INTRODUCTION

Among the various negative effects of smoking, there seem to be a classification of importance regarding health promotion and diseases prevention according to morbidity and mortality. Cardiovascular and oncological effects seem to be the most highlighted ones15. Studies of the effects of smoking on perception of taste and odor do not seem to raise the same interest in researchers, especially because symptomatology is relatively mild and does not generate public health problems. Similarly, few professions are directly dependent on taste and odor perceptions, such as wine specialists or perfume makers.

Investigation methods applied in different studies differ a lot and do not complement each other, resulting in lack of minimum standards. Variables such as use of both animals and humans as subjects, no use of sex matching groups4,18, small samples1,10,15, different formats (daily sessions or not), routes of administration (subcutaneous or inhaled), different perceived substances and use of smoking3,4,14,15 or only of some of its components (such as nicotine, for example) impair the development of minimal standardization1,5,8,10,13,16,17.

There are no studies that address concomitantly the effects on both sensorial systems, and there is also a predominance of studies that investigated taste perception. Therefore, we decided to conduct a review of the literature about consequences of smoking on each of the sensorial system.

LITERATURE REVIEW

a) Effects of smoking on taste perception

When evaluating taste perception, the main factors under study tend to be threshold for perception of stimuli, characterization of each of the, four main taste perceptions (acid - sour, sweet, salty and Bitter)11 and global hedonic impression of stimuli.

In studies with animals, distinction of taste perceptions is rarely approached. However, Parker et al.16 investigated the effect of nicotine (subcutaneous via) in the modification of taste for a tasty solution (sucrose) and a non-tasty - bitter - solution (quinine) in mice. Behavioral reactions of animals were recorded and analyzed, suggesting that nicotine suppressed adverse reactions to both types of solution, that is to say, it impaired (or inhibited) sensorial perception. In the same study, another significant finding was the increase in pleasant reaction with the solution of sucrose 21 days after discontinuation of nicotine. In order to characterize conditioned taste aversion produced by nicotine in mice, Kumar et al.13 proposed that nicotine was a powerful producer of taste aversion, clearly dose-related and showing central effect and association with posttrematic area (emetic effect).

In humans, studies showed a discordance of findings. As to determining perception thresholds, in 1961 Krut et al.12 reported that the threshold for bitter taste was higher in smokers, if compared to non-smokers. This effect would be related to quantity and duration of smoking, and not only to one cigarette. In a 18-month cohort study with 4 follow-up series comparing smokers and non-smokers, Peterson et al.18 identified an increased threshold for bitter taste in smokers compared to controls; however, there was no statistically significant difference among the three basic tastes (sweet, salty and acid). Nevertheless, the studied groups were not matched according to sex and age. Coats3 investigated concomitant effect of three variables on electrical threshold of taste perception using a taste electrogram (EGM). He believed that EGM values would depend on age, sex and smoking if the stimulus was the same for all subjects. Approximately 250 subjects were assessed. Women had lower thresholds than men, regardless of age and smoking. The effect of age alone did not show significance but the male group older than 60 years showed increased taste perception threshold. In women, this deterioration of taste perception as a result of aging was not noticed, at least not at a significant level.

Comparing smokers (n=8) and non-smokers (n=9) matched according to health status and demographic characteristics, Mela15 observed in smokers who had not smoked for 12 hours that there was a 2 to 4 fold increase in recognition threshold for sweet taste (sucrose), salty taste (sodium chloride) and bitter. taste (caffeine) - employed as solutions - than for non-smokers. Five minutes after smoking, in both groups, the results were the following: in the analysis of the group, there was no difference in recognition threshold in smokers and nonsmokers, all of them had difficulty to perceive all solutions - only caffeine produced a statistically significant difference. Among the groups, there was no statistically significant difference. Based on these data, one may suggest the development of tolerance in chronic users, since acute exposure did not produce a modified perception in the group. Moreover, smokers showed a reduction of taste perception.

Since these alterations of taste perception were objectively demonstrated, the influence of smoking on food preference of smokers is now finally understood. Smoking is associated with lower body weight average in smokers than in general non-smoker population, matched according to age, sex and height'. Weight gain after smoking cessation is one of the most frequent effects observed in former smokers20. The relation has been clearly stated; however, the reasons why are not so. Grunberg8 conducted two different studies: one in mice and another in humans. In the first one, there were four groups: one control group and three other groups submitted to increasing doses of nicotine. Exposed animals decreased total intake of calories, reducing significantly the intake of sweet solution, but not the total amount of food eaten during exposure period. When nicotine administration was interrupted, there was an increase in intake of calories when compared to control group. The second part of the study involved humans divided into three groups: smokers, smokers temporarily not smoking and non-smokers. Different kinds of foods (sweet, salty and bland) were offered to subjects. The measurement of consumption was conducted based on consumed weight. Similarly to the results obtained from the animal study, the intake of sweet foods was reduced in the group of smokers and temporarily non-smokers when compared to non-smokers. The consumption of other food items remained unchanged. The authors considered that there were alterations in the consumption of sweet foods caused both by cigarette (in humans) and nicotine (in mice). Therefore, changes in body weight and use of cigarettes may be related to a modification of taste perception, leading to compromised preference for sweet foods and high caloric content.

Following the same study design, but including perception of fat taste and not only concentration of sucrose, Perkins et al.17 administered either inhaled nicotine or placebo in 20 subjects - 10 smokers and 10 non-smokers - and presented them with random samples of milk that differed in its proportion of sucrose and fat, at two different sessions. Their findings were: acutely speaking, on one side, nicotine reduced the perception of fat taste in smokers and non-smokers, but had no influence on sweet taste and did not alter hedonic of food; on the other hand, chronically speaking, nicotine triggered modifications in the pleasure of eating, but did not modify perception of substances. The authors concluded that since there were no modifications between nicotine arid placebo in regards to taste perception, there would be no chronic effect of nicotine on perception (intensity), but rather, it would be altered in hedonic impression. The explanation for this fact lies in the non-nicotine components of cigarette that result in unbalance of taste pleasure. Moreover, taste pleasure is more associated to sweet than fat compounds. The importance of this study is that a large portion of sweet preparations also contains considerably high levels of fat.

Redington19 conducted a series of tests in smokers, 12-hour non-smokers and non-smokers in order to investigate if high levels of serum glucose would influence intensity of taste perception of sweet, salty and bitter taste, as well as the pleasure of taste. Subjects ingested a 25% glucose solution to partially mimic the effects of nicotine - increase in glycemia, that results in search for sweet food items. Prior to the administration of glucose, subjects assessed the same solutions and there were no statistically significant differences among the groups. After the intake of the solution, smokers considered very sweet solutions less agreeable and decreased scores for these solutions when compared to non-smokers. In addition, non-smokers and those who had interrupted smokin did not present a significant difference in taste for sweet solutions. After administration of solution of glucose, no tested subject modified significantly the intensity of perception, nor the test for salty and bitter tastes. The combination smoking-consumption of glucose caused a dramatic change in the pleasure of taste.

TABLE 1 - Main Effects of Smoking on Taste Perception:

• Increased perception threshold to bitter and salty substances and decrease of sweet perception.
• Aversion to sweet substances (modifications in food preference).
• Reduction of global hedonic perception (pleasure).

The differences obtained from various studies - sometimes producing contradictory results, may be explained by the variability of factors associated to taste perception among subjects, such as age, sex, genetics and food preference, among others. Variables of substances such as solubility, texture, odor and temperature are also essential (Table 1).

b) Effects of smoking on olfaction

The number of articles that intended to study the associations between smoking and the sense of smell is small. It may be caused by the fact that olfaction is still considered a second rank sense. In this field of study, there are animal experiments as well as investigations in humans.

Using electronic microscopy at ultrastructure level, Matulions14 investigated the status of respiratory epithelium of rats after exposure to smoking once or twice a day for six to nine days in three groups of animals from strains C57B1/6J and SWR/ J. Rats SWR/J were not affected by smoking. However, rats C57B1/6J showed marked changes in epithelium, including reduction of size and number of olfactory vesicles, and increase of sensorial cilia (especially those responsible for olfaction), impairing protrusion of vesicles above the surface of epithelium. Basal region of epithelium was still patent, as opposed to apical region. In the population of maintenance cells, an abnormal type of electrolucent cells was noticed among generally black cells, probably due to high-speed turnover of the exposed epithelium. In affected animals, the degree of morphological changes suggested that normal olfactory function had been compromised. It was also observed that olfactory epithelium of all animals belonging to the same strain did not react similarly to exposure, indicating that susceptibility to smoking was genetically determined.

By means of measurements conducted with an electroolfactogram (EOG), Edwards et al.5 suggested that nicotine produced EOGs with characteristics similar to those of known odors; however, sometimes there is longer duration. In addition, it was shown that a period loner than expected was necessary to make nicotine solution be diluted by olfactory mucus. The molecule of nicotine is accumulated in the cytoplasm, adding to cytotoxic effects, and its metabolites affect intracellular equilibrium. Something that has not been investigated yet is whether the fact of smoking continuously changes the physical-chemical characteristics of mucus. In order to enable molecules to reach olfactory chemoreceptors, they should be in a solution in nasal mucus7. Hall9 assessed electrical activity of olfactory bulb of 20 cats during inhaling cigarette smoking. In three out of 20 experiments they had transient periods of cortical desynchronization. The experiment was repeated with a canulla - that is, smoke was introduced deeper, to the region where most of the olfactory epithelium is located. In nine cases, there was bulbar desynchronization. They are rhythmical waves with wide oscillation (induced waves) - sometimes associated with temporal cortical desynchronization - different from oscillation waves of low amplitude (intrinsic waves) that characterize the spontaneous activities of olfactory bulb, in the absence of olfactory stimuli. Studies such as this raised the, interest of researchers to include second-hand smokers in studies1, since the olfactory perception of these people could also be affected.

Hummel et al.10 investigated topographic distribution of chemosensor potentials in relation to stimulation with nicotine, which produced not only the perception of odor, but also burning and pain sensations. This burning sensation would derive from high temperature of inhaled gases21. Three groups of patients were created, each of them with one specific sensation. Subjects scored intensity and duration of each sensation. They all had different recognition thresholds and duration of perception. Olfactory perceptions were immediately followed by painful sensation and, some seconds later, by burning sensation. Perception rates indicated that the more burning and pain increase as a result of higher concentrations of nicotine, the more olfactory perception increases; it reaches a peak of intermediary concentrations and from this point on it reduces as a result of increased concentrations. In other words, olfactory perception is not linearly related to concentration of stimuli. It suggested a suppression of olfactory perception by trigeminal system when there were high concentrations of nicotine.

One of the studies6 that had the largest sample (638 subjects - 262 non-smokers, 197 former smokers and 179 current smokers) aimed at studying the effects of smoking according to dose of use. Subjects were submitted to the Test of Identification of 40 Odors from University of Pennsylvania (UPSIT). Smoking showed to be associated with loss of capacity to identify dose-dependent odors in former and current smokers, suggesting long run changes in olfaction. Analysis of logistic regression revealed that current smokers have approximately two times more olfactory deficit that people who had never smoked [RC=1.9 (1.0-3.8)]. This deficit did not seem to be restricted to some substances. An improvement of olfaction would be present when smoking was interrupted; however, this is not something quick (for each two years of smoking two packs a day, it would be necessary to have one year of interruption).

Second-hand smokers do not normally form groups for comparison in terms of their olfactory discrimination capacity. Ahlstrom et al.1 were the pioneers to investigate that. They compared three groups (smokers, non-smokers and passive smokers) as to odor perception of two substances contained in cigarettes (pyridine and n-butane), because there were suggestions that there could be a selective reduction of odor perception of cigarette substances. This phenomenon is similar to the hearing loss "noise recruitment", caused by poor sensitivity to low stimuli and normal hearing for higher stimuli. Tests with pyridine showed that smokers had a behavior similar to the described "noise recruitment". In the group of second-hand smokers (n=15), it was observed that similarly to smokers (n=26), stimuli of n-butane were considered weaker than described by non-smokers (n=26). The reason, for such alteration was that smokers were more familiarized with the odor and then it would be a response of adaptation (or learned adjustment) and not olfactory deficit. While adaptation is characterized by control of central nervous factors, sensorial deficit is resultant from peripheral neuronal alterations. Since the reaction for both stimuli (pyridine and n-butane) was similar, reduction of olfaction in smokers would not be specific. This speculation is due to the significant presence of pyridine and the restricted presence of n-butane in tobacco, and it should, therefore, produce a proportional reaction and not an equivalent one. Based on this observation, it was suggested that the phenomenon could not be described as preferential (specific) anosmia, but rather as generalized hyposmia.

Apnea reflex, a defense mechanism of respiratory system, is also compromised in smokers. Cometto-Muniz and Cain 4 obtained results that demonstrated that smokers are 29% less sensitive to concentrations of reflex firing than non-smokers. It might be the clearest evidence of chemosensitive difference among smokers and non-smokers for an inhaling agent. This fact may be a consequence of the alteration of ciliary motility, very common in smokers because of the increase in mucus viscosity, functioning as a barrier to the perception of irritating substances. There is also an anatomical-physiological concurrent cause: the site of olfactory receptors is basically located on the superior region of nasal fossae, protecting olfactory cells from insults. As widely known, the best way to obtain olfactory perception is through deep inhalation, creating a turbulence capable of moving molecules to olfaction sites. During smoking, inhalation is not common.

An important aspect that has not seemed to be appropriately weighted, except for one study6, is accumulated smoking, a fact that could explain lack of uniformity in findings. Dichotomy between current smokers and controls ends up including in the latter group former heavy smokers who interrupted the use of cigarettes. Matching with age, sex and health status is not routinely done in studies either.

There is a potential bias in the analysis of data regarding intensity of stimuli, which could underestimate alterations of olfactory discrimination. In general, a 30% change in concentration of a substance is required to produce a change in intensity of perception by olfactory mucosa. In taste perception, a different of 30% is equally required. On the other hand, a 1% different in light intensity is usually detected by vision7.

Studies with more robust samples will be able to indicate more subtle differences in taste and odor perception and increase statistically significance, resulting in enhancement and improvement of different studies (Table 2).

FINAL CONSIDERATIONS

In addition to complexity and intrinsic difficulties of these two chemoreceptor systems, our context does not encourage the search for results either because it is not a situation of explicit morbidity, or because there is lack of financial interest by labs and medical professionals.

We do hope that scarcity of Brazilian scientific literature in the area is improved with time, thanks to the new publication of experiments and, studies. Internationally, the situation is different: in addition to publications specially dedicated to the topics of sensorial systems, such as Perception & Psychophysics, Chemical Senses and Perception, structured centers that are predominantly dedicated to studying these sensorial systems have been created, such as Smell and Taste Center at University of Pennsylvania, Clinical Olfactory Research Center at University of the State of New York and Monell Chemical Senses Center in Philadelphia-one of the first scientific institutes that conduct multidisciplinary research in olfaction, taste perception and chemosensitive irritation.

ACKNOWLEDGEMENT

We would like to acknowledge in memoriam Professor and Master Mário Rigatto - an unbeatable advocator of the fight against smoking, for his encouragement and clarification of fundamental ideas included in this study.

TABLE 2 - Main Effects of Smoking on Odor Perception.

• Ultrastructural alteration of olfactory epithelium (reduction in size and number of olfactory vesicles and number of cilia).
• Development of painful and burning sensation in olfactory mucosa.
• Reduction in the capacity of identifying non-specific and dose-dependent odors (hyposmia).

REFERENCES

1.AHLSTROM, R.; BERGLUND, B.; BERGLUND, U; ENGEN, T.; LINDVALL, T.- A comparison of odor perception in smokers, nonsmokers and passive smokers. Am. J. Otolaryngol., 8:1-6; 1987
2.ALBANES, D.; JONES, D.Y.; MICOZZI, M.S.; MATTSON, M.E. Associations between smoking and body weight in the U.S. population: analysis of NHANES II. Am. J. Public. Health., 77(4):439-44; 1987.
3. OATS, A.C.- Effects of age, sex and smoking on electrical taste threshold. Ann. Otol. Rhinol. Laryngol., 83(3): 365-9;1974.
4. COMETTO-MUNIZ, J.E.; CAIN, W.S.- Perception of nasal pungency in smokers and nonsmokers. Physiol. Behav., 29(4): 727-31; 1992. 5.EDWARDS, D.A.; MATHER, R.A.; SHIRLEY, S.G.; DODD, G.H. Evidence for an olfactory receptor which responds to nicotine - nicotine as an odorant. Experientia, 43(8): 868-73; 1987.
6. FRYE, R.E.; SCHWARTZ, B.S.; DOTY, R.L. - Dose-related effects of cigarette smoking on olfactory function. JAMA, 263(9):1233-36; 1990.
7. GANONG, W.F. - Review of Medical Physiology. 17° Edition, New Jersey, Prentice-Hall International Inc., 1995: 167-173.
8. GRUNBERG, N.E.- The effects of nicotine and cigarette smoking on food consumption and taste preferences. Addict Behav 7: 317-31; 1982.
9. HALL, G.H.- Effects of nicotine and tobacco smoke on the electrical activity of the cerebral cortex and olfactory bulb. Br. J. Pharmacol., 38(2): 271-86; 1970.
10. HUMMEL, T.; LIVERMORE, A.; HUMMEL, C.; KOBAL, G.Chemosensory event-related potentials in man: relation to olfactory and painful sensations elicited by nicotine. Electroencephalogr Clin. Neurophysiol., 84(2): 192-5; 1992.
11. JUNQUEIRA, L.; CARNEIRO, J.- Histologia Básica. 8ª edição, Rio de Janeiro; Editora Guanabara-Koogan S.A..1995. Pp. 3912.
12. KRUT, B.; ALLISON, H.- Taste perception in smokers and non-smokers. BMJ, 1: 384-87; 1961.
13. KUMAR, R.; PRATT, J.A.; STOLREMAN, I.P.- Characteristics of conditioned taste aversion produced by nicotine in rats. Br. J. Pharmacol., 79(1): 245-53; 4983.
14. MATULIONIS, D.- Ultrastructure of olfactory epithelia in mice after smoke exposure. Ann. Otol. Rhinol. Laryngol., 83(2):192201; 1974.
15. MELA, D.J. - Smokeless tobacco and taste sensitivity. N Engl J. Med 316(18):1165-66:1987.
16. PARKER, L.A.; DOUCET, K.- The effects of nicotine and nicotine withdrawal on taste reactivity. Pharmacol Biochem Behav, 52(1): 125-9; 1995.
17.PERKINS, K.A.; EPSTEIN, L.H.; STILLER, R.L.; FERNSTROM, M.H.; SEXTON, J.E.; JACOB, R.G.- Perception and hedonics of sweet and fat taste in smokers and nonsmokers following nicotine intake. Pharmacol Biochem Behav, 35(3): 671-6; 1990.
18. PETERSON, D.L.; LONERGAN, L.H.- Smoking and taste perception. Arch. Environ Health, 16(2): 219-22; 1968.
19. REDINGTON, K.- Taste differences between cigarette smokers and nonsmokers. Pharmacol Biochem Behav, 21: 2038; 1984.
20. RODIN, J.- Weight change following smoking cessation: The role of food intake and exercise. AddictBehav, 12: 303-17; 1987.
21. STAUFFER, J.L. - In: TIERNEY, L.M.; MCPHEE, SJ.; PAPADAKIS, M.A.. CURRENT: Medical Diagnosis & Treatment, 37th Edition,, New Jersey, Prentice-Hall International Inc., 1998: 251-332.

* Undergraduates, School of Medicine (FAMED) at Universidade Federal do Rio Grande do Sul (UFRGS).
** Otorhinolaryngologist, Joint Professor of the Department of Ophthalmology and otorhinolaryngology at FAMED/UFRGS.
*** Pneumologist; Faculty Professor of the Department of Internal Medicine at FAMED/UFRGS.

Affiliation: Faculdade de Medicina da Universidade Federal do Rio Grande do Sul, Hospital de Clínicas de Porto Alegre /RS.
Address for correspondence: Alexandre Armes Henriques - Av: General Barreto Viana, 547 - Chácara das Pedras - 91330-630 Porto Alegre /RS.
Tel: (55 51) 334-5530 - Fax: (55 51) 338-1535 - E-mail: aannesh@zaz.com.br
Article submitted on June 8, 2000. Article accepted on August 17, 2000.