| Sign In to gain access to subscriptions and/or personal tools. |
Alcohol Consumption Increases Periodontitis Risk
1 Department of Oral Health Policy & Epidemiology, Harvard School of Dental Medicine, 188 Longwood Avenue, Boston, MA 02115, USA; Correspondence: *corresponding author, waranuch{at}post.harvard.edu
Alcohol consumption impairs neutrophil, macrophage, and T-cell functions, increasing the likelihood of infections. We examined the association between alcohol consumption and periodontitis, prospectively, among 39,461 male health professionals aged 40 to 75 years and free of periodontitis at the start of follow-up. Alcohol intake was assessed at baseline and updated every 4 years by a food-frequency questionnaire. Periodontal disease status was self-reported and validated against radiographs. Multivariate analysis was adjusted for age, smoking, diabetes, body-mass index, physical activity, time period, and caloric intake. During 406,160 person-years of follow-up, there were 2125 cases of periodontitis. Compared with non-drinkers, the relative risk (95% confidence interval) among men reporting usual alcohol intake of 0.1-4.9 g/day was 1.24 (1.09, 1.42); 5.0 to 14.9 g/day, 1.18 (1.04, 1.35); 15 to 29.9 g/day, 1.18 (1.01, 1.38); and > 30 g/day, 1.27 (1.08, 1.49). The results suggest that alcohol consumption is an independent modifiable risk factor for periodontitis.
Key Words: alcohol drinking • epidemiology • periodontal diseases • periodontitis
Alcohol impairs neutrophil, macrophage, and T-cell functions, increasing the likelihood of infections (Szabo, 1999), possibly raising the risk of periodontitis. Despite the plausible mechanisms, information relating alcohol consumption to periodontitis risk is sparse. Previous cross-sectional (Sakki et al., 1995; Shizukuishi et al., 1998; Tezal et al., 2001) and case-control (Pan et al., 1998) studies have shown positive associations between alcohol use and periodontal disease; however, prospective data are not yet available. Furthermore, only one study has assessed the effects of different types of alcohol on the risk of periodontal disease (Tezal et al., 2001). We therefore examined prospectively the association between alcohol consumption and periodontitis among men who participated in the Health Professionals Follow-up Study (HPFS).
Study Population The HPFS is a prospective study of 51,529 male health professionals aged 40-75 years in 1986. The cohort included dentists (58%), veterinarians (20%), pharmacists (8%), optometrists (7%), osteopathic physicians (4%), and podiatrists (3%). Incident diseases and updated exposures were ascertained with biennial questionnaires. Responses to the questionnaires constituted informed consent to a protocol that was approved by the Institutional Review Board at Harvard School of Public Health. We excluded men who were deceased (n = 4), reported periodontitis (n = 8955), reported myocardial infarction or stroke (n = 1884), or provided inadequate dietary information (n = 1225) in 1986, leaving 39,461 men eligible for follow-up.
Assessment of Alcohol Consumption We evaluated the validity of the FFQ in a random sample of 136 men living in the Boston area (Giovannucci et al., 1991). Intake of alcohol reported over the previous year by the FFQ correlated highly with intake assessed by diet records completed over this period (Spearman r = 0.86, p < 0.001).
Assessment of Periodontitis
Statistical Analysis We used multivariate pooled logistic regression (D’Agostino et al., 1990) with two-year time intervals to approximate the Cox proportional hazards model. For the primary analyses, we modeled periodontitis risk and cumulatively averaged (Hu et al., 1999) alcohol consumption. In this analysis, if a person had angina, coronary artery bypass graft surgery, myocardial infarction, stroke, cancer, or asthma, we stopped updating his alcohol intake, because he might have changed consumption as a result of the event, and it may not reflect long-term intake. In additional analyses, we related incidence of periodontitis to intake of alcohol at baseline and to the most recent intake. The multivariate models adjusted for age, time period, smoking, diabetes, body mass index (BMI), physical activity (metabolic equivalents/wk), and total calories. Time-varying covariates including age, smoking, diabetes, physical activity, BMI, and total calories were updated every 2 yrs, because most recent status may be more relevant to the disease. We updated physical activity by using the cumulative average of activities during the period of follow-up to best represent long-term physical activity levels of individuals, and it reduced measurement error (Hu et al., 1999). We adjusted for energy as a surrogate measure of metabolic efficiency and the thermogenic effects of foods, which may be a potential source of residual confounding. The presence of a linear trend in relative risk (RR) across alcohol categories was tested with the medians within each category as an ordinal variable. We also conducted analyses separately among non-smokers and among participants who reported unchanged drinking habits during follow-up. To examine the presence of interactions, we performed the analyses stratified by age, smoking, and BMI. We used likelihood ratio tests to compare models with and without the interaction terms. All reported p-values are two-sided.
In this cohort, most participants (52%) had low-to-moderate alcohol consumption (0.1-14.9 g/day). Compared with men who reported drinking no alcohol, men who reported any regular alcohol intake were more likely to be smokers, were more physically active, consumed more calories, and were less likely to be diabetic (Table 1  ).
During 406,160 person-years of follow-up, 2125 participants reported periodontitis for the first time. Crude incidence of periodontitis was 4.3 per 1000 person-years among non-drinkers and varied from 5.2 to 6.9 per 1000 person-years among drinkers (Table 2 ). After adjustment for age, men who drank alcohol were at higher risk of periodontal disease compared with non-drinkers. Further adjustment for smoking slightly attenuated this association. There was a positive association between alcohol intake and periodontitis across all categories of intake after simultaneous adjustment for age, smoking, diabetes, BMI, physical activity, and total calories. Compared with non-drinkers, the multivariate RR among men reporting usual alcohol intake of 0.1-4.9 g/day was 1.24 (95% confidence interval [CI], 1.09, 1.42); and for > 30 g/day, 1.27 (95% CI, 1.08, 1.49). Further analyses restricted to never-smokers (800 cases) and to participants who reported unchanged drinking habits (1823 cases) provided similar results (data not shown). We observed no evidence of effect modification by age, smoking, and BMI (data not shown).
Alcohol intake assessed as a cumulative average, or as a single baseline measurement, had similar associations with periodontitis risk (Fig. ). The association between the most recent alcohol intake and periodontitis risk was weaker.
At baseline, total alcohol intake was associated with a significantly increased risk of periodontitis across all levels of intake (p for trend = 0.03) (Fig.  ). When we analyzed alcohol intake by the number of drinks of beer, white wine, red wine, or liquor (each adjusted for other types of alcohol), there was no clear association with periodontitis risk (Table 3). There was a modest inverse association with periodontitis risk in the third category of beer consumption, but the trend was not significant. Men who drank 2 or more glasses of red wine a day were at increased risk for periodontitis, RR = 1.50 (95% CI, 0.98, 2.30, p for trend = 0.05). White wine and liquor consumption had no appreciable effect on periodontitis. We did not find a significant difference in the test for trend between red wine and any of the other beverages.
In this large prospective study, we found a positive association between alcohol intake and periodontitis. Men who drank alcohol had an 18-27% higher risk of disease than did non-drinkers. These results were similar when baseline alcohol intake was used alone, when this measure was updated every 2 yrs with most recent reported information, or when the average intake over the follow-up period was used in the analysis. Recent intake of alcohol had the weakest association, as one may expect, since it covers the shortest induction period. Few studies have examined the possible relation between alcohol intake and periodontitis. Early studies observed increased prevalence and severity of periodontal disease among patients with cirrhosis (Sandler and Stahl, 1960; Movin, 1981), and attributed this to poor oral hygiene (Movin, 1981). Other studies reported worse periodontal conditions in alcoholic patients with and without cirrhosis than in healthy subjects (Dunkley and Carson, 1968; Novacek et al., 1995) and in non-alcoholic patients with cirrhosis (Novacek et al., 1995). There was a significant association between alcohol consumption and periodontal disease among Japanese factory workers, but only in the bivariate analysis (Shizukuishi et al., 1998). In a small study among dental patients, periodontal disease was positively associated with indicators of alcoholism among males only, but there were only 25 female participants (Kranzler et al., 1990). A cross-sectional study of 780 Finnish men and women showed an odds ratio (OR) of 1.76 among participants who drank < 7 drinks per 2 wks, and 2.52 among those who drank > 7 drinks per 2 wks in comparison with non-drinkers, controlling for dietary habits, smoking habits, and toothbrushing frequency (Sakki et al., 1995). In a case-control study in China, drinkers were 1.86 times more likely to have periodontitis than were non-drinkers (unadjusted analysis) (Pan et al., 1998). Recent findings from the Erie County Study also showed a positive relationship between alcohol consumption and more severe attachment loss and gingival bleeding (Tezal et al., 2001). Alcohol consumption of > 5 drinks/wk was associated with increased attachment loss, OR of 1.36 (95% CI, 1.02, 1.80), compared with consumption of < 5 drinks/wk. The OR was modestly stronger (OR = 1.44; 95% CI, 1.04, 2.00) when 10 drinks/wk were used as the threshold. In the same study, wine, beer, or liquor intakes had similar associations with periodontitis risk. We did not observe any clear pattern of association between specific beverages and periodontitis risk. High red wine intake raised the risk of periodontitis slightly more than that of the other beverages, but the result was not significant. We were somewhat limited in the beverage-specific analyses due to the limited number of cases in the heaviest drinkers; thus, the risk estimates should be interpreted with caution. Previous studies reported a J-shaped relationship between alcohol consumption and all-cause mortality (Camargo et al., 1997), driven by a reduction in risk of cardiovascular disease from moderate drinking and raised risk of cancer deaths from heavy drinking. We observed an increased risk of periodontitis with drinking any amount of alcohol. The difference in the results is due to substantially different hypothesized mechanisms to explain the association of alcohol with mortality, and with periodontitis.
Several plausible biological explanations exist for a detrimental effect of alcohol on periodontitis risk. Studies have shown that impaired neutrophil phagocytosis is associated with periodontal disease (Hart et al., 1994; Van Dyke and Vaikuntam, 1994). Alcohol impairs neutrophil function, contributing to bacterial overgrowth and increased bacterial penetration (Szabo, 1999) that may lead to periodontal inflammation. Second, evidence from in vitro (Cheung et al., 1995), animal (Farley et al., 1985; Turner et al., 2001), and human (Pepersack et al., 1992) studies suggests that alcohol may stimulate bone resorption and suppress bone turnover. Third, alcohol may have a direct toxic effect on periodontium as with other tissues of the oropharynx (Maier et al., 1994; Ogden et al., 1999). Finally, moderate alcohol intake reduces monocyte production of inflammatory cytokines such as TNF- Alcohol drinking may be associated with poor oral hygiene practices (Sakki et al., 1995), possibly raising periodontitis risk. Although we did not collect information on oral hygiene in the whole cohort, analysis of data from a sample of 152 men suggests that this population of health professionals had good oral hygiene. There was no significant association between oral hygiene practices and periodontal disease in this population (Merchant et al., 2002), as well as in other studies (Badersten et al., 1990; Machtei et al., 1993; AAP, 1996). Hence, oral hygiene is unlikely to confound the effect of alcohol on periodontitis in this cohort. This is the first prospective study to evaluate alcohol as a risk factor for periodontitis. The prospective design ensures temporality of the association and eliminates the possibility of recall bias. The high rate of follow-up reduced potential bias due to loss of follow-up. Men excluded due to inadequate dietary data were similar to those retained in the analysis with respect to age, smoking, physical activity, and BMI, so selection bias is unlikely. The participants are relatively homogeneous, thus minimizing confounding by race, socio-economic status, access to care, and oral hygiene practices. As with any observational study, we cannot exclude the possibility of residual confounding by other habits and lifestyle factors. Since smoking is an important risk factor for periodontitis and is correlated with alcohol drinking, some degree of the observed association may be due to residual confounding by smoking. In the analysis restricted to never-smokers, the results did not change substantially, indicating that residual confounding by smoking was unlikely. In the analysis excluding participants reporting substantial change in alcohol drinking habits (possibly because of health concerns), the results were similar to the main analyses. Another limitation includes the use of self-reports to assess the outcome. In such a large prospective study, it is impractical to perform clinical evaluation of periodontal disease. Our validation studies showed that self-reports of periodontitis (Joshipura et al., 1996, 2002) in the HPFS population were valid. Moreover, misclassification from self-reports tends to be random, resulting in an attenuated magnitude of association; with a perfect measure of periodontitis, the association would probably be stronger. In conclusion, the results support that alcohol drinking is an independent risk factor for periodontitis. Types of alcoholic beverages had no clear separate effect on periodontitis. Further research is needed to assess this association in other populations, and to determine the biological mechanisms of alcohol on periodontal disease. Health practitioners need to be aware that patients who drink may be at higher risk of periodontitis and could benefit from advice to quit smoking and maintain regular dental visits.
This research was supported by NIH Grants CA55075, HL35464, AA11181, and DE12102. Dr. Pitiphat is the recipient of the Royal Thai Government Scholarship. A preliminary report was presented at the 79th General Session of the International Association for Dental Research, June 27-30, 2001, Chiba, Japan. Received for publication August 15, 2002. Revision received February 20, 2003. Accepted for publication April 2, 2003.
Journal of Dental Research, Vol. 82, No. 7,
509-513 (2003) This article has been cited by other articles:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
, IL-1, and IL-6, possibly allowing for microbial proliferation (
