Smoking and drinking in relation to oral potentially malignant disorders in Puerto Rico: a case-control study

Smoking

Among the 86 subjects diagnosed with an OPMD, 52.3% reported ever smoking cigarettes while 4.7% and 2.3% had smoked cigars or pipes, respectively. Of the 155 subjects diagnosed with a benign oral tissue condition, the corresponding percentages were 36.1%, 2.6%, and 1.9%. None of the study subjects reported using smokeless tobacco.

Smoking level was defined in terms of average cigarette equivalents per day and pack-year equivalents. At each level of smoking, adjusted ORs were over twice as high for smokers relative to never smokers, and the tests of linear trend were statistically significant.

When we simultaneously included terms for both current smoking (current vs. noncurrent) and average cigarette-equivalents/day in the adjusted model, the OR for current smoking remained high (OR = 3.0, 95% CI: 1.28-7.03) while ORs for each smoking level were attenuated (i.e., adjusted ORs for average cigarette-equivalents/day declined from 1.0, 2.83, 2.21 (Table 2) to 1.0, 1.68, 1.23, P _trend = 0.70). When we conducted an analogous analysis that added current smoking and pack-year equivalents of smoking to the adjusted model, the OR for current smoking was 2.95 (95% CI: 1.24-7.03), and the ORs for pack-year equivalents of smoking declined from 1.0, 2.26, 2.83 (Table 2) to 1.0, 1.48, 1.45 (P _trend = 0.44).

Among ever smokers, the age at which smoking began did not differ significantly between the OPMD and benign conditions groups (OR = 0.81, 95% CI: 0.35-1.91).

Drinking

When we compared the mean number of alcoholic drinks consumed per week for ever drinkers, persons in the OPMD group had a higher mean total consumption compared to the benign conditions group but the difference did not approach statistical significance (18.8 vs. 16.2 drinks/wk, P = 0.68). On the other hand, among heavy drinkers (> 20 drinks/wk), the OPMD group consumed, on average, fewer drinks/wk than the benign conditions group (55.1 vs. 70.3 drinks/wk), but the difference was not statistically significant (P = 0.46). After stratifying on type of alcoholic beverage, consumption was marginally higher in the OPMD group (beer: 9.4 vs. 8.1; wine: 1.1 vs. 0.8; hard liquor: 8.3 vs. 7.4); but again, not statistically significant.

Smoking & Drinking Interaction

Additional Analyses

Point estimates for smoking and drinking did not change notably when we added variables for denture use and body mass index (BMI) to the adjusted model. In addition, adding lichen planus cases to the OPMD group had little impact on the adjusted ORs reported in Tables 2, 3 and 4. The adjusted ORs were also similar when age was entered into the model as a continuous, rather than categorical, variable.

Smoking

The favored form of tobacco use can vary across geographic areas and cultures around the world. In the United States, Europe, Australia, and Japan, cigarettes, cigars, and pipes are the major types of smoking tobacco, while chewing tobacco and snuff are the most common forms of smokeless tobacco [20]. Although the prevalence of tobacco smoking is high in India, Pakistan, China, and other areas of Asia, smokeless tobacco, used either alone or as part of a concoction and in forms including betel quid, bidis, paan, naswar, or nass, is also popular [20–23].

In our study, conducted in Puerto Rico, persons smoked tobacco primarily in the form of cigarettes, with few subjects in either the OPMD or benign conditions group claiming to smoke cigars or pipes. Although smokeless tobacco use has been implicated as a risk factor for both oral leukoplakia and oral cancer on the United States mainland [24–26], in our study, no subjects with an OPMD or a benign tissue condition reported using smokeless tobacco.

In the present study and after adjusting for drinking and other potentially confounding factors, current smoking was strongly associated with OPMD risk. There was a more than four-fold increased risk among current versus never smokers; however, the risk was notably attenuated among former smokers. These findings are consistent with numerous previous studies of oral leukoplakia [21, 27–32], OED [15, 33], erythroplakia [34], and oral cancer [16, 17, 35]. Our finding of a substantially decreased OPMD risk for former smokers relative to current smokers underscores the public health benefit associated with smoking cessation.

Previous studies have reported dose- and/or duration- response relationships between smoking and the risk of oral leukoplakia, erythroplakia, and OED [15, 21, 34, 36, 37]. In the current investigation, the risk of having an OPMD increased in a monotonic fashion with increasing pack-year equivalents of smoking. The linear trend for OPMD risk in relation to average cigarette equivalents smoked per day was also statistically significant, but did not increase in a stepwise fashion, perhaps due to the relatively small OPMD sample size. Of note, our findings suggest that current smoking status is a stronger risk factor for OPMDs than reported level of smoking.

Drinking

Alcohol consumption is an established risk factor for oral cancer [12, 16, 17, 35]. Nevertheless, findings from investigations into the association between alcohol consumption and either oral leukoplakia or other potentially premalignant oral diagnoses have not been consistent across studies. In the current investigation, we found little evidence that overall alcohol consumption is associated with increased OPMD risk. Our results are in keeping with several earlier studies reporting inverse or null findings for total alcohol consumption [21, 28, 29, 38, 39]; but not with others that reported a positive association for either leukoplakia in general, erythroplakia, or OED [15, 34, 40, 41].

Some studies of premalignant lesions and OED have found that risks associated with drinking are dependent upon the level of alcohol intake, with the highest risks observed for the highest level of alcohol intake [15, 34, 41]. After adjusting for various potential confounders; however, we found no evidence of an increased OMPD risk even among those persons who consumed > 20 drinks/wk.

Previous studies have reported that the risk of oral cancer and OPMDs can vary by beverage type, i.e., beer, wine, and hard liquor [15, 16, 39, 42–44]. In Puerto Rico, the consumption of hard liquor, particularly when used without a mixer, has been associated with a much stronger risk of oral cancer than beer or wine [44]. In our analysis of OPMD risk, however, we found little evidence that any type of alcoholic beverage consumption was associated with an increased OPMD risk in Puerto Rico.

Smoking and Drinking Interaction

Previous studies have reported evidence of a more than additive effect of both smoking and drinking on the risk of oral cancer and OED [12, 15–17]. Our exploratory analyses revealed no evidence that alcohol consumption modified the effect of smoking in terms of OPMD risk, although this finding is based on a relatively small sample that hindered an in-depth examination of the potential interaction.

Limitations

When interpreting our findings, it is important to consider study limitations. The relatively small sample size reflects a comparatively low rate of OPMD diagnosis in Puerto Rico [45] and limited our ability to conduct subgroup analyses related to smoking, drinking, and their interaction.

We utilized a comparison group comprised of persons diagnosed with a benign oral soft tissue condition rather than a traditional "disease-free" group. The strategy used provides an estimate of exposure patterns among persons diagnosed with asymptomatic oral lesions in Puerto Rico. It is possible, however, that one or another oral condition included in our benign conditions group is positively associated with either smoking or alcohol consumption; in that instance, the prevalence of smoking and drinking in our comparison group may exceed that of the background PR source population. Had we used a "disease-free" comparison group, ORs for smoking may have been higher than our current estimates, and ORs for drinking may have been closer to or exceeded the null value, 1.0.

Adjustment was made for a number of potential confounders, and further adjustment for denture use and BMI had little impact on the reported ORs; however, our results may reflect some uncontrolled confounding.

All exposures were measured based upon subject self-reports, and some degree of exposure misclassification may have occurred. In addition, since many OPMDs and benign oral tissue conditions are asymptomatic and may not be biopsied, we cannot rule out the possibility that undiagnosed cases are different from subjects included in the study in terms of the risk factors of interest.