C 2010 The American Laryngological, V

The Laryngoscope

Rhinological and Otological Society, Inc.

Effects of Smoking on Taste: Assessment with Contact Endoscopy and Taste Strips
Iordanis Konstantinidis, MD, PhD; Angelos Chatziavramidis, MD; Athanasia Printza, MD, PhD; Spyros Metaxas, MD, PhD; Jannis Constantinidis, MD, PhD

Objectives: This study aims to compare the taste function between smokers and nonsmokers with clinical testing, subjective ratings, and contact endoscopy of the tongue. Study Design: Cross-sectional survey. Methods: Data were collected from 38 smokers (mean age 37 years; 25 female, 23 male) and 34 nonsmokers (mean age 33.5 years; 18 female, 16 male). The parameters assessed were the number of fungiform papillae per square centimeter in a noncontact way and their morphology (surface, capillary vessels) by contact endoscopy. The morphology of the filiform papillae has also been assessed. In addition, clinical testing of gustatory function was performed by means of taste strips and subjective intensity ratings of natural taste stimuli. Results: No significant difference was found in clinical testing and intensity ratings between the two study groups. A trend toward significance was found in taste strip results for decreased bitter taste in heavy smokers (P .06). The number and the size of fungiform papillae did not significantly differ between the study groups. No sex-related differences were observed. Smokers exhibited significantly more keratin structures on the fungiform papillae surface, less tortuous capillary vessels, and a significant distortion of their filiform papillae. Conclusions: Taste function presents significant resistance to smoking, although changes in morphology of fungiform and filiform papillae have been observed especially in heavy smokers. Key Words: Contact endoscopy, taste, smoking, fungiform, filiform, papillae. Level of Evidence: 1b Laryngoscope, 120:19581963, 2010

INTRODUCTION
Smoking as a potential cause for taste dysfunction has not been extensively studied. As one of the most important health risk behaviors, smoking is under investigation for many systemic or local effects. Greece is among the countries with a significant tobacco production and a high prevalence of current smokers. Due to the high prevalence of smoking in Greece, any potential effect of smoking on taste function would contribute to a significant impairment of this sense in the general population. Smith first reported a correlation between the number of taste papillae and taste sensitivity a few decades ago, with a positive correlation between gustatory sensitivity and the number of stimulated fungiform papillae in a defined area after constant stimulus concentrations.1 Morphologic changes in fungiform papillae in addition to their decreased number have also been reported in patients with taste dysfunction after ear surgery.2 Finally, other studies reported a variety of effects of smoking on the capillary vessels.3 The aim of this study was to examine the effects of smoking on taste mainly at a peripheral level by examining with contact endoscopy three basic factors: the number of fungiform papillae, their morphology, and the morphology of their capillary vessels. These morphologic data were correlated with the clinical testing results and subjective ratings.

MATERIALS AND METHODS Subjects

The current study was conducted according to the guidelines of the Declaration of Helsinki on Biomedical Research Involving Human Subjects and was approved by the Ethic Committee of the Aristotle University, Thessaloniki, Greece. The population sample was randomly drawn from the outpatient clinic of the ORL Department. One hour prior to testing, subjects were asked not to eat or drink anything except water, not to smoke, and not to brush their teeth. Subjects with severe diseases that might affect taste perception (chronic renal failure, middle ear infections/operations, systemic use of drugs, etc.) were not included in the investigation. Ex-smokers were also excluded from the study. An age below 40 years was an inclusion criterion in order to avoid an age effect on taste function.4

From the Smell and Taste Clinic, 2nd Academic ORL Department, Papageorgiou Hospital, Aristotle University, Thessaloniki, Greece. Editors Note: This Manuscript was accepted for publication May 12, 2010. Send correspondence to Iordanis Konstantinidis, MD, PhD, ORLDept, Papageorgiou Hospital, Efkarpia, Ring road, 56403, Thessaloniki, Greece. E-mail: jordan_orl@hotmail.com DOI: 10.1002/lary.21098

Laryngoscope 120: October 2010

Konstantinidis et al.: Effects of Smoking on Taste

1958

Fig. 1. (A) Contact endoscopy of a 34-year-old nonsmoker male. Assessment requires open mouth and extended tongue. Then the endoscope is placed on the methylene-blue-stained surface of the tongue for visualization of the mucosa. (B) Noncontact endoscopic image of a square centimeter of the anterior tongue in the same subject (50). The fungiform papillae are clearly seen (white arrow) as numerous round, pale bumps in a dark background (filiform papillae).

In total, 38 smokers (20 females, 18 males, age range 2140 years, mean age 33.8) were assessed. The mean of smoking duration was 5.5 years (range 215 years). The control group consisted of 34 nonsmoker volunteers of similar age and gender (18 females, 16 males, age range 1940 years, mean age 31.9). The group of smokers was divided in two subgroups (heavy [n 12] and light smokers [n 26]) according to their cigarette consumption of more or less than 1 pack or 20 cigarettes/day, respectively.

Taste strips
Taste testing was performed by means of taste strips, a well-investigated clinical test.5,6 This test is based on filter paper strips impregnated with the four basic taste qualities: sweet, bitter, salty, and sour. Umami is not included in the test because Europeans identify this taste quality with a variable terminology. One hour before testing, subjects were asked not to eat or drink anything except water. The 16 taste strips were presented randomized regarding taste qualities in increasing concentrations, as a whole-mouth procedure in the middle of the anterior portion of the tongue. In addition, two blank strips were presented at the end of the test. Subjects were asked to identify the taste quality choosing one of five possible answers on a form (sweet, sour, salty, bitter, no taste). Before assessment of each taste strip, the mouth was rinsed with water. The taste score was the number of correctly identified taste strips. The following concentrations were used for the taste strips: sweet: 0.4, 0.2, 0.1, 0.05 g/ml sucrose; sour: 0.3, 0.165, 0.09, 0.05 g/ml citric acid; salty: 0.25, 0.1, 0.04, 0.016 g/ml sodium chloride; bitter: 0.006, 0.0024, 0.0009, 0.0004 g/ml quininehydrochloride. The score range is from 0 to 18, and scores higher than 8 are considered as within the normal range.

in a sitting position. Methylene-blue was used to stain the anterior tongue epithelia (Blue di metilene 50 mg/5 ml, Monico SPA, Venice, Italy). A filter paper strip delineating a tongue area of 1 cm2 was placed in a median position and approximately 1.5 cm posterior from the edge of the tongue for quantification of the number of fungiform papillae (Fig. 1A). Identification of fungiform papillae was performed first in a noncontact way (Fig. 1B). All subjects kept their mouth open and tongue extended for approximately 2 minutes, and the contact endoscope was placed on the methylene-blue-stained surface of the tongue for visualization of the mucosa. The digital images obtained were evaluated in an IBM-compatible working station for the number of fungiform papillae, the morphology of fungiform and filiform papillae, and general morphology and density of the subepithelial capillary vessels of the fungiform papillae. Subepithelial vessels according to their tortuosity were divided in five types with a classification used by Negoro et al.7 One investigator performed all the endoscopies and another investigator blinded to recordings performed the data analyses.

Subjective Ratings
Intensity ratings were performed with application of a drop (approximately 20 lL) of liquid tastant solutions (suprathreshold stimulations of sweet: sucrose, 10 g/ml; sour: citric acid, 2.5 g/ml; salty: sodium chloride, 5 g/ml; bitter: quinine hydrochloride, 0.05 g/ml) onto the extended tongue. The mouth was rinsed with water before each taste solution application. The subjects were asked to identify the quality of taste (sweet, sour, salty, bitter, or water) and to rate the intensity of the stimulation with a use of a visual analogue scale ranging from 0 (no taste) to 100 (very strong intensity).

Statistical Analysis Contact Endoscopy

A detachable camera (Image1TM, Karl Storz Company, Tuttlingen, Germany) was fitted to a contact endoscope (Karl Storz 7215 AA, 08, length: 23 cm, diameter: 4 mm, 30 , Karl Storz Company), and all findings were recorded and saved as digital data. Contact endoscopy was performed on the subjects Descriptive statistics are presented as means and standard errors of the mean (SE). An analysis of variance for repeated measures (rm-ANOVA) was used with subject factors sex and age. Group comparisons between smokers and nonsmokers were performed with the Student t test for independent samples, with regard to parametric data, and with

Laryngoscope 120: October 2010

Konstantinidis et al.: Effects of Smoking on Taste

1959

Fig. 2. (A) Contact endoscopy of fungiform papilla with a tortuous capillary tree (150). (B) Fungiform papilla with surface keratin structures (black arrows) and dotted granular-shape capillary vessels (150). the Mann-Whitney rank test, with regard to nonparametric data. Bonferroni tests were applied for post hoc comparisons. Pearson statistics were used for correlation analyses. The alpha level was set at 0.05 and SPSS 12.0 (SPSS, Inc., Chicago, IL) was used for statistical analyses.

RESULTS
Regarding factors affecting taste within the two groups, there was no main effect of the factor sex (P .25) or age (P .36) as tested with taste strips. In addition, taste strip results showed no significant differences between the study groups. Specifically, the mean of total taste strips score was 12.1 6 1.8 for the smokers and 12.6 6 2.2 for the nonsmokers. Subgroup analysis did not reveal any other significant difference. The percentage of correctly identified taste strips regarding each taste quality separately did not reveal significant differences in all concentrations with one exception. In the lowest concentration of bitter, the identification percentage was 52% for the nonsmokers and

31.1% for the smokers (P > .001). Subgroup analyses showed that this difference could be attributed to the subgroup of heavy smokers (identification percentage: 25%) rather than the subgroup of light smokers (identification percentage: 36.1%) (P .004). Regarding subjective ratings, no significant difference was seen in relation to age (P .31) and sex (P .065), although there was a trend for women to rate stimulations higher than men. No significant differences were found between the two study groups regarding intensity ratings for all taste qualities. Regarding contact endoscopy findings, the density of fungiform papillae in total had a 2.6-fold range and was similar in both groups (smokers 23.5 6 6.1 vs. nonsmokers 27.1 6 5.8). Correlation analyses indicated that fungiform papillae density in both groups was not affected by age (r32 0.21, P 0.3, r35 0.18, P .7) or sex (r32 0.31, P .08, r35 0.26, P .35). Changes in their morphology such as keratin structures on their surface (Fig. 2) after Bonferroni correction were

Fig. 3. Contact endoscopy of filiform papilla (A) in a nonsmoker and (B) in a heavy smoker (150). Note the significant distortion of the hairtype compartment in filiform papillae of Figure 3B.

Laryngoscope 120: October 2010

Konstantinidis et al.: Effects of Smoking on Taste

1960

TABLE I. Comparison of Morphologic Changes on Shape of Fungiform and Filiform Papillae.

Nonsmokers (n 34) Smokers (n 38) P Value

Surface structures on fungiform papillae

4 subjects/10.4%

22 subjects/57.5% Light smokers (n 26) 15 subjects/57.7% Heavy smokers (n 12) 7 subjects/58.3%

> .001 .7

Distorted filiform papillae

2 subjects/5.2%

12 subjects/31.7% Light smokers (n 26) 7 subjects/26% Heavy smokers (n 12) 5 subjects/41.6%

> .001 .003

Data are presented as number of subjects and percentage of the group or subgroup. P values correspond to differences between percentage incidences.

significantly higher in the group of smokers (22 subjects/ 57.5%) than those in nonsmokers (4 subjects/10.4%) (P > .001). A similar difference was seen in distorted filiform papillae (Fig. 3) with increased incidence in smokers compared with nonsmokers (12 subjects/31.7% vs. 2 subjects/5.2%) (P .002). Distorted filiform papillae were identified more frequently in heavy smokers. All these results are analytically presented in Table I. The two study groups showed significant differences in capillary vessels morphology with nonsmokers having distinguished tortuosity of their vessels. Data from fungiform papillae capillary vessels are presented in Table II. Correlational analyses between morphologic and functional measurements in both groups did not reveal significant differences (all P values > .2).

DISCUSSION
This is a study assessing the effects of smoking on taste at both the clinical and morphologic levels. Two major findings arose from this study. First, smoking affects the morphology of fungiform papillae in terms of capillary vessels and surface along with distortion of filiform papillae. Second, these morphologic changes do not affect the clinical testing results and subjective intensity ratings.

There is evidence that the number of fungiform papillae is related to gustatory sensitivity.1 However, the number of fungiform papillae varies in healthy subjects. In our study, there was a 2.6-fold range in their number regardless of smoking habit. This is in accordance with other authors reporting a similar variety in fungiform papillae density.8,9 Considering the even higher variability in the number of taste pores in healthy subjects10 and the difficulties of contact endoscopy to identify them,8,11 our study was based only in measurements of fungiform papillae. According to our results, the density of fungiform papillae is not affected by smoking as seen in other pathologies like chorda tympani transection in ear surgery.2 Experimental studies showed significant decrease of the taste buds in gerbils, rats, and hamsters within 3 weeks after chorda tympani/lingual nerve transection.12,13 The fact that the number of fungiform papillae is strongly affected by the neural status may indicate that smoking does not significantly affect the innervation of the papillae. The presence of keratin structures on the fungiform papillae surface can be interpreted as a sign of atrophy.14 Studies showed that the atrophy process results in keratinization similar to that of filiform papillae.15 This fact makes the identification of fungiform papillae difficult, and occasionally their distinction from filiform

TABLE II. Differences Between Smokers and Nonsmokers in Fungiform Capillary Vessels Morphology.
Classification of Capillary Vessels Nonsmokers Smokers Significance

Type A

Clear loop, wooden branch shape

59%

40%

P > .001 P > .001 P .004

Type B Type C

Unclear loop, wooden branch shape Elongated vessels without loops

36% 5%

42.5% 15%

Type D Type E

Dotted, granular shape Not seen

0% 0%

2.5% 0%

Laryngoscope 120: October 2010

Konstantinidis et al.: Effects of Smoking on Taste

1961

papillae is impossible.14 Our data suggest that smoking does not result in major changes in fungiform papillae even in heavy smokers regarding the number of identified papillae. However, signs of atrophy were more likely to be found in smokers than in nonsmokers, indicating a degeneration process. The resistance of taste system to smoking can be partially explained by the complex gustatory innervation of the oral cavity by three cranial nerves: the facial, glossopharyngeal, and vagal nerves. In addition, the fact that lateralized damage of the chorda tympani goes frequently unnoticed16 might explain the small number of individuals with taste dysfunction seen in smell and taste clinics. Several authors suggested that structural maintenance of fungiform papillae requires the normal function of chorda tympani or lingual nerve.13,14 No significant differences in the number of fungiform papillae have been demonstrated in relation to age. A reason for this may be the wide variability in their number. Although it may be hypothesized that fungiform papillae density decreases across life span, this correlation cannot be supported by our results. Regarding the capillary vessels, our data show that smoking habit does induce significant morphologic changes in the microcirculation of the fungiform papillae and that these changes are presented as a less visible capillary network with decreased loops and treeshape endings. The poorer microcirculation as noted in contact endoscopy may be a result of a decreased caliber of capillary loops, found in smokers by other authors. Lova et al. supported distorted microcirculation of smokers in a study, where the heavy smoker capillaroscopic pattern was often aneurysmatic distorted capillary loops.3 In general, there is evidence that acute cigarette smoking significantly impairs human cutaneous1 and coronary6 microcirculation. Additionally, chronic tobacco use increases postcapillary venous pressure but not precapillary arterial pressure in the mesenteric microcirculation of the rat7 and thus results in a decreased capillary flow. Experimental studies showed nicotine induced abnormal pattern of blood vessel formation with effects on vascular endothelial cells17 and extracellular matrix deposition.18 This fact is possibly related to the neoangiogenetic effect of nicotine. Another interesting finding of our study was the morphologic changes observed on the filiform papillae. This is in accordance with studies on other conditions such as the black hairy tongue, where smoking was found to be a pathogenic factor implicated in morphologic changes of the hair-type compartment of filiform papillae.19 Although filiform papillae are not considered to participate in taste function because they lack taste buds, they might indirectly influence taste input.20 Their presence is necessary to have a mechanically rough surface on the tongue, which is helpful for food handling and somatosensory function. Thus, morphologic changes of filiform papillae should be recorded in contact endoscopy tongue assessment for taste impairment. In a recent study, Vennemann et al. stated that smoking has a well-known negative impact on olfactory Laryngoscope 120: October 2010

function; however, taste function presents significant resistance.21 This observation may indicate that smoking effects probably have different thresholds in these senses. Our results are in agreement with this study as the observed morphologic changes do not reflect in clinical measurements. The only significant difference seen in taste strip results was that of a decreased threshold in bitter taste. However, this is probably not an effect of smoking but a genetic variation. Bartoshuk and colleagues linked the ability to taste a bitter compound called propylthiouracil with food preference, suggesting that 25% of the population are nontasters of it.22 This might partially explain why some individuals feel the bitter compounds of smoke less intensely and are more likely to start smoking. In our study, we did not assess former smokers or the duration of smoking as there is evidence that the effects of smoking on gustatory function are rather direct and of short term.20 Thus, in contrast to other negative outcomes such as coronary artery disease or ischemic stroke, the cessation of smoking would have an immediate effect on taste function. Limitations of our study should be noted. The use of taste strips as a whole-mouth procedure and the study population made the detection of taste impairment more difficult. However, in a larger population study with 1,312 participants, taste impairment was significantly related only to heavy smoking and not to lower cigarette consumption.21 The same authors explained that the effect of smoking on taste function for all smokers presented more intensely when additional independent variables existed, such as diabetes, myocardial infarction, and head trauma. Considering the age and the good general condition of subjects assessed in our study, we can explain the difficulty in finding significant differences between the study groups.

CONCLUSIONS
Smoking appears to significantly affect morphology and structure of taste buds; however, taste function appears to have a stronger resistance to smokings effect. This information should not be considered as a fact that promotes smoking, as its deleterious effects on chemical senses are well known.

BIBLIOGRAPHY
1. Smith DV. Taste intensity as a function of area and concentration: differentiation between compounds. J Exp Psychol 1971;87:163171. 2. Just T, Pau H, Witt M, Hummel T. Contact endoscopic comparison of morphology of human fungiform papillae of healthy subjects and patients with transected chorda tympani nerve. Laryngoscope 2006;116:12161222. 3. Lova R, Miniati B, Macchi C, et al. Morphologic changes in the microcirculation induced by chronic smoking habit: a videocapillaroscopic study on the human labial mucosa. Am Heart J 2002;143:658. 4. Gudziol H, Hummel T. Normative values for the assessment of gustatory function using liquid tastants. Acta Oto-Laryngologica 2007;127:658661. 5. Mueller C, Kallert S, Renner B, Stiassny K, Temmel AF, Hummel T, Kobal G. Quantitative assessment of

Konstantinidis et al.: Effects of Smoking on Taste

1962

6.

7.

8. 9.

10. 11.

12.

13.

gustatory function in a clinical context using impregnated taste strips. Rhinology 2003;41:26. Landis B, Welge-Luessen A, Bramerson A, Bende M, Mueller C, Nordin S, Hummel T. Taste stripsa rapid, lateralized, gustatory bedside identification test based on impregnated filter papers. J Neurol 2009;256:242248. Negoro A, Umemoto M, Fukazawa K, Terada T, Sakagami M. Observation of tongue papillae by video microscopy and contact endoscopy to investigate their correlation with taste function. Auris Nasus Larynx 2004;31: 255259. Miller IJ, Reedy FE. Quantification of fungiform papillae and taste pores in living human subjects. Chem Senses 1990;15:281294. Just T, Gafumbegete E, Kleinschmidt EG, Pau HW. Kontaktendoskopische quantifizierung der fungiformen papillen: korrelation zum schmeckvermogen? Laryngorhinootologie 2003;82:501507. Miller IJ, Reedy FE. Variations in human taste bud density and taste intensity perception. Physiol Behav 1990;47: 12131219. Segovia C, Hutchinson I, Laing DG, Jinks AL. A quantitative study of fungiform papillae and taste pore density in adults and children. Brain Res Dev Brain Res 2002;138: 135146. Nakashima T, Toyoshima K, Shimamura A, Yamada N. Morphological changes of taste buds and fungiform papillae following long-term neurectomy. Brain Res 1990;533: 321323. Oakley B, Lawton A, Riddle DR, Wu LH. Morphometric and immunocytochemical assessment of fungiform taste buds

14. 15. 16. 17. 18.

19.

20. 21. 22.

after interruption of the chorda-lingual nerve. Microsc Res Tech 1993;26:187195. Nagato T, Matsumoto K, Tanioka H, Kodama J, Toh H. Effect of denervation on morphogenesis of the rat fungiform papilla. Acta Anat (Basel) 1995;153:301309. Oakley B, Wu LH, Lawton A, de Sibour C. Neural control of ectopic filiform spines in adult tongue. Neuroscience 1990;36: 831838. Landis BN, Beutner D, Frasnelli J, Huttenbrink KB, Hummel T. Gustatory function in chronic inflammatory middle ear diseases. Laryngoscope 2005;115:11241127. Villablanca AC. Nicotine stimulates DNA synthesis and proliferation in vascular endothelial cells in vitro. J Appl Physiol 1998;84:20892098. Melkonian G, Le C, Zheng W, Talbot P, Martins-Green M. Normal patterns of angiogenesis and extracellular matrix deposition in chick chorioallantoic membranes are disrupted by mainstream and sidestream cigarette smoking. Toxicol Appl Pharmacol 2000;163:2637. Manabe M, Lim HW, Winzer M, Loomis CA. Architectural organization of filiform papillae in normal and black hairy tongue epithelium. Arch Dermatol 1999;135: 177181. Konstantinidis I. The taste peripheral system. B-ENT 2009; Suppl 13:115121. Vennemann MM, Hummel T, Berger K. The association between smoking and smell and taste impairment in the general population. J Neurol 2008;255:11211126. Bartoshuk LM, Duffy VB, Miller IJ. PTC/PROP tasting: anatomy, psychophysics, and sex effects. Physiol Behav 1994;56:11651171.

Laryngoscope 120: October 2010

Konstantinidis et al.: Effects of Smoking on Taste

1963