risk variables in the association between frequency of alcohol consumption and periodontitis

Risk variables in the associationbetween frequency of alcoholconsumption and periodontitisLages EJP, Costa FO, Lages EMB, Cota LOM, Cortelli SC, Nobre-Franco GC,Cyrino RM, Cortelli JR. Risk variables in the association between frequency ofalcohol consumption and periodontitis. J Clin Periodontol 2012; 39: 115–122. doi:10.1111/j.1600-051X.2011.01809.x.

AbstractObjective: Investigate the association between the frequency of alcohol consump-tion and periodontitis. Moreover, evaluate the influence of biological, behaviour-al, and social risk variables in this association.Methods: Sample was comprised by 542 subjects of both genders, 35–55 years ofage, who underwent a complete periodontal examination, and was divided intofour groups according to the frequency of alcohol use, based on alcohol use dis-orders identification test (AUDIT) and Cut-down, Annoyed, Guilty, Eye-opener(CAGE) instruments: (1) no or occasional alcohol use (NA), (2) moderate alcoholuse (MA), (3) intense alcohol use (IA) and (4) alcohol dependence (DA). Associa-tions between the occurrence of periodontitis and potential risk variables wereanalysed by univariate and multivariate logistic regression stratified by smokingstatus when appropriate.Results: The prevalence of periodontitis in NA, MA, IA and DA groups were17.2%, 24.0%, 29.6% and 53%, respectively. Alcohol odds ratio (OR) estimatessignificantly increased with an increase in consumption frequency (DA>IA>-MA>NA) and were approximately two times higher in smokers (OR = 3.43 to7.91) compared to non-smokers (OR = 1.22 to 3.02).Conclusion: Occurrence of periodontitis among alcohol users were high and thefrequency of alcohol consumption increased the odds of periodontitis incremen-tally mainly in smokers.

Eugenio J. P. Lages1, Fernando O.Costa1, Elizabeth M. B. Lages1, Luıs

O. M. Cota1, Sheila C. Cortelli2,Gilson C. Nobre-Franco2, Renata M.

Cyrino1 and Jose R. Cortelli2

1Department of Periodontology, Dentistry

School, Federal University of Minas Gerais,

Belo Horizonte, Brazil; 2Department of

Dentistry, Periodontics Research Division,

University of Taubate, Taubate, Sao Paulo,

Brazil

Key words: alcohol; periodontitis;

prevalence; risk factors

Accepted for publication 12 September 2011

It is well known that the complexinteraction between the microflora ofthe oral cavity and the host organ-ism is an important issue in theunderstanding of causal associa-

tions between systemic conditionsand periodontal diseases suscepti-bility. Classically, diabetes, smok-ing, immunosuppression, hormonalchanges, stress, medication and alco-holism are recognized factors thatcan alter the pathogenesis, theexpression and the clinical manage-ment of periodontal diseases (Alban-dar 2002, Costa et al. 2009, Sarfatiet al. 2010).

According to Tezal et al. (2001),the consumption of alcoholic bever-ages was associated with the severityof periodontal disease independentlyof oral hygiene levels. By contrast,

other studies suggested that the sever-ity of periodontitis was associated withbehavioural risk factors, although apoor oral hygiene level was a commonfinding among alcohol users (Nova-cek et al. 1995, Amaral et al. 2009).

The abusive alcohol use is consid-ered a serious global problem thatfavours a life style that can bringsocial, economic and health prob-lems (Casswell 2011). A study com-prising 8,589 subjects in Brazilreported that 68.7% of the samplewas composed by alcohol users and11.2% were considered alcoholdependents (Amaral et al. 2008).

Conflict of interest and source offunding statement

The authors declare that they have noconflict of interest. This study wassupported by grants from theResearch Support Foundation of Min-as Gerais (FAPEMIG-00207-11) andNational Research Council (CNPq/Brazil – 474235/2010-6)

© 2011 John Wiley & Sons A/S 115

J Clin Periodontol 2012; 39: 115–122 doi: 10.1111/j.1600-051X.2011.01809.x

In the last decade, conflictingfindings were reported on the associ-ation between alcohol consumptionand periodontitis (Sakki et al. 1995,Tezal et al. 2001, 2004, Pitiphatet al. 2003, Nishida et al. 2004, Shi-mazaki et al. 2005, Okomoto et al.2006, Amaral et al. 2009, Sarfatiet al. 2010). Therefore, studies withgreater methodological rigour on theassociation between alcohol con-sumption and periodontitis shouldbe conducted, especially focusing onthe alcohol exposure gradients andperiodontitis definition. Moreover,few studies have evaluated alcoholdependence (DA; Amaral et al.2008). According to previous studies,it was hypothesized that higher alco-hol consumption was associated withworse periodontal condition.

The present study aimed to inves-tigate the association between thefrequency of alcohol consumptionand periodontitis; in addition, inves-tigate the influence of biological,behavioural and social risk variablesin this association.

Methods

Study sample

In the present cross-sectional study,complete periodontal examinations,as well as interviews for collectingbiological, behavioural and socialvariables, were performed in a con-venient sample of 542 subjects, fromJune 2008 to December 2010. It wasapproved by the Ethics ResearchCommittee of the Federal Universityof Minas Gerais (COEP/UFMGprotocol #0094.0.203.000-10).

All subjects were selected from asingle medical/dental waiting list forhealth care in general practice of1420 subjects at three Health Centresfrom the west region of Belo Hori-zonte city, Brazil. Subsequent to theanalysis of medical records and theexclusion of subjects aged <35 and>55 years, 1010 subjects were deter-mined to be eligible and were invitedthrough telephone calls to partici-pate in the study. At that time, 197subjects could not be contacted and813 subjects were scheduled for clini-cal examinations and interviews.From this total, 173 did not meetthe inclusion/exclusion criteria, 79did not attend scheduled exams/interviews, 6 refused to undergo

periodontal examination allegingpotential discomfort, and 13 refusedto answer alcohol related question-naires. A flowchart of samplingstrategy is presented in Fig. 1. Thefinal sample for the present studycomprised 542 subjects from both en-ders, 35–55 years of age (47.3 ± 5.7),heterogeneous socio-economic andeducational levels. In attempt toreduce variation resulting from abroad age range, the present studyfocused on a main age group of 35–55 years.

The following two questionnaireswere used to determine the frequencyand intensity of alcohol consump-tion: (1) AUDIT – Alcohol Use Dis-orders Identification Test (Saunderset al. 1993), comprised by 10 ques-tions, being scores >8 indicatingalcohol use problems and (2) CAGE –Cut-down, Annoyed, Guilty, Eye-opener (Mayfield et al. 1974, Ewing1984), comprised by four questions,being scores � 2 indicating DA. Bothquestionnaires were validated for theBrazilian population (Masur &Monteiro 1983, Mendoza-Sassi &Beria 2003). In this manner, subjects

were categorized according toAUDIT and CAGE scores in fourgroups: (1) no use or occasional alco-hol use (NA) – never used or fre-quency of use less than monthly,AUDIT and CAGE scores = 0, (2)moderate alcohol use (MA) – fre-quency of use 2–4 times a month,AUDIT score � 8 and CAGEscore = 0, (3) intense alcohol use (IA)– frequency of use 2–3 times a week,AUDIT score � 8 and CAGE score� 1 and (4) DA – frequency of usefour or more times a week, AUDITscore >8 and CAGE score � 2.

After signing an informed con-sent, subjects were asked about theirmedical history, socio-demographicdata, AUDIT and CAGE question-naires by one examiner, followed byclinical examination by other exam-iner blinded to the interviews. Inaddition, subjects presenting peri-odontal treatment needs weredirectly referred to specialized publicDental Care Units. Subjects classi-fied as presenting intense alcohol use(IA group) or DA group wereinstructed to seek specialized treat-ment in the Psychotropic Drugs

1420 subjects in the waiting list for

medical/dental

treatment

410 subjects aged <35 to

>55 years were

excluded

1010 subjects

were determined to be eligible for

the study

197 subjects

were not found

173 subjects did not meet the inclusion

criteria

79 subjects did not attend the

scheduled

examinations

6 subjects refused to undergo

periodontal examination

Finalsample

n = 542

13 subjects refused to fulfil

questionnaires

141 subjects with <14 teeth;

9 subjects using antibiotics;11 pregnant women;

12 drug users.

Fig. 1. Sampling strategy flow.

© 2011 John Wiley & Sons A/S

116 Lages et al.

Care Unit of the Clinics Hospital ofthe Federal University of MinasGerais and in the Association ofAlcoholics Anonymous of Belo Hor-izonte city.

Exclusion criteria

Subjects were excluded from thestudy if: (1) pregnant women, (2)presenting debilitating diseases thatcompromise the immune system(HIV/AIDS, neoplasias, auto-immunediseases), (3) presenting drug-inducedgingival overgrowth, (4) age <35years or >55 years, (5) using illicitdrugs (cannabis, cocaine, crack,others), (6) using antibiotics in aperiod less than 3 months prior toexaminations and (7) presenting lessthan 14 teeth in the oral cavity.

Sample characterization

Data related to sample characteris-tics were collected: gender, age (35–55 years), family income [Brazilianminimum salary = approximately to$258 Euros (BMS <5 or � 5)], edu-cational level (<8 or � 8 years),cohabitation status (with or withoutcompanion), plaque index, last den-tal visit (last 2 years, 2–5 years,>5 years), tooth loss (mean numberof lost teeth), smoking [non-smokers,former smokers and smokers forthose who reported having smokedless than or more than 100 cigarettesduring life time, respectively (Tomar& Asma 2000)], diabetes [glycaemicvalues >126 mg/dl or those takinghypoglycaemic agents for more than2 weeks (American Diabetes Associ-ation, 2005)] and body mass index(BMI:<25 and � 25 Kg/m2).

Clinical periodontal examination

Periodontal parameters were col-lected for all subjects at four sites(vestibular, lingual, mesial and dis-tal) in all present teeth: probingdepth (PD), clinical attachment level(CAL), bleeding on probing [(BOP)evaluated 30–60 s after probing mea-surements and recorded as present/absent], lost teeth and plaque index(Silness & Loe 1964). The full-mouthperiodontal examination was per-formed using manual periodontalprobe (PCPUNC15BR – Hu-Frie-dy®). All examinations and inter-views were performed by two

trained periodontists (FOC andEJPL).

Periodontitis definition

Periodontitis was defined as the pres-ence of four or more teeth with oneor more sites with PD � 4 mm andCAL � 3 mm at the same site (Lo-pez et al. 2002).

Intra- and inter-examiner agreement

Measurements of PD and CAL wereperformed and repeated after 1 weekin 12 subjects randomly selectedfrom a pilot study sample (n = 60).Results showed weighted kappa val-ues >0.89 and intra-class correlationcoefficient >0.87 for intra- and inter-examiner agreement (p < 0.001). Atraining and calibration processregarding AUDIT and CAGE ques-tionnaires were also performed inthe pilot study. An agreement corre-lation coefficient was calculated forCAGE and AUDIT questionnaireanswers, and 100% of agreementwas reached in classifying individualsin the four proposed study groups.

Statistical analysis

Statistical analysis included adescriptive characterization of thesample according to variables ofinterest. Group comparisons bymeans of Chi-squared test and ANOVA

were performed when appropriate.When equal variances were assumed,variables were compared using ANO-

VA and Bonferroni post hoc test.When equal variances were notassumed, variables were comparedby means of Welch and Tamhanepost hoc test.

Distribution of independent vari-ables (alcohol use and explanatoryvariables) by periodontitis status,crude odds ratio (OR) and their95% confidence intervals (CI) werecalculated. The effect of variables inthe occurrence of periodontitis wasevaluated through multivariate logis-tic regression. To evaluate the inter-action effect of smoking and alcoholand to avoid the treatment of smok-ing as a confounding factor, threedistinct models logistic regressionstratified by smoking status (smok-ers, former smokers and non-smok-ers) were performed. Variablespresenting a p-value p < 0.25 in the

bivariate analysis were then includeda multivariate model of non-cases(0) and cases (1) of periodontitis andmanually removed (step by step)until the log-likelihood ratio testindicated that no variable should beremoved. Age, gender, familyincome, educational level and diabe-tes were included in the final modelsbecause they were variables stronglyassociated with periodontitis. Allvariables included in the final multi-variate model were determined to beindependent through the assessmentof their co-linearity. Areas under thereceiver operating curves (ROC), anestimate of the accuracy of the mod-els, were also reported.

All analysis was performed usingstatistical software SPSS 16.0 (Statis-tical Package for Social Sciences,Version 9.0 for Windows – SPSSInc., Chicago, IL, USA). Resultswere considered significant if p <0.05 were attained.

Results

Table 1 shows the distribution ofvariables of interest in study groups.The prevalence of alcohol consump-tion were of 45.2% (n = 245).A high prevalence of smokers andformer smokers in IA and DAgroups was observed. Mean timeof alcohol consumption (years)increased along the groups, that is,MA<IA<DA, respectively, 26.3 (±9.7),27.3 (±10.4), 28.6 (±9.7) (p < 0.05).

Periodontal variables were evalu-ated in all groups and are presentedin Table 2. Subjects from MA, IAand DA showed higher plaque indexvalues and mean number of siteswith BOP, mean and percentage ofsites with PD � 5 mm and CAL� 5 mm when compared to subjectsin NA group. These differences weresignificant among all pair of groupcomparisons, including DA versusIA, DA versus MA, DA versus NA,IA versus MA, IA versus MA, IAversus NA and MA versus NA.

Subjects in all groups presenteda high mean number of presentteeth (23.8 ± 2.3). A total of 51,598sites were examined totalling amean of 95.2 sites per subject. InTable 2, it can be observed thatthe average number of lost teeth inDA and IA groups was significantlyhigher than in NA and MAgroups.


Association between alcohol consumption and periodontitis 117

A total of 134 (24.7%) subjectswere diagnosed with periodontitis(cases) according to the definitionadopted in the present study(Table 3). The prevalence of peri-odontitis in the NA, MA, IA andDA groups was of 17.2% (refer-ence), 24.0% (crude OR = 1.52, CI1.07–2.65), 29.6% (crude OR = 2.03,

CI 1.15–3.57) and 53% (crudeOR = 4.72, CI 2.73–7.92), respec-tively. Therefore, the prevalence ofperiodontitis and alcohol OR esti-mates significantly increased with anincrease of alcohol exposure(DA>IA>MA; p < 0.05).

In the univariate analysis, indi-viduals that presented family income

<5 BMS, smokers and former smok-ers were significantly associatedwith occurrence of periodontitis(p < 0.001) (Table 3).

Significant variables and adjustedOR estimates for the occurrence ofperiodontitis in the final logistic mod-els stratified by smoking status were:(1) Smokers: alcohol use (MA group:

Table 1. Characteristics of alcohol consumption groups

Alcoholconsumption

No/occasionalalcohol use (NA)

Moderate alcoholuse (MA)

Intensive alcoholuse (IA)

Alcohol dependence(DA)

Total samplen = 542

Total n = 297 (54.8%) n = 96 (17.7%) n = 81 (15%) n = 68 (12.5%)

Gender*Women 262 (90%) 19 (6.4%) 6 (2%) 5 (1.6%) 292Men 41 (16.5%) 50 (19.8%) 87 (34.9%) 72 (28.8%) 250

Age� 35–45years 153 (61.3%) 28 (11.3%) 38 (15.2%) 31 (12.2%) 250>45–55years 155 (53.1%) 40 (13.8%) 48 (16.3%) 49 (16.8%) 292

Family income*<5BMS 141 (61.6%) 24 (10.3%) 36 (15.6%) 29 (12.5%) 230� 5BMS 132 (42.4%) 67 (21.3%) 88 (28.2%) 25 (8.1%) 312

Educational level<8years 120 (38.7%) 87 (28.1%) 64 (20.6%) 39 (12.6%) 310� 8years 78 (33.5%) 74 (32.1%) 52 (22.4%) 28 (12.1%) 232

Co-habitation status*With companion(family/friends)

119 (32.0%) 105 (28.3%) 94 (25.2%) 54 (14.5%) 372

Without companion 52 (30.6%) 40 (23.4%) 35 (20.7%) 43 (25.3%) 170Body mass index*

� 25Kg/m2 103 (61.6%) 21 (12.7%) 25 (14.9%) 18 (10.8%) 167>25Kg/m2 211 (56.3%) 6 (16.4%) 58 (15.6%) 44 (11.7%) 375

Last dental visitsLast 2years 102 (52.2%) 41 (21.2%) 30 (15.4%) 22 (11.2%) 1952–5years 62 (54.4%) 19 (16.7%) 18 (16.1%) 15 (12.8%) 114>5years 117 (50.2%) 41 (17.8%) 43 (18.3%) 32 (13.7%) 233

Diabetes*Yes 15 (46.8%) 6 (18.8%) 6 (18.8%) 5 (15.6%) 32No 282 (55.3%) 90 (17.6%) 75 (14.7%) 63 (12.4%) 510

Smoking status*Non-smokers 112 (43.6%) 49 (18.8%) 74 (28.7%) 23 (8.9%) 258Former smokers 21 (17.0%) 25 (19.6%) 41 (32.8%) 38 (30.6%) 125Smokers 35 (21.9%) 19 (12.2%) 38 (23.6%) 67 (42.3%) 159

Mean alcoholuse time (years)†

NA 26.3 (±9.7) 27.3 (±10.4) 28.6 (±9.7) 28.3 (±10.5)

*Chi-squared test (p < 0.001);†ANOVA (p < 0.05); NA, not applicable; BMS, Brazilian minimum salary (approximately $258.00 Euros).

Table 2. Periodontal variables according to alcohol consumption groups

Alcohol consumption No/occasionalalcohol use (NA)

Moderate alcoholuse (MA)

Intensive alcoholuse (IA)

Alcohol dependence(DA)

Plaque score (%)* 52.2 ± 21.24 58.3 ± 22.6 61.4 ± 23.7 64.9 ± 23.9Mean BOP† 10.3 ± 13.2 12.3 ± 8.9 12.9 ± 11.3 14.2 ± 13.8Mean PD (mm)† 2.5 ± 0.7 2.7 ± 0.6 2.9 ± 0.8 3.1 ± 0.9Mean CAL (mm)† 2.9 ± 0.5 3.1 ± 1.3 3.3 ± 1.7 3.8 ± 1.2Sites with PD � 5 mm (%)† 4.2 ± 7.8 6.5 ± 10.4 7.3 ± 9.9 9.6 ± 12.8Sites with CAL � 5 mm (%)† 11.2 ± 13.9 15.2 ± 19.7 19.3 ± 21.2 25.8 ± 24.7Lost teeth† 1.68 ± 2.81 1.75 ± 3.27 2.04 ± 3.30 2.53 ± 3.21

*Significant differences between alcohol groups at p < 0.05, using ANOVA and Bonferroni post hoc analysis.†Significant differences between alcohol groups at p < 0.001, using Welch test and Tamhane post hoc analysis.BOP, bleeding on probing; PD, probing depth; CAL, clinical attachment level.


118 Lages et al.

OR = 3.43, CI 1.04–3.98; IA group:OR = 5.48, CI 1.97–6.02; DAgroup: OR = 7.91, CI 2.96–12.18)and lost teeth (OR = 2.82, CI 1.07–5.78), (2) Former smokers: alcoholuse (MA group: OR = 2.82, CI 1.21–2.72; IA group: OR = 3.75, CI 1.21–4.13; DA group: OR = 4.12, CI 1.63–5.91) and lost teeth (OR = 2.73, CI1.02–3.82) and (3) Non-Smokers:alcohol use (MA group: OR = 1.22,CI 1.02–1.63; IA group: OR =2.17, CI 1.03–2.21; DA group:OR = 3.02, CI 1.51–5.96) and lostteeth (OR = 1.14, CI 1.01–1.98).These findings emphasize a stronginteraction effect of smoking-alcoholuse, as in all groups of alcohol usersOR estimates for the occurrence ofperiodontitis were approximately twotimes higher in smokers comparedwith non-smokers (Table 4).

Gender, age, plaque index, BMI,diabetes, educational level, family

income <5 BMS, cohabitation statusand last dental visit were variablesthat were not significantly associatedwith periodontitis in multivariateregression models (Table 4).

Discussion

The present study investigated theassociation between the frequency ofalcohol consumption and periodonti-tis and confirmed the hypothesis thathigher alcohol consumption wasassociated with prevalence and sever-ity of periodontitis. It is noteworthythat these findings were still morepronounced in smokers. Conflictingfindings regarding the associationbetween alcohol consumption andperiodontitis were reported in the lit-erature, being that some studiesreported a positive association(Novacek et al. 1995, Enberg et al.2001, Khocht et al. 2003, Amaral

et al. 2008, 2009) while othersreported no association (Torrungru-ang et al. 2005, Bouchard et al.2006, Jansson 2008). To our knowl-edge, few studies investigated theassociation between DA and peri-odontal disease (Novacek et al.1995, Enberg et al. 2001, Khochtet al. 2003, Amaral et al. 2008).

Biological mechanisms explainingthe effects of alcohol in periodontaltissues has been discussed (Tezalet al. 2001, Pitiphat et al. 2003).Alcohol can affect the host response;impair neutrophil, macrophage andT-cell functions; as well as increasethe frequency of infections. Further-more, it has toxic effects on the liverand interferes with protein meta-bolism. Ethanol stimulates boneresorption in vitro, suppresses boneturnover, and may have a directtoxic effect on periodontal tissues(Tezal et al. 2001). Excessive and

Table 3. Distribution of independent variables by periodontitis status

Variables Periodontitiscases (n = 134)

Periodontitisnon-cases (n = 408)

Crude OR (95% CI) p*

n % n %

Alcohol statusNo/occasional (NA) 51 17.2 246 82.8 – –Moderate (MA) 23 24.0 73 76.0 1.52 (1.07–2.65) 0.042Intensive (IA) 24 29.6 57 70.4 2.03 (1.15–3.57) 0.012Alcohol dependence (DA) 36 53.0 32 47.0 4.72 (2.73–7.92) <0.001

GenderMale 56 19.2 236 80.8 0.51 (0.38–0.77) 0.001Female 78 31.2 172 68.8

Age� 35–45years 64 25.6 186 74.4 1.09 (0.88–1.93) 0.663� 45–55years 70 24.0 222 76.0

Family income<5BSM 60 26.1 170 73.9 1.99 (1.29–3.05) <0.001� 5BSM 47 15.1 265 84.9

Educational level<8years 84 27.1 226 72.9 1.35 (0.96–2.12) 0.069� 8years 50 21.6 182 78.4

Co-habitation statusWithout companion 64 37.6 106 62.4 2.60 (0.73–3.9) <0.001With companion 70 18.8 302 81.2

Body mass index� 25Kg/m2 55 33 112 67.0 0.54 (0.36–0.81) 0.002>25Kg/m2 79 21.1 296 78.9

Last dental visitLast 2years 32 16.4 163 83.6 –2–5years 24 21.1 90 78.9 1.36 (0.75–2.44) 0.156>5years 78 33.5 155 66.5 1.34 (0.85–2.11 0.101

Diabetes (yes) 9 28.2 23 71.8 1.02 (0.54–2.67) 0.318SmokingNon-smokers 21 12.9 237 87.1 – –Former smokers 29 23.8 96 74.2 3.49 (1.85–6.27) <0.001Smokers 56 35.2 103 64.8 6.15 (3.53–10.61) <0.001

*Chi-square test.OR, odds ratio; CI, confidence intervals.



constant alcohol use may affect thehost response to infections caused bybacteria, thus increasing host vulner-ability (Amaral et al. 2008).

Prevalence rates of periodontitisin the present study were high(17.2% to 53%), as well as alcoholconsumption OR estimates for theoccurrence of periodontitis. Similar(Sakki et al. 1995, Tezal et al. 2001,2004, Amaral et al. 2008) and differ-ent (Shimazaki et al. 2005, Bouchardet al. 2006, Jansson 2008) resultswere reported. In the present study,findings reinforce the effect of thefrequency of alcohol consumption inthe occurrence of periodontitis.

It is important to highlight thatmethodological issues may have sig-nificantly influenced the conflictingresults of the association betweenalcohol and periodontitis presentedin the literature. Some studies pre-sented small samples and great vari-ability in the definition of alcoholdose-exposure, including differentcut-off points for the amount andfrequency of alcohol consumption(Torrungruang et al. 2005, Bouchardet al. 2006, Jansson 2008), while oth-ers presented less robust definitionsof periodontitis (Shizukuishi et al.1998, Okomoto et al. 2006).

Methods of measurement of alco-hol exposure levels and dependenceare complex and prone to informa-tion biases. However, the use of

AUDIT and CAGE questionnairescan minimize these biases. AUDITwas developed under the supervisionof WHO and it is composed of 10questions (three questions about theconsumption of alcoholic beverages,four questions about alcoholic bever-ages dependence, and three questionsabout problems related to alcoholicbeverages (Saunders et al.1993).CAGE is validated in Brazil and it iscomposed by four questions. CAGEpresents a specificity of 83% and asensitivity of 88% when the cut-offpoint of two positive answers isadopted to define alcoholic bever-ages (Masur & Monteiro1983, Ama-ral et al. 2008).

Sample of the present study com-prised a large number of subjects inan age group of 35–55 years. Somestudies demonstrated age as riskindicator associated to periodontitis(Do et al. 2003, Amaral et al. 2008).In this manner, the strategy of agerestriction in the sample of the pres-ent study could reduce the varianceof measurements and minimize theimpact of age as a confounder in theassociation with other variables ofinterest. Nevertheless, age was notassociated with periodontitis in thefinal logistic models.

Periodontitis case definition adoptedin the present study can be consideredrobust. This definition was used toavoid underestimation of periodonti-

tis prevalence. It has been recom-mend to be used in associationstudies (Lopez et al. 2002). In addi-tion to the higher prevalence of peri-odontitis observed among alcoholusers groups, worse periodontalparameters BOP, PD and CAL werealso observed in these subjects.These findings were also reported inprevious studies (Tezal et al. 2001,2004, Amaral et al. 2008, 2009, Kon-gstad et al. 2008).

Results from the present studyshowed a high plaque index, varyingfrom ~50% to 65%, among subjectsin the global sample. However, dentalbiofilm considered to be a primaryaetiological factor for periodontitis(Davenport et al. 1998, Albandar &Rams 2002, Costa et al. 2009) wasnot a significant variable in the finalmultivariate model. Hence, it can berationalized that dental biofilm is anecessary factor for the developmentof periodontal diseases, but variationsin biofilm accumulation during a per-iod of time has a less important rolein the variation of periodontitis path-ogenesis (Grossi et al. 1994). Similarfindings were also reported when theeffect of the variable plaque indexwas controlled in association studiesbetween alcohol use and periodontitis(Novacek et al. 1995, Tezal et al.2001, Amaral et al. 2008).

Periodontitis has been pointed asthe main cause of tooth mortality

Table 4. Final logistic regression models for periodontitis by smoking status

Variables Smokers (n = 159)* Former smokers (n = 125)** Non-smokers (n = 258)***

OR 95% CI p OR 95% CI p OR 95% CI p

Alcohol useModerate (MA) 3.43 1.04–3.98 <0.001 2.82 1.21–2.72 0.028 1.22 1.02–1.63 0.032Intensive (IA) 5.48 1.97–6.02 <0.001 3.75 1.21–4.13 0.012 2.17 1.03–2.21 0.031Alcohol dependence (DA) 7.91 2.96–12.18 <0.001 4.12 1.63–5.91 <0.001 3.02 1.51–5.96 0.026

Age� 45–55years 1.24 0.91–2.63 0.071 1.42 0.71–1.96 0.132 0.93 0.52–1.79 0.163

Gender*Male 1.56 0.38–2.72 0.283 1.27 0.41–2.12 0.285 1.02 0.23–1.92 0.452

Family income*<5BSM 1.73 0.78–3.23 0.042 0.96 0.74–3.21 0.062 1.01 0.91–2.13 0.067

Educational level*<8years 0.96 0.18–3.71 0.078 0.85 0.23–2.91 0.083 0.627 0.14–2.02 0.074

Diabetes (Yes) 1.14 0.92–3.96 0.061 0.83 0.23–3.18 0.382 0.94 0.25–3.21 0.071Lost Teeth 2.82 1.07–5.78 <0.001 2.73 1.02–3.82 <0.001 1.14 1.01–1.98 0.022

Statistically significant are show in bold.Dependent variable: occurrence of periodontitis.Area under receiver operating curve (ROC) for the model:*0.756;**0.719;***0.703.OR, odds ratio; CI, confidence intervals.


120 Lages et al.

after 45 years of age (Costa et al.2009) and it has been associated tostrong negative impacts in the qual-ity of life of periodontally susceptiblesubjects. Therefore, the number oflost teeth has been considered astrong risk indicator for periodonti-tis (Copeland et al. 2004, Costaet al. 2011). The mean number oflost teeth was low in the presentstudy. It is important to emphasizethat the inclusion criteria of havinga minimum of 14 teeth could haveinfluenced these results. However,this variable was significantly associ-ated with the occurrence of peri-odontitis in all groups of alcoholusers, especially in smokers (OR =2.82; CI 1.07–5.78).

Prevalence of alcohol users waslower among women. A low alcoholconsumption rate among Brazilianwomen due to social and behaviouralreasons was reported (Amaral et al.2008). However, an increase inalcoholic beverage drinking amongwomen, especially among those withyoung ages, was also reported (Ama-ral et al. 2009). Gender as a variablenot significant in the multivariatefinal models reinforces the strengthof the association between alcoholuse and periodontitis when explana-tory factors and confounders arecontrolled. Similar findings were alsoreported by Do et al. 2003 in a sam-ple of the Vietnam population.

A great number of studies havedemonstrated that smoking isstrongly related to the risk of peri-odontal attachment loss, as well asbone and tooth loss (Labriola et al.2005, Heasman et al. 2006). More-over, studies showed incremental ORestimates for the occurrence of peri-odontitis when smoking dose-expo-sure was evaluated (Grossi et al.1994, Tomar & Asma 2000; Torrun-gruang et al. 2005). Studies havereported a high prevalence of smok-ers among alcohol users (Tezal et al.2001, Amaral et al. 2008, Kongstadet al. 2008). Findings from the pres-ent study confirmed this statementshowing a higher prevalence ofsmokers and a strong associationbetween smoking and periodontitis.It was also demonstrated in the pres-ent study that the associationbetween periodontitis and alcoholuse was independent of smoking(non-smokers OR ranged from 1.22to 3.02). Similar findings were also

reported (Shizukuishi et al. 1998,Tezal et al. 2001, Copeland et al.2004, Torrungruang et al. 2005,Amaral et al. 2008).

The interactions between alcoholconsumption frequency and smokingin the occurrence of periodontitisshowed higher OR estimates, espe-cially for current smokers in the DAgroup. However, all groups ofalcohol users OR estimates for theoccurrence of periodontitis wereapproximately two times higher insmokers (OR = 3.43 to 7.91) com-pared with non-smokers (OR = 1.22to 3.02).Thus, these results of smok-ing-alcohol interaction can beexplained by a potentiated effect ofrisk variables that affect the immuneresponse and strongly influence thehost susceptibility. In addition, theeffect of smoking-alcohol interactionin the occurrence of periodontitisalso emphasizes the influence ofbehavioural factors in the pathogen-esis of periodontitis. In contrast, ina cross-sectional study by Tezalet al. 2001, no significant interactionwas found implying that the rela-tionship between alcohol consump-tion and periodontal diseasechanges with smoking. However,these authors reinforced the need ofevaluating interaction in such associ-ation studies.

Studies on the associationbetween periodontitis and some riskindicators such as diabetes, BMI,family income, educational level andfrequency of dental visits arereported in dental literature withconflicting findings (Albandar &Rams 2002; Bouchard et al. 2006,Naoki et al. 2008, Costa et al. 2011).In the present study, these risk indi-cators were not significantly associ-ated with the occurrence ofperiodontitis in the multivariate finallogistic models. Although controver-sial, socio-economic and cultural sta-tus can have an effect on subjects’behaviour and facilitate the incorpo-ration of noxious behavioural habitsin relation to general health (Coul-son et al. 2010, Casswell 2011).

Some limitations of the presentstudy should be pointed out. Althoughsignificant and well founded, thesample cannot be considered repre-sentative of the Brazilian population.Moreover, the cross-sectional designdoes not allow temporal inferencesregarding the abusive alcohol use

and periodontitis. In this manner,further studies with representativesamples and prospective design arerequired.

The present study provided someevidence of the high frequency ofalcohol consumption as a risk indi-cator for periodontitis. Preventiveand educational strategies withingeneral health services should giveattention to the hazardous effects ofalcohol consumption in periodontalstatus. It was concluded that theprevalence of periodontitis amongalcohol users was high and that thefrequency of alcohol consumptionand DA increased the odds of peri-odontitis incrementally, mainly insmokers.

References

Albandar, J. M. (2002) Periodontal diseases inNorth America. Periodontology 2000 29, 31–69.

Albandar, J. M. & Rams, T. E. (2002) Global epi-demiology of periodontal disease: an overview.Periodontology 2000 29, 7–10.

Amaral, C. S., Luiz, R. R. & Leao, A. (2008) Therelationship between alcohol dependence andperiodontal disease. Journal of Periodontology79, 993–998.

Amaral, C. S., Vettore, M. V. & Leao, A. (2009)The relationship of alcohol dependence andalcohol consumption with periodontitis: a sys-tematic review. Journal of Dentistry 37, 643–651.

American Diabetes Association (2005) Diagnosisand classification of diabetes mellitus. Positionstatement. Diabetes Care 1(Suppl.1), 37–42.

Bouchard, P., Boutouyrie, P., Mattout, C. &Bourgeois, D. (2006) Risk assessment for severeclinical attachment loss in an adult population.Journal of Periodontology 77, 479–89.

Casswell, S. (2011) Health and societal effects ofalcohol. Lancet 377, 463–464.

Copeland, L. B., Krall, E. A., Brown, L. J.,Garcia, R. I. & Streckfus, C. F. (2004)Predictors of tooth loss in two US adult popu-lations. Journal of Public Health Dentistry 64,31–37.

Costa, F. O., Cota, L. O. M., Lages, E. J. P., Lo-rentz, T. C. M., Oliveira, A. M. S. D., Oliveira,P. A. D. & Costa, J. E. (2011) Progression ofperiodontitis in a sample of regular and irregu-lar compliers under maintenance therapy: a3-year follow-up study. Journal of Periodontol-ogy 82, 1279–1287.

Costa, F. O., Guimaraes, A. N., Cota, L. O. M.,Pataro, A. L., Takeshi, K. S., Cortelli, S. C. &COSTA, J. E. (2009) Impact of different peri-odontitis case definitions on periodontalresearch. Journal of Oral Science 51, 199–206.

Coulson, C. E., Williams, L. J., Henry, M. J.,Berk, M., Lubman, D. I., Brennan, S. L., Nich-olson, G. C., Kotowicz, M. A., Korn, S. &Pasco, J. A. (2010) Patterns of alcohol use andassociated physical and lifestyle characteristicsaccording to new Australian guidelines. Austra-lian and New Zealand Journal of Psychiatry 44,946–951.

Davenport, E. S., Williams, C. E., Sterne, J. A.,Sivapathasundram, V., Fearne, J. M. & Curtis,



M. A. (1998) The east London study of mater-nal chronic periodontal disease and pretermlow birth weight infants: study design andprevalence data. Annals of Periodontology 3,213–221.

Do, G. L., Spencer, A. J., Roberts-Thomson, K.& Ha, H. D. (2003) Smoking as a risk indica-tor for periodontal disease in the midldle-agedVietnamese population. Community Dentistryof Oral epidemiology 31, 437–446.

Enberg, N., Wolf, J., Ainamo, A., Alho, H., Hei-nala, P. & Lenander-Lumikari, M. (2001) Den-tal diseases and loss of teeth in a group ofFinnish alcoholics: a radiological study. Actaodontologica Scandinavica 59, 341–347.

Ewing, J. A. (1984) ‘Detecting Alcoholism: theCAGE Questionnaire. Journal of the AmericanMedical Association 252, 1905–1907.

Grossi, S. G., Zambon, J. J., Ho, A. W., Koch,G., Dunford, R. G., Machtei, E. E., Norderyd,O. M. & Genco, R. J. (1994) Assessment ofrisk for periodontal disease. I. Risk indicatorsfor attachment loss. Journal of Periodontology65, 260–267.

Heasman, L., Stacey, F., Preshaw, P. M., McC-racken, G. I., Hepburn, S. & Heasman, P. A.(2006) The effect of smoking on periodontaltreatment response: a review of clinical evi-dence. Journal of Clinical Periodontology 33,241–253.

Jansson, L. (2008) Association between alcoholconsumptions and dental health. Journal ofClinical Periodontology 35, 379–384.

Khocht, A., Janal, M., Schleifer, S. & Keller, S.(2003) The influence of gingival marginrecession on loss of clinical attachment inalcohol-dependent patients without medicaldisorders. Journal of Periodontology 74, 485–493.

Kongstad, J., Hvidtfeldt, U. A., Gronbaek, M.,Stoltze, K. & Holmstrup, P. (2008) Amountand type of alcohol and periodontitis in theCopenhagen City Heart Study. Journal of Clini-cal Periodontology 35, 1032–1039.

Labriola, A., Needleman, I. & Moles, D. R.(2005) Systematic review of the effect of smok-ing on nonsurgical periodontal therapy. Perio-dontology 2000 37, 124–137.

Lopez, N. J., Smith, P. C. & Gutierrez, J. (2002)Periodontal therapy may reduce the risk of pre-term low birth weight in women with periodon-tal disease: a randomized controlled trial.Journal of Periodontology 8, 911–924.

Masur, J. & Monteiro, M. G. (1983) Validationof the “CAGE” alcoholism screening test in aBrazilian psychiatric inpatient hospital setting.Brazilian Journal of Medical and BiologicalResearch 16, 215–218.

Mayfield, D., McLeod, G. & Hall, P. (1974) TheCAGE questionnaire: validation of a new alco-holism screening instrument. American Journalof Psychiatry 131, 1121–1123.

Mendoza-Sassi, R. A. & Beria, J. U. (2003) Prev-alence of alcohol use disorders and associatedfactors: a population-based study usingAUDIT in southern Brazil. Addiction 98, 799–804.

Naoki, N., Morita, M. & Kawanami, M. (2008)Oral health Care-specific self-efficacy assess-ment predicts patients’ completion of periodon-tal treatment: a pilot study. Journal ofPeriodontology 79, 1041–1047.

Nishida, N., Tanaka, M., Hayashi, N., Nagata,H., Takeshita, T., Nakayama, K., Morimoto,K. & Shizukuishi, S. (2004) Association ofALDH(2) genotypes and alcohol consumptionwith periodontitis. Journal of Dental Research83, 161–165.

Novacek, G., Plachetzky, U., Potzi, R., Lentner,S., Slavicek, R., Gangl, A. & Ferenci, P. (1995)Dental and periodontal disease in patients withcirrhosis – Role of etiology of liver disease.Journal of Hepatology 22, 576–582.

Okomoto, Y., Tsuboi, S., Suzuki, S., Nakagaki,H., Ogura, Y., Maeda, K. & Tokudome, S.(2006) Effects of smoking and drinking habitson the incidence of periodontal disease andtooth loss among Japanese males. Journal ofPeriodontal Research 41, 560–566.

Pitiphat, W., Merchant, A. T., Rimm, E. B. &Joshipura, K. J. (2003) Alcohol consumptionincreases periodontitis risk. Journal of DentalResearch 82, 509–513.

Sakki, T. K., Kunuuttila, M. L., Vimpari, S. S. &Hartkainen, M. S. (1995) Association of life-style with periodontal health. Community Den-tistry and Oral Epidemiology 23, 155–158.

Sarfati, A., Bourgeois, D., Katsahian, A. S.,Mora, F. & Bouchard, P. (2010) Risk assess-ment for buccal gingival recession defects in anadult population. Journal of Periodontology 81,1419–1425.

Saunders, J. B., Aasland, O. G., Babor, T. F., dela Fuente, J. R. & Grant, M. (1993) Develop-ment of the alcohol use disorders identificationtest (AUDIT): WHO collaborative project on

early detection of persons with harmful alcoholconsumption – II. Addiction 88, 791–804.

Shimazaki, Y., Saito, T., Kiyohara, Y., Kato, l.,Kubo, M., Iida, M. & Yamashita, Y. (2005)Relationship between drinking and periodonti-tis: the Hisayama study. Journal of Periodonto-logy 76, 1534–1541.

Shizukuishi, S., Hayashi, N., Tamagawa, H.,Hanioka, T., Maruyama, S., Takeshita, T. &Morimoto, K. (1998) Lifestyle and periodontalhealth status of Japanese factory workers.Annals of Periodontology 3, 303–11.

Silness, J. & Loe, H. (1964) Periodontal disease inpregnancy. II. Correlation between oral hygieneand periodontal condition. Acta OdontologicaScandinavica 22, 121–135.

Tezal, M., Grossi, S. G., Ho, A. W. & Genco, R.J. (2001) The effect of alcohol consumption onperiodontal disease. Journal of Periodontology72, 183–189.

Tezal, M., Grossi, S. G., Ho, A. W. & Genco, R.J. (2004) Alcohol consumption and periodontaldisease. The third national health and nutritionexamination survey. Journal of Clinical Peri-odontology 31, 484–488.

Tomar, S. L. & Asma, S. (2000) Smoking-attrib-utable periodontitis in the United States: find-ings from NHANES III. National Health andNutrition Examination Survey. Journal of Perio-dontology 71, 743–751.

Torrungruang, K., Tamsailom, S., Rojanasoms-ith, K., Sutdhibhisal, S., Nisapakulton, K.,Vanichjakvong, O., Prapakamol, S., Premsiri-nirund, T., Pusiri, T., Jaratkulangkoon, O.,Unkurapinun, N. & Sritara, P. (2005) Riskindicators of periodontal disease in older Thaiadults. Journal of Periodontology 76, 558–65.

Address:Fernando O. CostaDepartment of PeriodontologyFederal University of Minas GeraisAntonio Carlos Avenue, 6627, Pampulha -PO Box 359, Zip Code 31270-901Belo HorizonteMG – BrazilE-mail: [email protected]

Clinical Relevance

Scientific rationale for the study:Data on the association betweenalcohol consumption and periodon-titis are controversial. In addition,few studies evaluated this associa-tion among subjects with alcoholdependence.

Principal findings: Prevalence of peri-odontitis among alcohol users werehigh (24–53%). Risk variables suchas alcohol use frequency, smokingand number of lost teeth were signif-icantly associated with the occur-rence of periodontitis.Practical implications: Increased fre-quency of alcohol consumption was

associated with increased odds ofhaving periodontitis, mainly insmokers. The association of peri-odontitis and increased alcohol use,especially in smokers, should beconsidered when developing pre-ventive strategies in general healthservices.


122 Lages et al.

risk variables in the association between frequency of alcohol consumption and periodontitis

Documents