J Med 82:213–225.

Morgott DA (1993): Acetone. In Clayton GD, Clayton FE (eds): ‘‘Patty’sIndustrial Hygiene and Toxicology.’’ New York: John Wiley & Sons, pp149–281.

Nelson KW, Ege JFJ, Morwick Ross (1943): Sensory response to certainindustrial solvent vapors. J Ind Hyg Toxicol 25:282–285.

OSHA (1989): ‘‘Air contaminants; final rule.’’ Federal Register 54:2332–2983.

Otto DA, Hudnell HK, House DE, Molhave L, Counts W (1992): Exposureof humans to a volatile organic mixture. I. Behavioral assessment. ArchEnviron Health 47:23–30.

Pennebaker JW, Skelton JA (1981): Selective monitoring of physicalsensations. J Person Soc Psychol 41:213–223.

Pierce JD Jr, Wysocki CJ, Aronov EV, Webb JB, Boden RM (1996): Therole of perceptual and structural similarity in cross-adaptation. ChemSenses 21:223–237.

Raleigh RL, McGee BS (1972): Effects of short, high-concentrationexposures to acetone as determined by observation in the work area. JOccup Med 14:607–610.

Schwartz BS, Doty RL, Monroe C, Frye R, Barker S (1989): Olfactoryfunction in chemical workers exposed to acrylate and methacrylate vapors.Am J Public Health 79:613–618.

Seeber A, Kiesswetter RR, Vangala M, Blaszkewicz M, Golka K (1992):Combined exposure to organic solvents: An experimental approach usingacetone and ethyl acetate. Appl Psychol 41:281–292.

Skov P, Valbjorn O, Pedersen BV (1992): Influence of personal characteris-tics, job-related factors and psychosocial factors on the sick buildingsyndrome. Scand JWork Environ Health 15: 286–295.

Swets JA, Tanner WP, Birdsall TG (1961): Decision processes in percep-tion. Psych Rev 68:301–340.

Williams CH, Lees-Haley PR (1993): Perceived toxic exposure: A reviewof four cognitive influences on perception of illness. J Soc Behav Personal8:489–506.

Winneke G (1992): Structure and determinants of psychophysiologicalresponse to odorant/irritant air pollution. Ann NYAcad Sci 641:261–276.

Wysocki CJ, Green BG, Malia TP (1992): Monorhinal stimulation as amethod for differentiating between thresholds for odor and irritation. ChemSenses 17:722–723.

Wysocki CJ, Dalton P, Brody MJ, Lawley HJ (1996): Odor and irritationthresholds for acetone in acetone-exposed factory workers and (occupation-ally non-exposed) control subjects. Presented at the American IndustrialHygiene Conference Exposition, May 1996.


Perceived odor, irritation, and health symptoms following short-term exposure to acetone

Documents

Transcript of Perceived odor, irritation, and health symptoms following short-term exposure to acetone