Tobacco use is related to increased periodontal disease, tooth loss, and decreased success of orthodontic appliances, and it may inhibit orthodontic tooth movement. Most smokers start during adolescence. Since most cessation attempts fail, prevention appears necessary.
A cross-sectional sample of orthodontic patients reported hypothesized risk factors for smoking and susceptibility to tobacco use initiation. Exploratory analyses regressed susceptibility to tobacco initiation on each hypothesized predictor variable in a separate logistic model that included a standard set of covariates.
Significant odds ratios (OR) were found for the presence of a smoker in the home (OR, 2.168; 95% confidence interval [CI], 1.144-4.107), a friend having no-smoking rules in his or her home and car (OR, 0.337; 95% CI, 0.128-0.886), having been offered a cigarette (OR, 4.526; 95% CI, 1.190-17.207), and exposure to tobacco advertisements (OR, 1.910; 95% CI, 1.044-3.496).
Peer, family, and environmental factors appear to increase children’s susceptibility to smoking in orthodontic populations. Attention to such factors could help dental clinicians to more effectively identify susceptible young patients in need of antismoking advice. Prospective and experimental studies are required to confirm the role that dental clinicians might play in youth smoking prevention.
Factors appear to increase susceptibility to smoking in orthodontic populations.
Understanding these risk factors may help clinicians prevent youth smoking.
More studies are needed to confirm the role clinicians can play in tobacco control.
Smoking remains the number 1 cause of preventable death, disease, and disability in the United States and costs the nation over $300 billion annually in direct medical expenditures and lost productivity. Early intervention is needed to prevent tobacco use initiation among youth, since 88% of adult smokers start by age 18. Although youth tobacco use has been declining for decades, the decline is slowing, with 2% of middle school and 9% of high school students smoking cigarettes. Prevention efforts appear necessary because only 5% to 10% of cessation attempts result in long-term abstinence.
General risk factors for smoking initiation among youth include, but are not limited to, exposure to tobacco advertising and media portrayal of smoking, family and peer use or approval of tobacco, low socioeconomic status, low academic achievement, low self-esteem, aggressive behavior, and lack of tobacco refusal skills. Whereas physicians are often tasked with preventing smoking among their young patients, pediatric dentists and orthodontists are in a good position to do the same because of frequent contact with their patients and predominantly younger patient populations. Tobacco prevention is relevant for dental clinicians, since tobacco use damages oral health through increased periodontal disease, tooth loss, and decreased success of orthodontic appliances. Furthermore, inhibited orthodontic tooth movement from exposure to tobacco smoke has been demonstrated in animal models.
A systematic review of 4 studies of youth tobacco initiation prevention by clinicians showed that all 4 produced lower initiation rates in the intervention group. However, reductions were statistically significant in only 1 study, and in the study implemented by orthodontists, a significant protective intervention effect was found among patients whose peer group thought smoking was socially desirable.
Few studies have evaluated risk factors for tobacco initiation among youthful orthodontic patients to inform dental clinician-based interventions. Since smoking rates among orthodontic patients are lower than in the general population, it is possible that a unique set of risk factors influences their smoking initiation. One cross-sectional study of a youth orthodontic sample found that at baseline an increased odds of tobacco use was associated with older age, white race, abnormal weight, alcohol use, not wearing a seatbelt, not getting 8 hours of sleep, not regularly brushing or flossing, having a tobacco user in the home, and having friends who thought smokers were cool. A 2-year follow-up study from the same sample found that male sex, considering oneself overweight, engaging in risk behaviors such as drinking alcohol and not using seat belts, not engaging in healthy behaviors such as flossing, peer influences, and increased age promoted smoking initiation. Similarly, a third study found that alcohol and marijuana use, insufficient sleep, poor academic performance, and poor dental hygiene are linked with tobacco use in orthodontic populations. Poor academic performance, peer influence, and family influence were risk factors observed in both the general population and the orthodontic studies discussed above. Understanding whether the 2 populations have other risk factors in common is limited due to few studies measuring the same risk factors. A more thorough understanding of potential risk factors would better inform which patients are most in need of future orthodontist-delivered intervention.
Researchers have defined persons as susceptible to becoming smokers if they do not report that they will not smoke in the future. Those categorized as susceptible have been found to be at increased risk for experimenting with tobacco. Studies have assessed the relationships between risk factors and likely future tobacco smoking by using a susceptibility to initiation measure. This study used a cross-sectional analysis of baseline data from an orthodontics-based health promotion trial to explore risk factors for susceptibility to future tobacco initiation among orthodontic patients in parts of Southern California and Mexico. Identifying the most at-risk orthodontic patients might maximize the potential of clinician-based tobacco prevention interventions.
Material and methods
Healthy Smiles, an orthodontist program, funded by the National Institutes of Health, was a randomized, controlled trial designed to test a minimal intervention to improve diet and activity practices, or to reduce tobacco use and exposure, of 8- to 16-year-old youth. Orthodontic practices in Southern California (San Diego, Orange, and Riverside Counties) and the northern border region of Baja California, Mexico, were recruited to deliver the program to their patients. All study procedures were approved by San Diego State University’s Institutional Review Board. All participants provided informed consent, or informed assent, for participation in the experimental research.
Orthodontic practices were identified using the American Association of Orthodontists’ membership directory, online searches, telephone directory advertisements, and referrals from participating orthodontists. Of the orthodontists in the United States and Mexico with whom contact was attempted between 2009 and 2013, by telephone or letter, or in person, 32 (8%) enrolled in the program. The low rate of enrollment stemmed from orthodontists’ lack of time or interest, exclusion due to retirement or part-time work status, and unsuccessful contact.
Enrolled orthodontic practices informed their patients of the research study by letter or in person. Patients permitting contact by researchers were screened for eligibility. Patients were excluded if they were not between 8 and 16 years old, had plans to move within a year, had less than a year remaining in treatment, were unable to care for themselves, had been diagnosed with severe depression or an eating disorder, had participated in organized sports or activities 3 or more times per week for 9 or more months of the year, or had been prohibited by a physician from engaging in regular physical activity. Of 4737 age-eligible patients receiving treatment at 1 of the 32 orthodontic offices enrolled in the Healthy Smiles Program, 2982 were contacted (63%), 858 (18%) refused to be screened, 990 (21%) qualified, and 693 (15%) enrolled. Twenty-two offices in the United States were private practices (participants, 519), 2 offices in the United States were part of a large dental health maintenance organization (participants, 32), and 8 orthodontic offices were in Tijuana, Mexico (participants, 142). Participants received $10 for completion of baseline measures, and parents were included in a quarterly raffle for $100.
Research staff conducted baseline measures at completion of the in-person parent and child consent process. Demographic data were collected in a self-administered questionnaire completed by the child’s parent. The child’s previous tobacco exposure, tobacco use, susceptibility, reported norms, tobacco rules, and peer influence, as well as parent tobacco use and exposure were collected as part of a 3-day series of computer-assisted telephone interviews.
The dependent variable, susceptibility to future smoking, was computed from responses to the 3 questions shown in Figure 1 , similar to those used by Lessov-Schlagger et al. Response options for each question were “definitely yes,” “probably yes,” “probably not,” and “definitely not.” The sampling frame was first reduced by including only those with no smoking history: ie, those who responded “no” to the question, “have you ever tried or experimented with cigarettes?” Then, following the logic originally introduced by Evans et al, nonsmokers were categorized as susceptible to future smoking if they responded with anything other than “definitely not” to all 3 questions; they were defined as not susceptible only if they responded “definitely not” to all 3 questions. This 3-question approach to measuring resulted in a susceptibility variable correlated with later tobacco initiation.
Independent variables were assessed using the questions from Figure 2 . Additionally, the presence of a smoker living in the home was reported by the parent. The child’s family and friends finding smoking to be uncool was assessed with questions 1 and 2, and was dichotomized as both family and friends thought smoking was uncool vs at least one who did not. Question 3 asked whether the child had ever been offered a cigarette. The child’s exposure to no-smoking rules was assessed with questions 4 through 7. For both the child’s and friend’s environments, no-smoking rules were coded as present only if the child responded “no one is allowed to smoke inside, ever” for both the home and the car. The no-smoking rules were coded as absent if the child reported other response choices such as “smoking is ok only for some people, places, or times.” Question 8 assessed reported norms and was then dichotomized according to whether the reported number of perceived smokers was above or below the average number of youth in California who smoke at school (middle school or high school). Questions 9 and 10 assessed whether the child had been advised by his or her orthodontist or another health care worker not to smoke. Question 11 assessed exposure to tobacco advertisements and was dichotomized as “no exposure” or “exposure” to tobacco advertising. Questions 12 through 16 assessed peer influence to smoke. Children with 2 or more affirmative responses were categorized as subject to high influence, and those with 1 or none were categorized as subject to low influence.
Analyses were performed with SPSS software (version 22.0; IBM, Armonk, NY). Descriptive analyses of the study sample at baseline included child race or ethnicity, child age, child smoking status, child smoking susceptibility, family income, and home smoking status. Means and standard deviations were calculated for age, whereas categorical percentages were calculated for all other descriptive variables. Controlling for a smoker living in the home was considered because it would likely have a strong influence on the child’s susceptibility to tobacco use initiation and could be related to other study variables. A smoker living in the home was significantly related to susceptibility to tobacco initiation (chi square test, P <0.05) and was therefore included with basic demographics as a covariate in all regression models.
Logistic regression was used to explore the relationship of each hypothesized risk factor to susceptibility of future smoking while controlling for whether a smoker lived in the home, child sex, child age, family income, and non-Hispanic white status. Each risk factor was run in a separate logistic regression model along with the 5 covariates, as opposed to 1 full model containing all risk-factor variables. Large variations for whom there were missing values for the different risk factors resulted in a drastically reduced sample size when all variables were analyzed in 1 model, leading to the decision to run separate analyses on each independent variable of interest.
The study sample had a mean age of 12.1 years (SD, 1.9). Frequency statistics for the demographic variables of sex, race, ethnicity, income, smoking status, smoker living in the home, and susceptibility to smoking are given in Table I . Presence of a smoker in the home was significantly associated with a child’s susceptibility to future smoking (chi-square test, n=340; OR, 2.168; 95% Confidence Interval [CI], 1.144-4.107). The sample (57% of households having income >$70,000) was more affluent than San Diego County as a whole (50% of households having income >$63,996). Frequencies for the independent variables are shown in Table II .
|Demographic variable||n||%||Independent variable||n||%|
|Pacific Islander||3||0.4||Smoker lived in home|
|Hispanic||279||43||Susceptible to smoking|
|Independent variable||n||%||Independent variable||n||%|
|Child’s rules||Advertisement exposure|
|No rule||55||13||No exposure||203||35|
|Friend’s rules||Been offered cigarette|
|Orthodontist advice||Friend influence|
|Other clinician advice||Friend/family think uncool|
|No||439||76||At least 1 does not||366||64|
|Below California average||315||64|
|Above California average||176||36|