Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

The concept of exposure when selecting comparison groups for determining individual susceptibility to addiction to cigarette smoking

  • Indiara W. Henn,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Writing – original draft, Writing – review & editing

    Affiliation Graduate Program in Dentistry, Pontifícia Universidade Católica do Paraná, Curitiba, PR, Brazil

  • Luciana R. A. Alanis,

    Roles Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliation Graduate Program in Dentistry, Pontifícia Universidade Católica do Paraná, Curitiba, PR, Brazil

  • Adriana Modesto,

    Roles Conceptualization, Funding acquisition, Project administration, Validation

    Affiliation Departments of Oral Biology and Pediatric Dentistry, School of Dental Medicine, University of Pittsburgh, Pittsburgh, PA, United States of America

  • Alexandre R. Vieira

    Roles Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliations Graduate Program in Dentistry, Pontifícia Universidade Católica do Paraná, Curitiba, PR, Brazil, Departments of Oral Biology and Pediatric Dentistry, School of Dental Medicine, University of Pittsburgh, Pittsburgh, PA, United States of America

The concept of exposure when selecting comparison groups for determining individual susceptibility to addiction to cigarette smoking

  • Indiara W. Henn, 
  • Luciana R. A. Alanis, 
  • Adriana Modesto, 
  • Alexandre R. Vieira


Smoking is a leading cause of preventable death. The effect of tobacco is even more contundent in people with mental illness and, in general, cigarette smoking addiction is influenced by genetic factors. The opioid system is involved in the mesolimbic reward system, which is of great importance in addictive behaviors, such as smoking and is influenced by genes such as the OPRM1. The aim of this study was to evaluate if selecting a comparison group that include light smokers versus people that never smoked impacts the results of genetic association studies. In addition, to evaluate the genetic association in different groups of smokers by analyzing independent covariates such as mental illness and clinical dental data. All subjects were participants of the Dental Registry and DNA Repository project. Genotyping was carried out using TaqMan chemistry for two markers in OPRM1 (rs553202 and rs7755635). Logistic regression analyses were performed as implemented in PLINK. The established value for alpha was 5%, and the Hardy-Weinberg equilibrium was evaluated by the chi-square test with one degree of freedom for each marker. 1,897 patients were included, which were allocated to eight distinct groups, according to the frequency and quantity of cigarettes smoked and mental illness status. There was no significant association between the two markers in OPRM1 and smoking. When mental illness and dental clinical data (tooth loss, dental caries, and periodontitis) were used as covariates, there were associations between heavy smoking and OPRM1, when non-smokers were used as comparison. We did not have diet or microbiome data to consider for these dental analyses and suggest that these kinds of data should be always incorporated in the future. Significant results were found only when the covariables mental illness and oral clinical data were added to the analysis.


Smoking is a leading cause of preventable death, resulting in more than 7 million deaths annually worldwide and a huge impact on global health. There are around 1 billion smokers in the world, with 80% living in low and middle-income countries [1]. These smokers represent approximately one-third of the global population aged 15 years or older. Even knowing the serious consequences with which smoking is associated, such as high rates of cancer, cardiovascular and pulmonary diseases, and consequently death, about 45.3 million adults in the U.S. smoke cigarettes, and more than 68% claim to want to quit [2].

After starting smoking, the individual can rapidly develop dependence on tobacco, even in the case of young and irregular smokers [3]. Soon after smoking the first cigarette, the neurophysiological processes underlying tobacco dependence are initiated [4,5]. The younger one starts smoking, the greater the risk of addiction in later stages of life [6]. Tobacco dependence occurs through nicotine, which is the main psychoactive component in tobacco [7].

The data on the effect on tobacco is even more contundent in people with mental illness. A meta-analysis of more than 40 studies conducted in 20 countries showed that smokers with severe mental illness were significantly heavier smokers than smokers in the general population. In addition, the prevalence of smoking in patients with schizophrenia was 62%, which represents 5.3 times more than in the general population [8]. In Western countries, the prevalence of smokers in patients with bipolar disorder is 66%, and 60% of patients with depression also smoke [9]. A national survey of people living with psychotic illness in Australia found that 73% of men and 56% of women smoked tobacco [10].

In general, cigarette smoking addiction is influenced by genetic factors [11]. Heredity for initiation of smoke is estimated to be approximately 48–60% [12, 13, 14], and heritability seems to be higher in men, but the difference between the sexes is small [15].

We still do not know why people with mental illnesses smoke excessively but knowing that the opioid and endogenous dopaminergic systems are involved in the strengthening of smoking, the study of these genes is justified. Hirasawa-Fujita et al. [16] observed an association among smoking, schizophrenia and the opioid receptor OPRM1, and that an alteration of the DRD2 (dopamine D2) function increased the smoking behavior in women with schizophrenia. The mu-opioid receptor (MOR), encoded by OPRM1, is a natural regulator of analgesic pain response and also controls the gratifying effects of many drugs of abuse such as opioids, tobacco, and alcohol [17]. Vink et al. [18] performed a genome-wide association study comparing ever smokers with never smokers and current smokers versus nonsmokers and found evidence of association between OPRM1 and smoking initiation by network-based analysis.

In the study of infectious diseases, affected and unaffected individuals are typically recruited from endemic areas where groups are naturally exposed to the microbial challenge and the resistant and susceptible phenotypes are thus unveiled [19, 20]. However, in the context of smoking, the possibility to impact exposure by discouraging people to smoke interferes with the exposure factor, and consequently a population of non-smokers (the usual control of traditional case–control smoking studies) is theoretically composed of both susceptible and resistant subjects, although their genotypic nature remains uncertain. Therefore, it is possible to affirm that the traditional case–control smoking studies clearly disregard the classic case–control study definition, which states that a case–control study is designed to determine if an exposure is associated with an outcome [21]. Here we hypothesize that a reappraisal of the case–control design based in the observance of the case–control study definition may significantly impact the odds of identification of genetic factors associated with smoking addiction risk. In this context, we propose that patients that smoke only occasionally fulfill the exposure concept and could be a more suitable control for case–control studies representing an apparent resistant phenotype in opposition to uncertain phenotype of subjects that never smoked. To test our hypothesis, in this study we compared genotype frequencies of markers of OPRM1 by the traditional case–control approach (i.e. smokers versus non-smokers) and the proposed new design (heavy smokers versus light smokers). In addition, to evaluate a genetic association in different groups of smokers and covariates such as mental illness and clinical dental data.

Materials and methods

All subjects were participants of the Dental Registry and DNA Repository (DRDR) project at the University of Pittsburgh School of Dental Medicine [University of Pittsburgh Institutional Review Board (IRB) approval # 0606091]. This project was started in September 2006 and, as of this date, all the patients treated at the Dental Clinics of the University are invited to be part of the registry. All subjects provided written informed consent authorizing the extraction of dental and medical information from their records and provided a saliva sample for future genetic studies. All 5,800 subjects recruited by July of 2017 were included in this study. The data extracted from the registry database included clinical information and self-reported medical history.

Definition of controls from published studies looking at smoking and mental illness

In order to confirm how case-control studies on smokers define their appropriate controls for smokers with mental illness, we performed a systematic review in June 2017 with the question (PECO): "What would be the best controls used for a group of dependent smokers with mental illness?". The search was performed using the keywords: "tobacco smoking" OR "nicotine dependence" AND "control group" AND mental. We obtained a total of 61 articles (12 in Cochrane, 29 in Scopus and 20 in PubMed) and, after reading each abstract, we excluded 56 articles because they did not show the group selection information, were not related only to smokers, or were related to cigarette cessation tests. In the end, five articles (four case-control articles and one cohort) were selected for analysis (Table 1). During the peer-review process, we repeated our search in February 2019 with the quotation markers and without the quotation marks. These searches gave us an additional four papers when the search was done with quotations and 191 papers when no quotations were used. However, no papers could be added to the ones we selected in our original search.

Table 1. Identified studies describing controls for smokers with mental illness.

We observed a lack of concordance among the studies for the definition of "smoker", with no information related to the quantity and frequency of tobacco used. In addition, the definition of "non-smokers" also did not appear standardized. Zhang et al. [26] considered nonsmokers those who smoked less than 100 cigarettes throughout their lives, while Guney et al. [23] considered non-smokers those who never smoked. Since we could not answer the question of the systematic review and to avoid grouping participants who smoked little or very much in a single group of smokers, and since quantity and frequency of tobacco may be influenced by the genetic predisposition to dependence, and to consider the existence of ex-smokers, we used a few different groups as comparison for the genetic association study we performed.

Study participants

In order to select the patients who would be part of this study, we classified smokers into different groups, according to the amount and frequency of smoking, and also in two subcategories, with and without mental illness: (1) heavy smokers with mental illness; (2) heavy smokers without mental illness; (3) ex-smokers with mental illness; (4) ex-smokers without mental illness; (5) light smokers with mental illness; (6) light smokers without mental illness; (7) non-smokers with mental illness; and (8) non-smokers without mental illness.

Light smokers were those who smoked less than 10 cigarettes (half a pack) per day and heavy smokers, those who smoked more than 11 cigarettes per day. Ex-smokers were considered those who had stayed at least five months prior to the survey, without smoking. Nonsmokers were those who had never smoked in their lives, according to the report of the medical history. The cutoff point between light and heavy smokers was decided based on the way the data were reported in our project: smoking up to half a pack or more than half a pack of cigarettes per day.

For the covariant mental illness, thinking of not subdividing the groups in relation to all the mental illnesses found in the records and thus obtaining very small and unrepresentative samples from each group, we decided to aggregate all the mental illnesses found: depression, bipolar disorder, anxiety, schizophrenia, panic disorders, high stress, eating disorders, and, most often, the association of more than one psychic disorder. In this way, we divided the patients according to presence or absence of mental illness.

Dental clinical data

Dental clinical data were collected from the medical records of the study participants: tooth loss, dental caries, and periodontitis. Dental caries experience data were obtained through the DMFT (Decayed, Missing due to caries, Filled Teeth) index and tooth loss was defined when there was a lack of one or more teeth, extracted due to dental caries or periodontitis. Periodontitis was defined in individuals presenting at least three teeth exhibiting sites of clinical attachment loss equal or greater to 5 mm in two different quadrants. All clinical conditions were evaluated by professionals in training under the supervision of experienced dentists. These data were dichotomized as to their presence or absence. Dental caries and periodontitis were also analyzed in combination (individuals with both conditions versus individuals with none of the conditions) per the current joint suggestion of the European Organization for Caries Research and the European Federation of Periodontology (EFP) [27].


In addition to the dopaminergic, serotonergic and nicotinic receptor genes, there are other genes linked to addictive substances that can be found in the endogenous opioid system. The opioid system is involved in the mesolimbic reward system, which is of great importance in addictive behaviors, such as smoking [28]. Within this system, a gene often studied is OPRM1 [28, 29, 30, 31]. OPRM1 occupies a region of 200kb in the long arm of chromosome 6 and is a receptor for endogenous opioids, such as beta-endorphin and endomorphin [17]. We selected two markers flanking OPRM1 (rs553202 and rs7755635) in 6q24-q25 that were not in strong linkage disequilibrium with each other and had good heterozygosity (30% or more) for this study.

DNA extraction and genotyping

Genomic DNA was extracted from oral cells of unstimulated saliva (2mL) collected using Oragene DNA kits (DNA Genotek, Ottawa, Ontario, Canada) and extraction protocol according to the manufacturer's specifications.

Genotyping was performed by polymerase chain reactions using the Taqman method [32] with an ABI PRISM QuantStudio 6 Flex instrument (Foster City, CA). The pre-designed probes were supplied by Applied Biosystems (Foster City, CA).

Statistical analysis

Data were analyzed using the SPSS (Statistical Package for the Social Sciences), version 22 (IBM SPSS Statistics, Chicago, IL) package, and the PLINK software [33].

Linkage disequilibrium between the two genetic markers tested was calculated using D’ statistics implemented in the Haploview 4.2 software [34].

Logistic regression analyses of each genetic marker was performed as implemented in PLINK. Dental clinical data (tooth loss, caries and periodontitis) were used as covariates in logistic regression analysis. The established value for alpha was 5%, and the Hardy-Weinberg equilibrium was evaluated by the chi-square test with a degree of freedom for each marker.

The samples were tested for association with OPRM1 as follows:

1. Heavy and Light Smokers versus Ex-Smokers and Non-Smokers

2. Heavy smokers versus Light Smokers

3. Heavy Smokers versus Ex-Smokers

4. Heavy smokers versus Non-Smokers

5. Heavy Smokers versus Light Smokers, Ex and Non-Smokers

6. Heavy Smokers versus Light Smokers and Ex-Smokers

7. Heavy Smokers, Light Smokers and Ex-Smokers versus Non-Smokers

8. Light Smokers versus Non-Smokers

9. Light Smokers versus Ex-Smokers

10. Ex-Smokers versus Non-Smokers

Power calculations [35], assuming that our marker alleles were in complete linkage disequilibrium with the genetic variant contributing to smoking dependence, showed that we have a sufficient statistical power of 80% under the following parameters specified in the calculations: high-risk allele frequency set at 0.1, the disease prevalence in the general population set at 0.2, and the genotype risks for the Aa and AA genotypes relative to the baseline aa genotype risk of at least 1.5 and 2.5. If the relative risk for heterozygotes is 1.5, we need at least 700 individuals in each comparison group for 80% power. If the relative risk for heterozygotes is 2.0, 200 individuals in each comparison group will give 80% power.


Initially, a total of 2,017 patients were included in this study, which were allocated to eight distinct groups: (1) heavy smokers with mental illness (n = 165), (2) heavy smokers without mental illness (n = 450), (3) ex-smokers with mental illness (n = 126), (4) ex-smokers without mental illness (n = 396), (5) light smokers with mental illness (n = 68), (6) light smokers without mental illness (n = 226), (7) non-smokers with mental illness (n = 353), (8) non-smokers without mental illness (n = 233). However, after processing of all saliva in preparation for genotyping, 120 samples were excluded due to poor quality and the final sample consisted of 1,897 patients (1: 152; 2: 447; 3: 110; 4: 372; 5: 64; 6: 217; 7: 325; and 8: 210).

Table 2 describes the studied sample according to sex, age, and geographic origin. Even though the comparison groups were not matched by sex, age, or geographic origin, we did not observe differences in their distribution.

Table 2. Samples included in the study (1: heavy smokers with mental illness, 2: heavy smokers without mental illness, 3: ex-smokers with mental illness, 4: ex-smokers without mental illness, 5: light smokers with mental illness, 6: light smokers without mental illness, 7: non-smokers with mental illness, and 8: non-smokers without mental illness).

The 10 case-control group comparison options evaluated were in Hardy Weinberg equilibrium. D’ between the two markers was 0, suggesting they provide independent information about the locus. There was no significant association between OPRM1 (rs553202 and rs7755635) and smoking (Table 3).

Table 3. Genetic analysis (allelic association and Hardy Weinberg) of gene OPRM1 and smoke.

Logistic regression analyses with mental illness as covariate showed associations between heavy smoking and OPRM1 for two comparisons of different case-control frameworks (Table 4):

  1. 4 = Heavy Smokers versus Non-Smokers
  2. 7 = Heavy, Light and Ex-Smokers versus Non-Smokers
Table 4. Logistic regression analysis results with covariates: Mental illness and Oral Conditions.

Our results showed that this association is only present when we compared the group of heavy smokers (the heavy smoking group alone or associated with all individuals who smoke and have smoked) with the group of those who have never smoked in their lives (non-smokers).

Logistic regression, using dental clinical data as independent covariables and correlating them with OPRM1 genotypes and tobacco dependence, found some associations. Tooth loss in the presence of OPRM1 genetic variation was more likely in people who currently smoked or smoked in the past (comparison 4—Heavy smokers versus Non-Smokers, 7—Heavy Smokers, Light Smokers and Ex-Smokers versus Non-Smokers, and 10—Ex-Smokers versus Non-Smokers). These data show that the chance for tooth loss is higher for smokers, regardless of the amount and frequency of tobacco use, compared to individuals who never smoked. Dental caries in the presence of OPRM1 genetic variation was less likely to occur for comparisons 7—Heavy Smokers, Light Smokers and Ex-Smokers versus Non-Smokers and 10—Ex-Smokers versus Non-Smokers, and more likely for comparison 9—Light Smokers versus Ex-Smokers. This result suggests dental caries in less likely in ex-smokers when compared to those who never smoked. However, when active smokers were considered, dental caries was more likely to occur compared to those who stopped smoking. Periodontitis in the presence of OPRM1 genetic variation was less likely to occur in smokers in comparisons 2—Heavy smokers versus Light Smokers, 5—Heavy Smokers versus Light Smokers, Ex and Non-Smokers, and 6—Heavy Smokers versus Light Smokers and Ex-Smokers, suggesting that heavy smokers were less likely to have periodontitis, when compared to light smokers. When we analyzed individuals with both dental caries and periodontitis in comparison with those without any of these two conditions, we observed that individuals who smoked more were less likely to have dental caries and periodontitis. Similarly, the comparison done with individuals who never smoked or were light smokers (2—Heavy smokers versus Light Smokers, 4—Heavy smokers versus Non—Smokers, 10—Ex—Smokers versus Non—Smokers) show the same results. However, when we evaluated group 9—Light Smokers versus Ex-Smokers, we observed that there was an increased risk of having dental caries associated with periodontitis for those who smoked compared to those who stopped smoking (Table 4).


It should be emphasized the importance of the development of a research in this field because of the high percentage of smokers in the world [1], particularly among individuals with mental illness [8], linked to the possibility of the dependence to nicotine is correlated with genetic factors.

We did not find an association between smoking and OPRM1, which corroborates with the results of Kleinjan et al. [31]. On the other hand, other studies described significant association between smoking and OPRM1. OPRM1 was shown to be associated with higher reward activity in response to nicotine intake because of the greater effectiveness at the μ-opioid receptor to β-endorphin binding, which leads to increased dopaminergic activity. This potential for opioid receptor binding appeared to be significantly higher for the A allele of OPRM1 (A118G/rs1799971) [30]. A meta-analysis performed by Arias et al. [28] on the A118G variant of the OPRM1 gene in relation to other substances (alcohol, opioids and heroin) evaluated 19 case-control studies and showed that the OPRM1 gene did not affect the risk of dependence on these substances. However, of the 7 studies that had a significant effect on the A118G variant, 4 of them reported that the G allele was the risk variant, while the other 3 studies demonstrated that G variant was protective for dependence. Another meta-analysis performed by Schwantes-An et al. [36] investigated the non-specific risk of OPRM1 with dependence on "general" substances (alcohol, opioids, marijuana, cocaine and nicotine), comparing substance-dependent cases with controls that were not dependent on the substances evaluated. The G allele of rs1799971 has been shown to have a modest protective effect on the risk of general substance dependence in samples of European ancestry. Similar effects have been reported for each individual substance, probably due to the small sample size. The authors concluded that rs1799971 contributed to the mechanisms of dependency liability that are shared between different substances that cause addiction. The divergence of the results among the studies may be related to several factors, such as the use of different genetic markers (rs1799971, rs553202 and rs7755635), sample size, as well as different compositions of the case / control groups.

Based on studies that we found on our systematic review, we observed a lack of concordance among the studies for the definition of "smoker", with no information related to the quantity and frequency of tobacco use in these studies. In addition, the definition of "non-smokers" also did not appear standardized in our systematic review. Therefore, we divided the group of smokers into four categories (heavy smoker, light smoker, ex-smoker and non-smoker). The characterization of smokers that we used, with four different groups in the same study, was not observed in the literature and makes our study unique. For example, Verde et al. [37] reported that smokers were those that smoked more tan 10 cigarettes per day at the time of the study and other studies that analyzed the association between smoking and genetic variants did not mention data on the quantity and frequency of tobacco use among smokers [38, 39]. We compared the addiction to smoke risk odds using the traditional approach (i.e. smokers versus non-smokers) and our proposed new design (heavy smokers versus light smokers).

Nicotine is the essential factor in tobacco addiction, and its action on nicotinic cholinergic receptors in the brain is primarily to increase the release of various neurotransmitters, including dopamine, noradrenaline, acetylcholine, glutamate, GABA, vasopressin, endorphin and others. The release of neurotransmitters is thought to mediate psychological effects such as arousal, relaxation, cognitive improvement, stress relief, and depression. Repeated exposure to nicotine increases its locomotor and booster effects in rodents, a phenomenon known as sensitization, common to psychostimulant drugs. In addition, repeated administration of nicotine enhances the dopamine stimulatory effects [40]. High doses of nicotine produce reward stimuli greater than low doses, and rapid arrival to the brain brings more reward than slow arrival. It is determinant for the establishment of dependence that the cigarette offers high concentrations of nicotine rapidly to the central nervous system [41]. The μ-opioid receptor (OPRM1) plays an important role in nicotine dependence because of its ability to bind β-endorphins and enkephalins, which are released after nicotine ingestion. This, in turn, leads to the release of dopamine into various areas of the brain, providing feelings of reward by increasing the additive properties of the cigarette [31]. Thus, the importance of smoking classification according to tobacco quantity and frequency is highlighted. Even with the classification of the groups and the various combination options for comparisons, we did not find any significant results associating smoking and the OPRM1 gene, which suggests indeed OPRM1 influence in smoking addiction may be too small to ever become of clinical significance or target for therapy, unless underlying mental disorder is present. Indeed, we found significant results for the association between smoking in individuals with mental illness and OPRM1. Our results showed that this association is only present when we compared the group of heavy smokers (the heavy smoking group alone or associated with all individuals who smoke and have smoked) with the group of those who have never smoked in their lives (non-smokers). Zerdazi et al. [42] also showed a similar result, finding an association between TLR4 and bipolar disorder. Hirasawa-Fujita et al. [16] observed that OPRM1 is associated with increased smoking in patients with schizophrenia, and DRD2 has also been associated with increased smoking behavior in women with schizophrenia. Thus, according to our results and the above-mentioned reports, we showed that genetic variation in OPRM1 associates with smoking addiction in patients with mental illness.

Oral health in indeed impacted by tobacco dependence. Tooth loss is higher for smokers, regardless of quantity, when compared to individuals who never smoked (non-smokers). These data corroborate the results of a cohort study that assessed the effects of smoking on tooth loss even after cessation of smoking, with a 46-year follow-up in northern Finland, that showed that smoking has effects on the loss of teeth in the long run [43]. The results for dental caries were less clear: it was less likely for the ex-smokers when compared to those who never smoked, and more likely for smokers compared to those who quit smoking. The last result represents that the cessation of smoking can lead to improvement in oral health and, consequently, decrease of the caries experience. Benedetti et al. [44], in a systematic review of tobacco and dental caries, showed that smoking was associated with an increased risk of dental caries. However, the poor overall quality of the studies did not produce validation for such association. Badel et al. [45] investigated the experience of caries and tobacco use in 19-year-old Croats and concluded that non-smokers had fewer decayed teeth. In our study, periodontitis was less likely in heavy smokers compared to light smokers. The high number of cigarettes would not bring more harm to the periodontium compared to daily smokers with low cigarettes (less than 10 cigarettes / day). This result is controversial in the literature; some studies show consistent data for tobacco as a risk factor for periodontal disease [46, 47, 48, 49] and other studies report that there is no association, but that this non-association may be weak or take more time to happen [50, 51]. Smoking is identified as a risk factor for periodontitis, but risk estimates diverge between studies [47].

Combining dental caries and periodontitis as a composite phenotype has been suggested [27]. Since smoking impacts both conditions, as well as increases the chance for their ultimate negative outcomes (tooth loss) to occur, the suggestion is warranted. Our results consistently showed that the individuals used for comparison (light smokers, ex-smokers, non-smokers, or a combination of them) tended to have more often dental caries and periodontitis in the presence of OPRM1 genetic variation. We considered the interaction between mental illness and oral health outcomes. This interest has occurred because this interaction has been studied. We also observed that tooth loss, caries, and periodontitis are "outcomes" present in both mental illness and tobacco dependence. Therefore, it is critical to emphasize that poor general health can lead to poor oral health. Although much emphasized lately [52], our data do not support the idea that poor oral health is driving the risk for mental illness.

Being concerned about multiple testing, we avoided applying the strict Bonferroni correction and increasing the type II error. If we had used Bonferroni correction, we would have lowered the alpha to 0.0025 (0.05/20). Therefore, here we report all results with P values below 0.05. However, our data must be carefully interpreted because it is expected that some of the P values below 0.05 can be due to chance. Also, our power analysis showed that we had sufficient statistical for a range of comparisons. But we know that we could have a lack of association that may be due to lack of statistical power.

The reason why non-smokers or light smokers were more likely to have dental caries associated with periodontitis when compared to heavy smokers is intriguing. One possibility is that the population studied is possibly exposed in high numbers to second hand smoking, low quality air, or other chemical irritants and that would account for the non-intuitive results. These subjects are from the greater Pittsburgh area; for the most part, individuals with low socioeconomic status and hence high risk for all oral and overall health problems tend to be over-represented in the patient pool [53].


In this study, no associations were found between OPRM1 and smoking, but when added to the analysis, mental illness and oral clinical data, significant associations were found. Our data show that the comparison group matters depending on the oral health outcome being measured. Tooth loss risk increases in the presence of OPRM1 genetic variation if one smokes at any frequency. Impact of OPRM1 in dental caries or periodontitis risk when someone smokes appears to depend of other factors that were not measured. The counterintuitive results for dental caries and periodontitis in these analyses may suggest that the effect of smoking in these diseases may be the consequence of dietary or microbiological changes that happen when someone stops smoking. We did not have diet or microbiome data to consider for these analyses and suggest that these kinds of data should be always incorporated in the future. Another limitation of this study is that the data collection was performed by different students / teachers, as it was carried out for several years. Because we work with secondary data from this database, we cannot prevent this. We also suggest that multiple comparison groups comprised by individuals that never smoked and individuals that smoked at one point and stopped should be utilized.

Supporting information

S1 File. Raw clinical data.

Raw clinical data.



Data for this work was obtained from the University of Pittsburgh School of Dental Medicine Dental Registry and DNA Repository project. I.W.H. was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (CAPES)—Finance Code 88881.131583/2016-01.


  1. 1. WHO, World Health Organization. Tobacco. 2017.
  2. 2. Centers for Disease Control and Prevention (CDC). MMWR Morb. Mortal. Wkly. Rep. 2011.
  3. 3. Difranza JR, Wellman RJ. A sensitization- homeostasis model of nicotine craving, withdrawal, and tolerance: integrating the clinical and basic science literature. Nicotine. Tob. Res. 2005; 7(1): 9–26. pmid:15804674
  4. 4. Breslau N, Kilbey MM, Andreski P. DSM-III-R nicotine dependence in young adults: prevalence, correlates and associated psychiatric disorders. Addiction. 1994; 89(6): 743–754. pmid:8069175
  5. 5. Audrain-Mcgovern J, Al Koudsi N, Rodriguez D, Wileyto EP, Shields PG, Tyndale RF. The role of CYP2A6 in the emergence of nicotine dependence in adolescents. Pediatrics. 2017; 119(1): 264–274.
  6. 6. Stolerman IP, Jarvis MJ. The scientic case that nicotine is addictive. Psychopharmacology. 1995; 117(1): 2–10. pmid:7724697
  7. 7. Pidoplichko VI, Debiasi M, Williams JT, Dani JA. Nicotine activates and desensitizes midbrain dopamine neurons. Nature. 1997; 390: 401–404. pmid:9389479
  8. 8. De Leon J, Diaz FJ. A meta-analysis of worldwide studies demonstrates an association between schizophrenia and tobacco smoking behaviors. Schizophr. Res. 2005; 76: 135–157. pmid:15949648
  9. 9. Li X, An F, Ungvari GS, Ng CH, Chiu HFK, Wu P, et al. Prevalence of smoking in patients with bipolar disorder, major depressive disorder and schizophrenia and their relationships with quality of life. Sci. Rep. 2017; 7:8430. pmid:28814728
  10. 10. Jablensky A, Mcgrath J, Herrman H, Castle D, Gureje O, Morgan V, et al. People Living with a Psychotic Illness: An Australian Study 1997–98. Canberra, Australia: Department of Health and Aged Care. 1999.
  11. 11. Rose RJ, Broms U, Korhonen T, Dick DM, Kaprio J. Genetics of smoking behavior. In: Handbook of Behavior Genetics. Kim YK (Ed.). Springer. 2009: 411–431.
  12. 12. Heath AC, Kirk KM, Meyer JM, Martin NG. Genetic and social determinants of initiation and age at onset of smoking in Australian twins. Behav. Genet. 1999; 29(6): 395–407.
  13. 13. Sullivan PF, Kendler KS. The genetic epidemiology of smoking. Nicotine. Tob. Res. 1(Suppl. 2), 1999: S51–S57.
  14. 14. Li MD, Cheng R, Ma JZ, Swan GE. A meta-analysis of estimated genetic and environmental effects on smoking behavior in male and female adult twins. Addiction. 2003; 1: 23–31.
  15. 15. Broms U, Silventoinen K, Madden PA, Heath AC, Kaprio J. Genetic architecture of smoking behavior: a study of Finnish adult twins. Twin. Res. Hum. Genet. 2006; 9(1): 64–72. pmid:16611469
  16. 16. Hirasawa-Fujita M, Bly MJ, Ellingrod VL, Dalack GW, Domino EF. Genetic Variation of the Mu Opioid Receptor (OPRM1) and Dopamine D2 Receptor (DRD2) is Related to Smoking Differences in Patients with Schizophrenia but not Bipolar Disorder. Clin. Schizoph. Relat. Psychoses. 2017; 11(1): 39–48.
  17. 17. Crist CC, Berrettini WH. Pharmacogenetics of OPRM1. Pharmacol. Biochem. Behav. 2014; 123: 25–33. pmid:24201053
  18. 18. Vink EG, Smit JM, de Geus AB, Sullivan EJC, Willemsen P, Hottenga, PG, et al. Genome-wide Association Study of Smoking Initiation and Current Smoking. The American Journal of Human Genetics. 2009; 84(3): 367–379. pmid:19268276
  19. 19. Sakthianandeswaren A, Foote SJ, Handman E. The role of host genetics in leishmaniasis. Trends in Parasitology. 2009; 25: 383–391. pmid:19617002
  20. 20. Vannberg FO, Chapman SJ, Hill AV. Human genetic susceptibility to intracellular pathogens. Immunological Reviews. 2011; 240: 105–116. pmid:21349089
  21. 21. Zondervan KT, Cardon LR, Kennedy SH. What makes a good case–control study? Design issues for complex traits such as endometriosis. Human Reproduction. 2002; 17: 1415–1423. pmid:12042253
  22. 22. Smolka MN, Budde H, Karow AC, Schmidt LG. Neuroendocrinological and neuropsychological correlates of dopaminergic function in nicotine dependence. Psychopharmacology. 2004; 175(3): 374–381. pmid:15114432
  23. 23. Guney F, Genc BO, Kutlu R, Ilhan BC. Auditory P300 event-related potential in tobacco smokers. Clin. Neurosci. 2009; 16(10): 1311–1315.
  24. 24. Yip SW, Sacco KA, George TP, Potenza MN. Risk/reward decision-making in schizophrenia: A preliminary examination of the influence of tobacco smoking and relationship to Wisconsin Card Sorting Task performance. Schizophrenia Research. 2009; 110(1–3): 156–164. pmid:19269138
  25. 25. Boumaza S, Lebain P, Brazo P. Tobacco smoking and psychiatric intensive care unit: Impact of the strict smoking ban on the risk of violence. L’Encéphale. 2015; 41: S1–S6.
  26. 26. Zhang XY, Chen DC, Tan YL, Luo X, Zuo L, Lv MH, Shah NN, Zunta-Soares GB, Soares J.C. Smoking and BDNF Val66Met polymorphism in male schizophrenia: A case-control study. Journal of Psychiatric Research. 2015; 60: 49–55. pmid:25455509
  27. 27. Chapple IL, Bouchard P, Cagetti MG, Campus G, Carra MC, Cocco F, et al. Interaction of lifestyle, behavior or systemic diseases with dental caries and periodontal disease: consensus report of group 2 of the joint EFP/ORCA workshop on the boundaries between caries and periodontal diseases. J. Clin. Periodontol. 2017; 44Suppl 18: 39–51.
  28. 28. Arias A, Feinn R, Kranzler HR. Association of an Asn40Asp (A118G) polymorphism in the mu-opioid receptor gene with substance dependence: a meta-analysis. Drug. Alc. Depend. 2006; 83(3): 262–268.
  29. 29. Raja R, Ruparelb K, Newbergc A, Wileytoa EP, Lougheadb JW, Divgid C, et al. Human Mu Opioid Receptor (OPRM1 A118G) polymorphism is associated with brain mu-opioid receptor binding potential in smokers. PNAS. 2011; 108(22): 9268–9273. pmid:21576462
  30. 30. Verhagen M, Kleinjan M, Engels RCME. A systematic review of the A118G (Asn40Asp) variant of OPRM1 in relation to smoking initiation, nicotine dependence and smoking cessation. Pharmacogenomics. 2012; 13(8): 917–933. pmid:22676196
  31. 31. Kleinjan M, Engels RCME, Difranza JR. Parental smoke exposure and the development of nicotine craving in adolescent novice smokers: the roles of DRD2, DRD4, and OPRM1 genotypes. BMC Pulmonary Medicine. 2015; 15: 115. pmid:26449981
  32. 32. Ranade K, Chang MS, Ting CT, Pei D, Hsiao CF, Olivier M, et al. High-throughput genotyping with single nucleotide polymorphisms. Genome. Res. 2001; 1: 1262–1268.
  33. 33. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et al. PLINK: a toolset for whole-genome association and population-based linkage analysis. American Journal of Human Genetics. 2007; 81.
  34. 34. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005; 21(2): 263–265. pmid:15297300
  35. 35. Purcell S, Cherny SS, Sham PC: Genetic power calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics 2003;19:149–150. pmid:12499305
  36. 36. Schwantes-An T, Zhang J, Chen L, Hartz SM, Culverhouse RC, Chen X, et al. Association of the OPRM1 variant rs1799971 (A118G) with non-specific liability to substance dependence in a collaborative de novo meta-analysis of European-ancestry cohorts. Behav. Genet. 2016; 46(2): 151–169. pmid:26392368
  37. 37. Verde Z, Santiago C, Gonzalez-Moro JMR, Ramos PL, Martin SL, Bandres F, et al. ‘Smoking Genes’: a genetic association study. PLoS ONE. 2011; 6(10): e26668. pmid:22046326
  38. 38. Kita-Milczarska K, Sieminska A, Jassem E. Association between CHRNA3 and CHRNA5 nicotine receptor subunit gene variants and nicotine dependence in an isolated population of Kashubians in Poland. Med. Sci. Monit. 2016; 22: 1442–1450. pmid:27127891
  39. 39. Zlomuzica A, Machulska A, Roberts S, Von Glischinski M, Rinck M, Lester KJ, et al. The dopamine D2 receptor mediates approach-avoidance tendencies in smokers. Eur. Arch. Psychiatry. Clin. Neurosci. 2018; 268(3): 261–268. pmid:28364268
  40. 40. Di Matteo V, Pierucci M, Di Giovanni G, Benigno A, Esposito E. The Neurobiological Bases for the Pharmacotherapy of Nicotine Addiction. Current. Pharmaceutical. Design. 2007; 13: 1269–1284. pmid:17504235
  41. 41. Benowitz NL. Pharmacology of nicotine: addiction and therapeutics. Annu. Rev. Pharmacol. Toxicol. 1996; 36: 597–613. pmid:8725403
  42. 42. Zerdazi E, Oliveira J, Vorspan F, Bennabi M, Jamain S, Etain B, et al. TLR4 gene polymorphism associated with lifetime cigarette smoking in bipolar disorder. Journal of Neuroimmunology. 2017; 305: 96–101. pmid:28284355
  43. 43. Similä T, Auvinen J, Timonen M, Virtanen JI. Long-term effects of smoking on tooth loss after cessation among middle-aged Finnish adults: The Northern Finland Birth Cohort 1966 Study. BMC Public Health. 2016; 16: 867. pmid:27557640
  44. 44. Benedetti G, Campus G, Strohmenger L, Lingström P. Tobacco and dental caries: a systematic review. Acta. Odontol. Scand. 2013; 71(3–4): 363–371. pmid:23088732
  45. 45. Badel T, Pavicin IS, Carek AJ, Segovic S. Dental Caries Experience and Tobacco Use in 19-year-old Croatian Army Recruits. Coll. Antropol. 2014; 38(2): 671–675. pmid:25145005
  46. 46. Do Loc G, Gary S. Smoking attributable periodontal disease in the Australian adult population. J. Clin. Periodontol. 2008; 35: 398–404.
  47. 47. Jacob SP. Smoking as a risk factor for periodontitis: A literature review. Rev. Odonto. Cienc. 2010; 25(4): 406–411.
  48. 48. Neto JBC, Rosa EF, Pannuti CM, Romito GA. Smoking and periodontal tissues: a review. Braz. Oral. Res. 2012; 26(Spec Iss 1): 25–31.
  49. 49. Leite FRM, Nascimento GG, Scheutz F, López R. Effect of Smoking on Periodontitis: A Systematic Review and Meta-regression. Am. J. Prev. Med. 2018; 54(6): 831–841. pmid:29656920
  50. 50. Lopez R, Fernandez O, Jara G, Baelum V. Epidemiology of Clinical attachment loss in adolescents. J. Periodontol. 2001; 72: 1666–1674. pmid:11811502
  51. 51. Persson RE, Kiyak AH, Wyatt CCI, Macentee M, Persson GR. Smoking, a weak predictor of periodontitis in older adults. J. Clin. Periodontol. 2005; 32: 512–517. pmid:15842268
  52. 52. Barnett ML. The oral-systemic disease connection: An update for the practicing dentist. JADA. 2006; 137Suppl: 5S–6S.
  53. 53. McGarvey EL, Leon-Verdin M, Killos LF, Guterbock T, Cohn WF. Health disparities between Appalachian and non-Appalachian counties in Virginia USA. J. Community. Health. 2011; 36: 348–356. pmid:20862529