Influence of Personal, Environmental, and Community Factors on Cigarette Smoking in Adolescents: A Population-Based Study from Taiwan

Understanding the factors that influence cigarette smoking among adolescents is critical. We identified personal, community, and environmental factors associated with current cigarette smoking among adolescents. This population-based cross-sectional analysis study was conducted using the 2012 Taiwan Global Youth Tobacco Survey and the sociodemographic statistics of the city or county from Taiwan’s Ministry of the Interior. A total of 27,524 participants (age: 12–18-years) was included. The associated factors were identified through logistic regression. A path analysis was performed to examine the pathway from the associated factors to current cigarette smoking. According to this analysis, the following factors were prominently and positively associated with adolescent cigarette smoking: one personal factor (pocket money), five environmental factors (home secondhand smoke (SHS) exposure, smoker friends, outside SHS exposure, school SHS exposure, and smoker parents), and two community factors (free cigarettes from tobacco companies and indigenous population). By contrast, five personal factors (feeling less comfortable smoking at social occasions, feeling indifferent about smoking or not smoking at social occasions, female sex, feeling that quitting is difficult, and feeling that quitting after having smoked is harmful to health) and one environmental factor (school antismoking education) had negative effects. Thus, comprehensive interventions promoting the perception of harm caused by smoking and interrupting access to cigarettes through social networks can reduce cigarette smoking in adolescents.


Introduction
Since the recognition of cigarette smoking as the leading cause of preventable disease and death [1], its prevalence has progressively declined in Taiwanese adolescents [2]. The implementation of the Anti-Smoking Act may have limited results for modern teens due to the increased availability of alternative tobacco products (such as e-cigarettes) and advertising targeted at teenagers [3]. Nearly 80% people who have ever smoked daily smoked their first cigarette before they were 18 years old [4,5] and more than onethird of the adults who have ever smoked a cigarette smoked their first cigarette were 14 years of age [6]. Earlier smoking initiation was associated with sustained smoking through adulthood [6]. Thus, understanding the factors that influence adolescent cigarette smoking in adolescents is crucial. Research has highlighted the importance of multilevel influential factors for determining health behaviors in adolescents [7,8]. According to the socioecological developmental model [9] and social cognitive theory [10], smoking The Taiwan Global Youth Tobacco Survey (GYTS) is a national sample of Taiwanese middle-and high-school students (i.e., 12-18-year-old junior high, senior high, vocational school, and night-school students) from all 22 cities and counties of Taiwan. This GYTS was developed by the Taiwan Health Promotion Administration in conjunction with the US Center for Disease Control and Prevention for surveillance of tobacco use behavior, perceptions, and attitudes. It involves a self-administered questionnaire with items focusing on cigarette smoking behavior, perceptions, and attitudes toward smoking hazards, SHS exposure, and smoking ban implementation on campuses. Basic demographic variables are also recorded. In 2012, the Taiwan GYTS used a multistage sample design using regional-level stratification, with schools selected in proportion to their enrolment sizes and a representative sample of students was also selected. The applicable classrooms were randomly chosen from within the selected schools, and all students in these selected classrooms were eligible for participation. Moreover, sociodemographic statistics on each city or county in which the schools were located were collected from Taiwan's Ministry of the Interior.

Study Design
This was a population-based cross-sectional analysis study.

Study Population
The questionnaire from the 2012 Taiwan GYTS was completed by 34,552 students, with an overall response rate of >90.0%. Participants aged >18 years and with missing data were excluded. Thereafter, responses of only 27,524 participants were included.

Outcome
Current cigarette smoking was defined through self-report of having smoked cigarettes on at least 1 day in the past 30 days. The related question in the questionnaire was "How many days in the past 30 days did you smoke?" The participants responding to this question with 1 or more days were categorized as current cigarette smokers.

Potential Associated Factors
The following potential associated factors were evaluated: personal (knowledge of health risks related to smoking and harmful effects of SHS and attitudes toward smoking); environmental (parents and peers who smoke, exposure to school antismoking education, quantity of SHS exposure (days of SHS exposure at home, school, and outside of home and school over the past 7 days)); and community (TAPS exposure, sociodemographic statistics of the school region (i.e., indigenous population, low-income population, college and university graduate population, migrant population, temple density, crime rates, fertility rate of under 19 years old, divorce rate, and communicable diseases burden)).

Statistical Analysis
The chi-square test was used to compare the distributions of personal, community, and environmental factors among current and noncurrent smoking participants. Logistic regression analysis was used to explore the adjusted effect of explanatory variables on current smoking after mutual controlling. Odds ratios (ORs) and their 95% confidence intervals (CIs) are reported. Values of p < 0.05 were considered statistically significant.
After the predicting factors were ascertained through logistic regression, path analysis was performed to examine the pathways from the possible factors to current cigarette smoking. Modification indices were used to assess the model fit. Five statistical tests were used to evaluate the overall goodness-of-fit of the correction model: standardized root mean square residual, root mean square error of approximation, comparative fit index, Tucker-Lewis index, and parsimonious goodness-fit-index. After the model was corrected, the effects of factors associated with adolescents' current smoking were estimated and significantly positive and negative associations were reported. Finally, all data were split into two data sets (training and testing) to evaluate the consistency of the results through validation analysis. The data analysis was performed using SPSS (version 23) and Amos (version 21).

Ethical Approval
The Institutional Review Board of National Taiwan Normal University (Protocol Number: 201902HM005) approved the study. Written consent was exempted because the data were obtained from Taiwan Youth Tobacco Survey, which contains de-identified secondary data released for research purposes. Table 1 presents the results of the descriptive analysis of the background information and study variables. Overall, 27,524 students were included in the analysis. Of these, 50% were females, and 49.1% were aged 13-15 years. Approximately 9.2% of the participants were current smokers, of which 7.0% were aged 13-15 years and 11.3% were 16-18 years. Of the current smokers, most (80.5%) had received antismoking education at school; of these, 74.9% were considering ceasing smoking because of its harmful effects and 94.8% reported knowing that SHS is harmful. Furthermore, the parents of 65.7% of the current smokers were smokers themselves, and most of the friends of the 39.4% of the current smokers were also smokers. Moreover, 52.6% and 62.3% of the participants reported SHS exposure at school and home, respectively, over the previous 7 days, and 23.8 % had received free cigarettes from a tobacco company. The distribution of these factors was significantly different between the nonsmokers and smokers.

Factors Associated with Current Cigarette Smoking
The community factor indigenous population was associated with an increased likelihood of current smoking (OR 1.01, 95% CI 1.01-1.02, p < 0.001). Moreover, adolescents from areas where tobacco companies offered free cigarettes presented significantly higher smoking risk than those from other areas (p < 0.001). Current cigarette smoking was lower for some personal factors, such as female gender (OR 0.54, 95% CI 0.48-0.60, p < 0.001) and low acceptability of smoking during social occasions (OR 0.31, 95% CI 0.26-0.35, p < 0.001). School antismoking education and rules, which are community factors, were associated with a lower likelihood of current smoking (OR 0.74, 95% CI 0.62-0.89, p = 0.001).

Effects of Associated Factors on Current Cigarette Smoking
According to the goodness-of-fit statistics, the final correction models were acceptable (Supplementary Table S1). Figure 1 presents the final path analysis model of the total effects on current cigarette smoking incidence. Table 3 presents the factors associated with direct, indirect, and total effects on current cigarette smoking incidence. Eight associated factors exhibited a direct positive effect on smoking incidence: free cigarettes from tobacco companies (0.182), home SHS exposure in the previous 7 days (0.153), smoker friends (0.104), SHS exposure outside of the home and school in the previous 7 days (0.061), school SHS exposure in the previous 7 days (0.046), indigenous population (0.037), smoker parents (0.019), and pocket money (0.08). Six associated factors displayed a direct negative effect on smoking incidence: feeling less comfortable smoking on social occasions (−0.236), feeling indifferent to smoking or not smoking on social occasions (−0.170), female gender (−0.068), feeling that quitting is difficult (−0.047), receipt of school antismoking education (−0.021), and feeling that quitting after having smoked in harmful to health (−0.015). The training model results were consistent with the results of the testing model (Supplementary Table S2) and with original data. Table 3. Direct, indirect, and total effect of associated factors on current cigarette smoking incidence.

Effects of Associated Factors on Current Cigarette Smoking
According to the goodness-of-fit statistics, the final correction models acceptable (Supplementary Table S1). Figure 1 presents the final path analysis m the total effects on current cigarette smoking incidence. Table 3 presents Table S2) and with original data.  Table 3. Direct, indirect, and total effect of associated factors on current cigarette smoking inc

Discussion
In this study, we examined the personal, environmental, and community factors associated with current cigarette smoking in adolescents through multivariate logistic regression and path analyses. After simultaneously assessing the associations, one personal factor (pocket money), five environmental factors (home SHS exposure, smoker friends, SHS exposure, school SHS exposure outside of the home and school, and smoker parents), and two community factors (free cigarettes from tobacco companies and indigenous population) were found to be associated with a significant increase in cigarette smoking incidence. By contrast, five personal factors (feeling less comfortable smoking on social occasions, feeling indifferent about smoking or not smoking on social occasions, female gender, feeling that quitting is difficult, and feeling that quitting after having smoked is harmful to health), and one environmental factor (antismoking education) negatively affected current smoking behavior. The factors affecting youth smoking during the developmental stages of adolescence are highly complicated. Reducing smoking among adolescents may require comprehensive interventions tailored toward the promotion of adolescent perceptions regarding the harmful effects of smoking, as well as toward preventing tobacco companies from advertising through social media or offering free cigarettes to adolescents.
Cigarette smoking is linked to gender. The present study confirmed a previous finding that the rates of current cigarette smoking are lower in females [30] However, this result is inconsistent with the worldwide trend, in which no gender difference has been noted in adolescents' smoking behavior. In the sociocultural context, Taiwan has an East Asian culture, in which the social acceptance of smoking by women is low. By contrast, in various Western countries, smoking among women is as acceptable as it is among men. We also observed that teenagers who felt less comfortable smoking during social occasions had a lower likelihood of being current smokers. However, this observation may represent the sociocultural influence of smoking behavior on teenagers, as a result of which they are reluctant to publicly reveal their smoking habit.
Smoking is a learned behavior. Better knowledge of and attitudes toward smoking have positive effects on current smoking in adolescents. Increased access to antitobacco information at school can increase students' knowledge of its risks and change their attitudes toward smoking [31]. Lower levels of knowledge on smoking hazards [32] and positive attitudes toward smoking [33] promote smoking among Asian adolescents. A GYTS in South Asia suggested that adolescents who are not informed about the harmful effects of cigarette smoking at school or home are more likely to smoke [34]. We observed that nonsmokers who received school antismoking education believed that quitting was difficult, and believed that it was harmful to smoke and then quit. Thus, establishing antismoking behaviors through education is valuable. Antismoking intervention programs have led to a decrease in smoking behaviors [35,36]. Such programs provide smokingrelated health information through mass media, schools. In general, success with a single strategy is difficult. For instance, a study investigating education about the harm caused by smoking in schools reported no association of this education with current smoking in adolescents [37]. The results might have been inconsistent due to other factors, including SHS exposure or the desire to mimic adult behavior, despite knowledge of the risks [38].
Parents and peers can influence smoking in adolescents. Having smoker parents and peers was positively associated with smoking initiation in adolescents [39] and higher smoking susceptibility; this is also a significant predictor of adolescents' current smoking status [40]. The risk of smoking in childhood and adolescence increases if at least one parent smokes, and this risk increases almost threefold when both parents are smokers [41]. Social development can also influence adolescents' smoking behavior. Students who are under peer pressure are more likely to smoke cigarettes [42]. A 3-year cohort study investigated the effect of parent and peer influence on adolescent smoking and found no parental effect with age. This study also indicated that best friends and friends had a stronger influence on younger adolescents, whereas friends and same-grade students had a stronger influence on older adolescents [43]. Moreover, smoker parents and friends were strong determinants of all types of SHS exposure [44]. SHS exposure is a significant risk factor for current smoking in adolescents, and increases the risk of cigarette smoking among young adolescents [45]. Under the Tobacco Harm Prevention Act, smoking is completely prohibited in most indoor or outdoor public spaces in Taiwan, except in designated smoking areas. However, some private spaces that are free of law enforcement, such as homes, could be major areas of SHS exposure for youths if their parents smoke. Some SHS exposure among youths could be unintentional. If someone lives in a smoking family, their parents are less likely to ban smoking and may even provide adolescents with tobacco products. In such cases, adolescents are in a vulnerable position, and incapable of avoiding such smoking environments. A shown in Table 3, SHS and contact with smokers influence youth smoking. Home SHS exposure implies frequent exposure to smokers. Our findings also suggest that SHS exposure is associated with an increased likelihood of current smoking in adolescents. Comprehensive smoke-free legislation that prohibits smoking in all indoor public spaces, including schools, workplaces, bars, and restaurants, can reduce SHS exposure and encourage people to quit smoking [46].
Easy access to cigarettes is a key factor contributing to smoking initiation among adolescents [47]. Access to free cigarettes, either from tobacco companies or smoker friends, was the strongest predictor of cigarette smoking in this study. Selling or providing tobacco products to younger people is illegal in Taiwan. Nonetheless, adolescents, particularly current smokers, reported that they were able to obtain cigarettes from retail outlets, such as convenience stores, supermarkets, tobacco sales, cyber cafés, weddings, funerals, Kondatam television (KTV), and music television, in the 30 days prior to the questionnaire, and approximately 3-5% of smokers had been offered free cigarettes, which implies that strict regulation is necessary. This is valuable information for policy makers because it highlights the need for surveillance in the retail industry. Adolescents' exposure to TAPS is also a critical factor influencing smoking initiation. The promotional activities of tobacco companies, particularly their distribution of free cigarettes, were significantly associated with increased smoking risk [48]. Adopting and enforcing interventions that prevent adolescents from accessing tobacco products should be included in strategies for reducing smoking initiation. Adolescents are more sensitive to cigarette pricing policies [49]. Pocket money was positively associated with current smoking. A cigarette pricing policy could bar adolescents from accessing cigarettes [50].
The path analysis in this study was not used to test the overall model validity but to clarify the specific direct and indirect relationships across the variables. Our study provides a possible relational structure but does not eliminate other possible modes. Despite its limitations, this study has several key strengths. We used a two-step analysis to examine the pathway from associated factors to current cigarette smoking; this has rarely been reported. After the consideration of all the possible associations, our findings on the associations of personal, environmental, and community factors with cigarette smoking in adolescents were consistent with prior results. These factors, particularly access to free cigarettes, smoker friends, and SHS exposure, play a major role in tobacco use among adolescents. More comprehensive antismoking acts and effective strategies are required to prevent adolescent cigarette smoking.
In conclusion, the present study demonstrated the influences of personal, environmental, and community factors on current smoking among Taiwanese students aged 12-18 years, and found that single interventions were not effective. Comprehensive interventions promoting the perception of the harmful effects of smoking, as well as interrupting access to cigarettes through social networks, can reduce cigarette smoking in adolescents. Taiwan follows the framework convention on tobacco control implemented by the Tobacco Harm Prevention Act, promoting various preventive measures. Therefore, the prevalence of smoking among 12-15-year-olds and 16-18-year-old adolescents decreased to 61% and 43%, respectively, from 2008 to 2019. However, the perception of cigarette smoking as being trendy and cool and the access to advanced-technology electronic products, such as electronic cigarettes, are influencing young people to use cigarettes. SHS exposure, smokers, and the offer of free cigarettes to adolescents combine to influence youth smoking, and comprehensive control is urgently required.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/healthcare10030534/s1. Table S1: Overall indices related to the goodness-of-fit model. Table S2: Validation analysis. Informed Consent Statement: Patient consent was waived due to the fact that the data were obtained from the Taiwan Youth Tobacco Survey, which contains de-identified secondary data released for research purposes.

Data Availability Statement:
The data used to support the findings of this study are included within the article.

Conflicts of Interest:
The authors declare no conflict of interest.