ARTIGOS ORIGINAIS ORIGINAL ARTICLES
Saúde bucal, práticas de higiene bucal e ocorrência de câncer da cavidade oral
Luzia A MarquesI; José ElufNetoII; Rejane A O FigueiredoI; José F de GóisFilhoIII; Luiz P KowalskiIV; Marcos B de CarvalhoV; Márcio AbrahãoVI; Victor WünschFilhoI
IDepartamento de Epidemiologia. Faculdade de Saúde Pública. Universidade de São Paulo. São Paulo, SP, Brasil
IIDepartamento de Medicina Preventiva. Faculdade de Medicina. Universidade de São Paulo. São Paulo, SP, Brasil
IIIDepartamento de Cabeça e Pescoço. Instituto do Câncer Arnaldo Vieira de Carvalho. São Paulo, SP, Brasil
IVDepartamento de Cabeça, Pescoço e Otorrinolaringologia. Hospital do Câncer A.C.Camargo. São Paulo, SP, Brasil
VDepartamento de Cabeça e Pescoço. Hospital Heliópolis. São Paulo, SP, Brasil
VIDepartamento de Otorrinolaringologia, Cabeça e Pescoço. Hospital São Paulo. Universidade Federal de São Paulo. São Paulo, SP, Brasil
OBJECTIVE: To assess the association between oral health and hygiene practices and oral cancer.
METHODS: Hospitalbased casecontrol study in the metropolitan area of São Paulo, southeastern Brazil, from 1998 to 2002. A total 309 patients with squamous cell carcinoma of the mouth and the pharynx and 468 controls matched by sex and age were included in the study. Cases were recruited in seven reference hospitals and controls were selected in five out of the seven participating hospitals. Detailed information on smoking, alcohol consumption, schooling, oral health status and hygiene practices were obtained through interviews. Odds ratios (OR) and 95% confidence intervals (95% CI), adjusted by sex, age, schooling, smoking, alcohol consumption as well as the variables oral health status and hygiene practices were estimated using unconditional logistic regression analyses.
RESULTS: The use of complete dental prosthesis was not associated with oral cancer but regular gum bleeding showed a strong association (OR 3.1; 95% CI 1.27.9). Those who never attended a dental visit were more likely to have oral cancer (OR 2.5; 95% CI 1.34.8). Daily mouthwash use showed a stronger association to pharynx (OR 4.7; 95% CI 1.812.5) than mouth cancer (OR 3.2; 95% CI 1.66.3).
CONCLUSIONS: Gum bleeding, no dental care, and daily mouthwash use were factors associated with oral cancer regardless of tobacco and alcohol consumption.
Descriptors: Mouth Neoplasms. Oral Hygiene. Oral Health. Risk Factors. CaseControl Studies.
OBJETIVO: Avaliar a associação entre saúde e higiene bucal na ocorrência de câncer oral.
MÉTODOS: Estudo casocontrole de base hospitalar, realizado entre 1998 e 2002 na área metropolitana de São Paulo, SP. Foram incluídos 309 pacientes com carcinoma epidermóide de boca e orofaringe e 468 controles, pareados com os casos por sexo e idade. Os casos foram levantados em sete hospitais que concentram a assistência médica a pacientes com a doença e os controles rastreados em cinco hospitais gerais dentre os sete. Informações pormenorizadas sobre tabagismo, consumo de álcool, escolaridade, saúde bucal e práticas de higiene bucal foram obtidas por entrevista. Por meio de análise de regressão logística não condicional foram calculados odds ratios (OR) e intervalos com 95% de confiança (IC 95%), ajustados por sexo, idade, nível educacional, tabagismo e consumo de álcool, bem como para as demais variáveis de saúde e higiene bucal.
RESULTADOS: Uso de prótese bucal total não se associou a câncer oral, mas sangramento gengival freqüente apresentou alta associação (OR=3,1; IC 95%: 1,2;7,9). Nunca ter consultado dentista mostrou associação com câncer oral (OR=2,5; IC 95%: 1,3;4,8). Uso diário de enxaguatórios bucais apresentou associação mais intensa com tumores de faringe (OR=4,7; IC 95%: 1,8;12,5) do que com tumores de boca (OR=3,2; IC 95%: 1,6;6,3).
CONCLUSÕES: Sangramento gengival, ausência de consultas com dentistas e uso regular de enxaguatórios bucais foram fatores associados com câncer oral, independentemente de tabagismo e consumo de bebidas alcoólicas.
Descritores: Neoplasias Bucais. Higiene Bucal. Saúde Bucal. Fatores de Risco. Estudos de Casos e Controles.
Oral cancer accounts for 7% of all new cancer cases worldwide, around 270,000 cases annually. In developing countries, it is the fifth most common cancer in men and the seventh in women.15 In South America, Brazil has the highest rates of oral cancer in men.19 The city of São Paulo has the highest incidence of oral cancer in the Brazil, 25.3 cases per 100,000 males and 4.9 per 100,000 females.20
Smoking and alcohol consumption are widely recognized as risk factors for oral cancer,10,1,2,3 but also nutrition,6 occupation1 and metabolic gene polymorphisms9 seem to have specific roles for the development of oral cancer. Previous epidemiological studies have provided evidence of the effect of oral health status and oral hygiene practices on the development of oral cancer. Poor oral hygiene seems to cause additional cancer risks. Irregular toothbrushing, dental prostheses, and mouthwash use have all been associated with oral cancer.3,14,17 A casecontrol study conducted in Puerto Rico18 reported nonsignificant risk of frequent mouthwash use for oral cancer in nonsmokers and nonalcohol users. Even though associations between mouthwash use and oral cancer have been reported in the literature, there is no conclusive evidence of increased risk.7 Alcohol is a component of many commercial mouthwash formulations, which could explain the observed increased risk with their use for oral cancer. In Brazil, mouthwash products are freely marketed and some brands may contain as much as 27% alcohol.
The objective of the study was to assess the association between oral health status and hygiene practices and oral and pharynx cancer.
Hospitalbased casecontrol study conducted in the metropolitan area of São Paulo (Southeastern Brazil) between November 1998 and March 2002. Cases were recruited in seven reference hospitals in this area. Controls were selected from five out of these seven general hospitals. Cases and controls had to be living in the metropolitan area of São Paulo for at least six months.
Eligible cases were all patients newly diagnosed during the study period and histologically confirmed as oral squamous cell carcinoma, and coded according to the International Classification of Diseases and Related Health Problems 10th revision (ICD10): oral cavity (C00, C02, C03, C04, C05.0, C06) and pharynx (C01, C02.4, C05.1, C05.2, C09, C10). Patients with cancers of the vermilion border, parotid gland and other major salivary glands, nasopharynx, paranasal sinus, and hypopharynx, or classified as nonspecific sites of the oral cavity, lips, or pharynx, were excluded.
Out of 573 cases recruited, 210 were considered ineligible according to one of exclusion criteria: 74 were living outside the study area; 79 did not have their cancers histologically confirmed; 32 had undergone previous treatment for oral cancer; six had cancer involving pharynx and larynx anatomical areas of unknown origin; seven patients had tumor primary sites other than those established in the study; 12 cases were not confirmed as squamous cell carcinoma. Of 363 eligible cases, 28 were severely ill and unable to undergo an interview and 26 refused to participate in the study. The overall response rate was 85.1%. A total of 309 cases were interviewed and included in the analysis.
Cases and controls were matched by sex and age (5year age groups). Controls were selected from patients admitted in the participating general hospitals during the study period. Controls had no previous history of oral cavity or pharynx cancer. Patients with diseases associated with oral cancer risk factors (tobacco smoking, alcohol consumption and occupation) such as emphysema, cirrhosis, asthma, ischemic heart diseases, cancers, and chronic sinusitis were excluded.
Of 527 controls selected, 24 were excluded as they were living outside of the study area. Of the remaining patients, 31 were severely ill and unable to undergo an interview; and four refused to participate in the study. A total of 468 controls were interviewed and included in the analysis.
All contacts and interviews with cases and controls were carried out by trained interviewers who were not blinded to the patient status as case or control, but did not have any knowledge on the study hypothesis.
Data were collected through facetoface interviews using a standardized questionnaire applied to all cases and controls. The questionnaire included items on socioeconomic status, occupation, smoking, alcohol consumption, family history of cancer, infectious diseases, diet, sexual behavior, and oral health and hygiene practices. The questionnaire was piloted on a group of ten cases and ten controls, and few adjustments were made in the instrument before its application.
The study subjects were inquired regarding the period before the diagnosis of the conditions for which they were hospitalized. General oral health/hygiene practices were assessed using standard criteria for the following variables: use of complete dental prosthesis, gum bleeding, frequency of dental visits, tooth brushing and frequency of mouthwash use.
Subjects were classified into the following categories of alcohol consumption: nondrinkers, former drinkers (no alcohol consumption for at least one year prior to the interview), or current drinkers. Additionally, average daily alcohol intake was estimated in grams per liter per day (g/L/day). Subjects' consumption of different beverages was converted into liters of alcohol as follows: beer, 5%; wine, 12%; spirits (cachaça, whiskey, vodka, rum), 41%; and liqueurs, 30%. Then they were converted into grams of alcohol (one liter is equivalent to 798 g of alcohol).1 Five categories were created: <1 g/L/day (nondrinker or drinking at most one glass of wine every ten days); 1 to <40 g/L/day; 40 to <80 g/L/day; 80 to <120 g/L/day; and >120 g/L/day.
Smoking status was categorized into nonsmokers, former smokers (no smoking for at least one year prior to the interview), and current smokers. Cumulative exposure was determined based on the number of daily packs of cigarettes consumed per year (packyears). The following criteria were used: one cigarette is equivalent to 1 g of tobacco; the content of one pipe is equivalent three cigarettes; and one cigar is equivalent to four cigarettes.2 One pack contains 20 cigarettes. Five categories were created for tobacco consumption: <1 packyear (nonsmoker or smoker of at most 0.05 cigarettes per day); 1 to <20 packyears; 20 to <40 packyears; 40 to <60 packyears; and >60 packyears.
Use of complete dental prosthesis (both jaws) was considered a dichotomous variable (yes/no). Gum bleeding was ascertained by asking subjects about bleeding during tooth brushing (no; sometimes; always; or almost always). The study hypothesis was that subjects who reported gum bleeding every or almost every time they brushed their teeth would have higher risk of oral cancer than those with no or sporadic gingival bleeding.
Frequency of dental visits in the last twenty years was categorized as never, occasionally (intervals >2 years), and regular annual visits. For tooth brushing frequency, subjects were divided into four groups: no brushing; less than once a day; once or twice a day; and three or more times a day. Frequency of mouthwash use was categorized as never; less than once a day; and more than once a day.
Sample size calculation considered the population prevalence of use of complete dental prosthesis, one of the main independent study variables. In a survey conducted by the Health Department of the State of São Paulo in 19984 it was estimated that around 29% of the state's adult population used dental prostheses. Then, assuming a=0.05 and b=0.20, 309 cases and an equivalent number of controls, the study would detect an odds ratios (OR) of 1.6.16
The association between oral health and hygiene practices and oral cancer was estimated by OR and 95 percent confidence intervals (95% CI) using unconditional logistic regression analysis.5 For each oral health status or hygiene practice variable the OR was adjusted by all other oral health variables and by sex, age (five age groups), smoking (packyears), alcohol consumption (g/L/day), and schooling (illiterate, elementary, high school, and college education). Linear trend was estimated for alcohol consumption and smoking by fitting the categorical variable into the logistic regression model as a continuous variable.10
The study was approved by the Research Ethical Committees of all participating medical sites and Universidade de São Paulo School of Public Health, as well as by the National Human Research Ethics Committee of the Brazilian Ministry of Health. Cases and controls were informed of the objectives and uses of the research project both in person and through a consent form, which was signed by them.
The study sample included 198 cases of mouth cancer and 111 cases of pharynx cancer. The most common anatomical site of cancer was the tongue (ICD10 code C01 base of tongue 5.5%; ICD10 code C02 other and unspecified parts of tongue 32%). Digestive conditions were generally reported as the main reason for hospital admission in 24.1% of controls, followed by cardiovascular diseases (21.6%). However, the distribution according to specific diseases was lower than 10% for each diagnosis. For instance, in the great group of chapter XI of ICD10 (diseases of the digestive system), the most common diagnosed disease among controls was ICD10 code K40.9 (unilateral or unspecified inguinal hernia, without obstruction or gangrene). There were 28 patients diagnosed with this disease, only 6.0% of all controls. No patient with alcohol or tobaccorelated digestive or cardiovascular diseases was included as control in the study.
Maletofemale ratio was 5.7:1 and 84.2% aged between 40 and 69 years. About 70% of both cases and controls were classified by interviewers as white skin. Schooling was slightly higher among controls, but this difference was not statistically significant (Table 1).
Current drinkers showed higher risk of oral cancer than former drinkers and current smokers had higher risk compared to former smokers. Although a clear doseresponse effect was seen in both tumor sites, increased alcohol consumption was associated to higher risk of pharynx than mouth cancer. As for tobacco smoking, no clear risk differences were seen between pharynx and mouth cancers. Only in those in the strata of tobacco smoking of 40 to 59 packyears the risk was higher for pharynx than mouth cancer (Table 2).
The use of complete dental prosthesis was not associated with increased risk of oral cancer (Table 3). The risk of oral cancer was higher among subjects who reported regular gingival bleeding during tooth brushing. Those who reported never attending dental visits had a statistically significant oral cancer risk of 2.5 compared to those who attended annual dental visits. In the crude analysis, those who reported never brushing their teeth showed higher risk of oral cancer compared to those who brushed their teeth three times a day (OR=2.6, 95% CI: 1.1;5.9), but the association lost its statistical significance after adjustment. Subjects who reported using mouthwash products more than once a day showed a statistically significant 3fold increase in the risk of oral cancer compared to those who never used them.
The effect of distinct oral health status and health care variables on mouth and pharynx cancer was assessed separately. For those with frequent gingival bleeding, no dental visits, and mouthwash use, the risk of pharynx cancer was greater than mouth cancer (Table 4).
A possible limitation of the present study includes the risk of misclassification of exposure history. The sensitivity of selfreported oral health status is lower than objective bucal examination. However, mouth examination in patients with oral cancer could be compromised in order to obtain information on gingival bleeding previously to the disease. However, selfreported oral health status and practices may have introduced a nondifferential misclassification that would dilute risk estimates.
There is no simple explanation for the high incidence of cancer of oral cavity and oropharynx found in São Paulo,20 since the prevalence of tobacco smoking and alcohol consumption seems not be different from other Brazilian regions.13 The results in the present study for the effect of tobacco smoking and alcohol consumption confirm what is long known, that these two exposures are the main risks for oral cavity and oropharynx cancers.10
No association was seen between complete dental prosthesis and oral cancer, corroborating other studies.8,11,12 On the other hand, high association of frequent gum bleeding during toothbrushing and oral cancer was found. Maier et al12 reported among oral cancer cases that 40.9% had 3 mm or more dental calculus or tartar and 28% had chronic gingival inflammation, thus indicating that patients with oral cancer could have poorer oral care and health. Other studies carried out in Brazil8 and India2 showed increased oral cancer risk among subjects with poor oral hygiene practices. In the present study, information on missing and decayed teeth (DMF index) and the index of periodontal disease were not used; instead, oral health status was assessed through the variables dental prostheses and frequency of gum bleeding during toothbrushing, as selfreported by interviewees.
In regard to oral hygiene practices, a nonsignificant association for oral cancer was seen among subjects that never brushed their teeth compared to those who brushed them three or more times a day. Patients who had never attended dental visits had a higher risk of oral cancer than subjects who reported attending a visit at least once a year. This finding is similar to that found in studies in Germany12 and Poland.11 Those who have frequent dental visits are generally more careful about their oral hygiene, and this behavior could be considered an indicator of good oral care, which can help eliminate or prevent mucosa exposure to certain carcinogens.
There seems to be collinearity between the variables frequency of gingival bleeding and frequency of dental visits, since subjects with poor oral health are probably those who do not often attend dental visits. On the other hand, the findings for toothbrushing could be downplayed, considering that subjects with gingival diseases are less likely to brush their teeth because of gingival bleeding during this procedure.
In the present study, the proportion of cases who reported using mouthwash products one or more times a day was a little over twice that of controls (15.4% vs. 6.9%, respectively). Association with oral cancer was 3fold higher in subjects who frequently used mouthwash products than among those who never used them. This finding confirms in Brazil the results from Winn et al18 (2001) that found higher risk, though not statistically significant, among a small number of nonsmoking and nondrinking subjects mouthwash users. There is no available data on mouthwash use in Brazilian population. The prevalence among controls in this study (6.9%) could be assumed as a rough estimate, even though a higher prevalence of mouthwash use could be found among those hospitalized, since they could use mouthwash only because their diseases. There are several brands of mouthwash products marketed in Brazil, some of them with alcohol in their formulae. Alcohol is used in mouth rinses as a solvent of other ingredients and a preservative. Certain brands of mouthwash agents commercially available contain as much as 27% alcohol, and these products are sold over the counter. Mouthwash products are in contact with the oral mucosa as much as alcoholic beverages and may cause chemical aggression of the cells. Similarly to the findings of alcohol consumption, subjects who reported frequent use of mouthwash products also showed higher risk of pharyngeal than mouth cancer. In general, when washing their mouth, people gargle the mouthwash product and the substances of these products come into direct contact with the pharynx mucosa. Special attention should be paid to the differences that may exist between mouth and pharynx mucosas, and to the consequently distinct effects of alcohol, present in beverages and mouthwash formulas, on cancers in these anatomical sites. Also, the mechanisms of alcohol carcinogenesis and the role of acetaldehyde, as a solvent for other carcinogens, and ethanol in DNA damage, or reduced immune surveillance carcinogenicity of nonethanol compounds, need to be further explored,4 as they could provide better understanding on cancer development in the head and neck.
The potential bias related to the fact that subjects may use mouthwash to reduce the aftertaste left by smoking or drinking was controlled for in the multiple logistic regression analysis adjusted for smoking and alcohol consumption.
Although it was stressed to the interviewers to ask cases and controls information on oral health status and practices before disease diagnosis (oral cancer for cases and other diseases for controls), the occurrence of reversed causality cannot be excluded. Particularly, patterns of gingival bleeding and mouthwash use could be different after oral cancer patients perceived their disease. Oral neoplasia could make oral care difficult and contribute for the development of gingivitis and, consequently, gingival bleeding. On the other hand, more advanced oral cancers can cause bad breath and make patients try to minimize this inconvenience using mouthwash.
In conclusion, frequent gum bleeding, no dental visits and mouthwash use more than once a day are highly associated with oral cancer. However, reverse causation should be also taken in consideration as a potential alternative explanation.
1. Andreotti M, Rodrigues AN, Cardoso LM, Figueiredo R, Eluf Neto J, WünschFilho V. Ocupação e câncer da cavidade oral e orofaringe. Cad Saude Publica. 2006;22(3):54352.
2. Balaram P, Sridhar H, Rajkumar T, Vaccarella S, Herrero R, Nandakumar A, et. al. Oral Cancer in southern India: the influence of smoking, drinking, paanchewing and oral hygiene. Int J Cancer. 2002;98(3):4405.
3. Blot WJ, Winn DM, Fraumeni JF Jr. Oral cancer and mouthwash. J Natl Cancer Inst. 1983;70(2):2513.
4. Boffetta P, Hashibe M. Alcohol and cancer. Lancet Oncol. 2006;7(2):14956.
5. Breslow E, Day E. Statistical methods in cancer research. The analysis of casecontrol studies. Lyon: IARC Press; 1980. (IARC Scientific publications, 82).
6. Marchioni DML, Fisberg RM, Góis Filho JF, Kowalski LP, Carvalho MB, Abrahão M, et al. Dietary patterns and risk of oral cancer: a casecontrol study in São Paulo, Brazil. Rev Saude Publica. 2007;41(1):1926.
7. Elmore JG, Horwitz RI. Oral cancer and mouthwash use: evaluation of the epidemiologic evidence. Otolaryngol Head Neck Surg. 1995;113(3):25361.
8. Franco EL, Kowalski LP, Oliveira BV, Curado MP, Pereira RN, Silva ME, et al. Risk factors for oral cancer in Brazil: a casecontrol study. Int J Cancer. 1989; 43(6):9921000.
9. Gattás GF, Carvalho MB, Siraque MS, Curioni OA, Kohler P, Eluf Neto J, et al. Genetic polymorphisms of CYP1A1, CYP2E1, GSTM1, and GSTT1 associated with head and neck cancer. Head Neck. 2006;28(9):81926.
10. Hashibe M, Brennan P, Benhamou S, Castellsague X, Chen C, Curado MP, et al. Alcohol drinking in never users of tobacco, cigarette smoking in never drinkers, and the risk of head and neck cancer: pooled analysis in the International Head and Neck Cancer Epidemiology Consortium. J Nat Cancer Inst. 2007;99(10):77789.
11. Lissowska J, Pilarska A, Pilarski P, SamolczykWanyura D, Pickarczyk J, BardinMikollajczak A, et al. Smoking, alcohol, diet, dentition and sexual practices in the epidemiology of cancer in Poland. Eur J Cancer Prev. 2003;12(1):2533.
12. Maier H, Zoller J, Herrmann A, Kreiss M, Heller WD. Dental status and oral hygiene in patients with head and neck cancer. Otolaryngol Head Neck Surg. 1993;108(6):65561.
13. Marcopito LF, Rodrigues SS, Pacheco MA, Shirassu MM, Goldfeder AJ, Moraes MA. Prevalência de alguns fatores de risco para doenças crônicas na cidade de São Paulo. Rev Saude Publica. 2005;39(5):73845.
14. Mashberg A, Barsa P, Grossman ML. A study of the relationship between mouthwash use and oral and pharyngeal cancer. J Am Dent Assoc. 1985; 110(5):73134.
15. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin. 2005;55(2):74108.
16. Schlesselman JJ. Casecontrol studies. Design, conduct, analysis. New York: Oxford University Press; 1982.
17. Velly AM, Franco EL, Schlecht N, Pintos J, Kowalski LP, Oliveira BV, et al. Relationship between dental factors and risk of upper aerodigestive tract cancer. Oral Oncol. 1998;34(4):28491.
18. Winn DM, Diehl SR, Brown LM, Harty LC, BravoOtero E, Fraumeni JF Jr, et al. Mouthwash in the etiology of oral cancer in Puerto Rico. Cancer Causes Control. 2001;12(5):41929.
19. WünschFilho V, Camargo EA. The burden of mouth cancer in Latin America and the Caribbean: epidemiologic issues. Semin Oncol. 2001;28(2):15868.
20. WünschFilho V. The epidemiology of oral and pharynx cancer in Brazil. Oral Oncol. 2002;38(8):73746.
Departamento de Epidemiologia
Faculdade de Saúde Pública USP
01246904 São Paulo, SP, Brasil
Supported by the European Commission Contract No IC18CT970222 and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP Proc. No 2001/017682).
1 International Agency for Research on Cancer. Alcohol drinking. Lyon; 1988. (IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, 44).
2 International Agency for Research on Cancer. Tobacco smoking. Lyon; 1986. (IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, 38).
3 International Agency for Research on Cancer. Tobacco smoke and involuntary smoking. Lyon; 2004. (IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, 83).
4 Secretaria de Estado da Saúde de São Paulo. Levantamento epidemiológico em saúde bucal: Estado de São Paulo, 1998. São Paulo: Núcleo de Estudos e Pesquisas de Sistemas de Saúde da FSPUSP; 1999.