Tobacco control programs span through a myriad of approaches, including relevant legislation, tobacco control policies, anti-tobacco population campaigns, school-based preventive programs, and smoking cessation services1. A smoking ban at home, schools, children’s playgrounds and indoor environments in general constitutes an integral part of tobacco control efforts aiming to prevent the uptake of smoking among young teenagers2,3. While the change in beliefs and attitudes about smoking among teenagers imposes complex challenges, home no-smoking rules represent a relatively simple and practical approach for smoking prevention4. Specifically, this rule has been shown to lower the risk of smoking initiation among adolescents, even if their friends5 or both parents are smokers3,6. Furthermore, it empowers tobacco control attitudes and lowers teenagers’ perceived prevalence of smoking7. In addition, it eliminates images of role models of smoking at home8, and increases the intention to quit smoking among family members leading to successful smoking cessation9, and overall decreased risk of adolescent smoking10-12. Lastly, a home no-smoking rule protects children from secondhand and thirdhand smoke2,13, and may also serve as a non-smoking norm at home even if parents are smokers14.

In contrast, smoking at home facilitates the exposure of teenagers to smoking behavior and negative parental role modeling and has been associated with positive beliefs, attitudes and expectations towards tobacco. In addition, it leads to a higher intention to smoke as well as a higher probability of smoking among teenagers15-18. Furthermore, it facilitates access to cigarettes and other tobacco products, while it undermines the value of role modeling of non-smoking parents or other adult household members19,20.

While the above-cited studies have shown an association of the home no-smoking rule with different aspects of adolescent smoking, there is little in the international literature on whether teenagers may also serve as vehicles of change at home using a bottom-up approach. Therefore, our study aimed to improve adolescents’ knowledge so that they can act as facilitators of change in enforcing the home nosmoking rule following school-based anti-tobacco programs.


The study was implemented during the 2016–2017 academic year among middle-school students in Athens, Greece, under permission from the Greek Ministry of Education and approval from the Bioethics Committee of Evangelismos Hospital (16/6/2016, Protocol number 131), an affiliated hospital of the National and Kapodistrian University, Medical School. The study was implemented in five middle schools. Students from 1st to 3rd middle-school grade were invited to participate in either in the intervention or control group. Written informed consent was signed by students’ parents who participated in the study.

The intervention used an experiential learning approach that included the reading of an excerpt from the Histories Book 5: Terpsichore [Herodotus’ Histories (5.92)] 21. This is a parable from the story of the tyrant of Corinth, Periandros, who was wondering how to protect his power and was told that the best way was to destroy the strong sheaves of the field, representing the most accomplished members of the society. Using this parable, adolescents were paralleled with the strongest sheaves, which have the potential to become the modifiers of society and play an active role in shaping a healthier community away from substance abuse and addiction. This was followed by a class discussion of its meaning through a focused retrieval and decoding of archetypal symbols, as reflected and applied in real-life scenarios, including tobacco control messages. After the intervention, students participated in role-playing and encouraged to declare their contribution as the strong sheaves of the community. As a follow-up activity, students were guided to write a letter to their parents to motivate them to implement the no-smoking rule at home.

Our intervention occurred in school classrooms of 20–25 students during their everyday program and lasted for two school hours plus the in-between break. Students from both groups completed anonymous self-administered questionnaires twice (before and after the intervention). Data collection lasted for about seven months (academic year 2016– 2017) at five public schools in the northwestern suburbs of Athens. To avoid performance bias, the experiential learning intervention was carried out by the same investigator in all class sessions. The home no-smoking rule was documented by students’ reports at baseline and at the follow-up at 3 months using the following multiple-choice item: ‘What rules are applied at your home related to smoking?’. Responses included: 1) No one is allowed to smoke inside the house, 2) Smoking at home is allowed only for guests, and 3) Smoking is freely allowed for everyone. A pre- and post-intervention item was also used to evaluate students’ intention to smoke regarding their likelihood of taking up smoking in the future (12 months) following the intervention.

Data were entered into a computerized database and analyzed using the statistical package for social sciences (IBM SPSS v 22). Associations were assessed between a home no-smoking rule and the intention to smoke among students in the intervention and control groups at baseline and at the follow-up at 3 months.

Chi-squared tests were used for categorical variables and t-tests for continuous variables. To discretely assess students’ knowledge, attitudes and intention toward smoking, a total score was calculated for each outcome. For corresponding questions: positive answers were tabulated as 1, negative answers as -1, and neutral answers as 0. The score of the knowledge and attitudes scale takes values from -16 (incomplete knowledge and attitudes) to +16 (full knowledge and attitudes). Attitudes scores ranged from -9 to +9, respectively. A two-tailed p-value of ≤0.05 was considered statistically significant.


A total of 351 students participated in our study, with 181 (51.6%) in the intervention group and 170 (48.4%) in the control group. The mean age of student participants was 13.00±0.96 years; 12.76±0.91 years in the intervention and 13.26±0.95 years in the control group (p<0.001). The majority of participants were female (191; 54.5%), and the distribution of middleschool students across grades was 202 (57.5%) in Grade 1, 68 (19.4%) in Grade 2, and 81 (23.1%) in Grade 3. There was a significant difference in age distribution between the intervention and control groups.

In the intervention group, 80.7% of students before the intervention and 94.5% (p<0.001) after the intervention replied that they believe they will not smoke in the future (Table 1). There was no impact of gender between the intervention and control group (p=0.335) and the gender ratio was very close to 1:1. Within males, the belief that they will not smoke increased (75.9% pre vs 91.9% post, p=0.015). In females, similar results emerged (85.1% pre vs 96.8% post, p=0.017).

Table 1

Intention to smoke

IntentionIntervention group (n=181)Control group (n=170)p
n (%)n (%)
I will smoke in the future (Before intervention)0.764
Not possible146 (80.66)140 (82.35)
Possible30 (16.57)24 (14.12)
Very possible5 (2.76)6 (3.53)
I will smoke in the future (After intervention)<0.001
Not possible171 (94.48)128 (75.29)
Possible9 (4.97)35 (20.59)
Very possible1 (0.55)7 (4.12)
Males87 (48.07)73 (42.94)
Females94 (51.93)97 (57.06)
I will smoke in the future (Intervention group: n=174 males)Before interventionAfter intervention0.015
Not possible66 (75.86)80 (91.95)
Possible18 (20.69)6 (6.90)
Very possible3 (3.45)1 (1.15)
I will smoke in the future (Intervention group n=188 females)Before interventionAfter intervention0.017
Not possible80 (85.11)91 (96.81)
Possible12 (12.77)3 (3.19)
Very possible2 (2.13)0 (0.00)
I will smoke in the future (Intervention group n=362)Before interventionAfter intervention<0.001
Not possible146 (80.66)171 (94.48)
Possible30 (16.57)9 (4.97)
Very possible5 (2.76)1 (0.55)

In the intervention group, reporting that no one is allowed to smoke at home increased after the intervention (47.5% vs 61.3%, p=0.016), whilst in the control group, no significant change was noticed (51.8% vs 50%, p=0.922) (Table 2). Reporting that both parents do not smoke did not statistically change in either the intervention group (48.1% pre vs 42.5% post, p=0.291) or in the control group (50% vs 53.5%, p=0.515). In the intervention group, the existence of a no-smoking ban was statistically significantly associated with the intention to smoke among teenagers (p<0.001) both before and after the intervention (Table 2). The intention to smoke in the next 12 months and parental non-smoking status were neither associated before (p=0.053) nor after the intervention (p=0.435) (Table 2).

Table 2

Association between the no-smoking rule at home and the intention to smoke

No one allowed to smoke at homeOnly visitors allowedAllowedp
n (%)n (%)n (%)
Home no-smoking rule**
Intervention group0.016
Before intervention86 (47.5)39 (21.5)56 (30.9)
After intervention111 (61.3)35 (19.3)35 (19.3)
Control group0.922
Before intervention88 (51.8)39 (22.9)43 (25.3)
After intervention85 (50.0)42 (24.7)43 (25.3)
Parents smokingBoth parents smokeParents do not smoke
Intervention group0.291
Before intervention94 (51.9)87 (48.1)
After intervention104 (57.5)77 (42.5)
Control group0.515
Before intervention85 (50.0)85 (50.0)
After intervention79 (46.5)91 (53.5)
Home no-smoking rule and intention to smoke*No one allowed to smoke at homeOnly visitors allowedAllowed
Intervention group
Will not smoke in the future74 (86.0)36 (92.3)37 (66.1)Before intervention p<0.001
Possible to smoke in the future12 (14.0)3 (7.7)15 (26.8)
Very possible to smoke in the future0 (0.0)0 (0.0)4 (7.1)
Will not smoke in the future109 (98.2)30 (85.7)32 (91.4)After intervention p<0.001
Possible to smoke in the future2 (1.8)4 (11.4)3 (8.6)
Very possible to smoke in the future0 (0.0)1 (2.9)0 (0.0)
Control group
Will not smoke in the future82 (93.2)28 (71.8)30 (69.8)Before intervention p<0.001
Possible to smoke in the future5 (5.7)8 (20.5)10 (23.3)
Very possible to smoke in the future1 (1.1)3 (7.7)3 (7.0)
Will not smoke in the future77 (90.6)27 (64.3)24 (55.8)After intervention p<0.001
Possible to smoke in the future7 (8.2)12 (28.6)16 (37.2)
Very possible to smoke in the future1 (1.2)3 (7.1)3 (7.0)
Parents smoking and intention to smokeBoth parents smokeBoth parents do not smoke
Intervention group
Will not smoke in the future70 (47.6)77 (52.4)Before intervention p=0.053
Possible to smoke in the future21 (70.0)9 (30.0)
Very possible to smoke in the future3 (75.0)1 (25.0)
Control group
Will not smoke in the future98 (57.3)73 (42.7)After intervention p=0.435
Possible to smoke in the future6 (66.7)3 (33.3)
Very possible to smoke in the future0 (0.0)1 (100.0)

* Adolescents’ intention to smoke in the future (following 12 months); their belief /prediction about the likelihood of themselves to smoke in the following 12 months.

** No smoking rule at home; smoking is not allowed at home as a family rule.

We observed a statistically significant improvement in knowledge acquisition in the intervention group, which was not seen in the control group. Overall, the mean knowledge score was increased by 1.24 standard deviations in the intervention group and by 0.19 standard deviations in the control group, which indicates that the intervention is effective in the improvement of knowledge that leads to the enhancement of the no-smoking rule at home. Further, we observed an increase in non-smoking intention in the intervention group, as students reported that they were unlikely to smoke in the 12 months following the intervention.


We found that the observance of the home nosmoking rule was associated with students’ lower intention to smoke in the future, in agreement with the literature, which documents that environments free from tobacco smoke protect students from exposure to secondhand smoke8,13,22 and from the uptake of smoking behavior3,6,12,15. Furthermore, our school-based intervention contributed to the diffusion of tobacco control messages to students’ parents and to a higher percentage of enforcement of the home no-smoking rule.

The relevant literature shows that parents’ smoking behavior has a significant effect on children smoking choices. Children may also exert a positive impact on parents’ attitudes and behavior as well as influence community norms and serve as vehicles of change23. Genuine open discussions between parents and their children regarding smoking behavior may also have positive outcomes in preventing teenagers’ smoking initiation24.

Experiential learning programs at schools appear effective in engaging students in active participation. In addition, the combination of experiential learning with multidisciplinary approaches using ancient Greek literature, applied in our program, provided students with the opportunity to deviate from the passive role of listener and engage in active participation, developing creative and critical thinking, and providing a level platform for all students to be involved irrespective of their academic performance25-27.

Our study provided evidence that adolescents could act as vehicles of non-smoking messages from school to home. Therefore, our findings could encourage students and children to undertake active roles in educational and community settings to diffuse public health messages in general and tobacco control messages in the home, with family and in community environments.

Strengths and limitations

The strengths of our study are that we used a simple, solid, clear home no-smoking rule as an easily implemented guideline, which has a significant and multi-level effect on adolescents’ and adults’ smoking. We also used an experiential learning method that is a successful approach for engaging students without triggering resistance to transfer knowledge and skills related to tobacco control messages at home, where an innovative component, classic texts of the ancient Greek literature, was added.

The limitations of our study include the use of a convenience student sample and a self-reported questionnaire that may lead to socially acceptable responses. However, although the student participants were enrolled from five schools in the northwest region of Athens based mainly on the collaboration of school directors, we believe that the selected student population may be relatively representative of the students attending public schools in Athens because the student population of the public education system is more or less homogeneous with respect to socioeconomic status and school performance, as research has indicated that socioeconomic status can mediate adolescents’ SHS exposure28. Additionally, most of the literature concerning adolescents’ smoking refers to the use self-reported questionnaires and not biochemical markers, as smoking in adolescence is usually occasional. Moreover, the effectiveness of the intervention was evaluated only at the follow-up at 3 months, however, peer reviewed literature supports that short-term positive finding are usually persistent over longer periods29 . Data collection was available from students who were present on the day of intervention and at follow-up. In accordance with wider literature, absent students tend to demonstrate lower academic achievement and appear more susceptible to addictive behaviours30,31. However, potentially high-risk students were not differentially absent from one of the two comparison groups. Finally, students who did not participate may introduce selection bias since they may originate from households with lower awareness about smoking32. Nevertheless, in our study, the proportion of student smokers was similar in both groups.


The findings of our study support that our experiential learning program was successful in improving adolescent students’ knowledge about smoking. It also re-confirmed that the home no-smoking rule had a measurable and significant effect on a negative intention to smoke among adolescents. Children from a home with indoor smoking restrictions were less likely to express an intention to smoke in the next 12 months. In addition, we documented that students from the intervention group were effective facilitators of tobacco control messages to their home environment, thereby increasing the percentage of home no-smoking rules enforced at follow-up. This is an important finding, showing that students may also be successful transmitters of anti-smoking messages and may have a loud voice on such matters. Increasing the prevalence of home no-smoking rule leads to several advantages including the decrease of second- and third-hand smoke exposure and associated risks for adolescents and adults, decreased risk of smoking uptake from adolescents, and propagation of such rules in homes of students who will become adults and future parents.