Beliefs about thirdhand smoke and voluntary smoking bans in homes and cars: a cross-sectional study in Türkiye, 2023 Authors: Cansu Özbaş, Esra Gültekin Koç, Hakan Tüzün, Asiye Uğraş Dikmen Published: 2026 ABSTRACT AbstractBackgroundIn this study, conducted in Ankara, Türkiye in 2023, the prevalence and determinants of voluntary smoking bans implemented by adults in their homes and vehicles were investigated, as were their relationship with beliefs about THS.MethodsThis cross-sectional study used convenience sampling and included adults aged ≥ 18 years who presented to a university hospital in Ankara in June 2023. Data were collected using a 33-item questionnaire administered via face-to-face interviews. Beliefs about thirdhand smoke were measured using the Turkish-validated Beliefs About Thirdhand Smoke (BATHS) scale, which includes health effects and environmental persistence subdimensions. Complete smoking bans in homes and cars were defined as dependent variables. Adjusted Odds Ratios (aOR) with 95% Confidence Intervals (CI) were calculated using logistic regression. Home models were adjusted for age, income, smoking status, secondhand smoke exposure, and the presence of a smoker in the home; car models were additionally adjusted for gender, education, and the presence of a child in the home.ResultsA total of 57.4% of the participants reported enforcing a complete smoking ban at home and a total of 63.2% reported enforcing a complete smoking ban in their cars. Each 1-point increase in the total BATHS score was significantly associated with smoking bans at home (aOR = 1.043, 95% CI: 1.018–1.068) and in cars (aOR = 1.044, 95% CI: 1.015–1.074). In the absence of smokers at home, the ban is enforced more frequently (home: aOR = 8.91, car: aOR = 8.34). Nonsmokers are more likely to enforce the ban at home (aOR = 1.57) and in cars (aOR = 2.02). A higher level of education increases the likelihood of enforcing a smoking ban in vehicles (aOR = 1.93). Individuals who have not been exposed to secondhand smoke in the past 30 days are more likely to enforce the smoking ban in cars (aOR = 2.15).ConclusionBeliefs about THS predict voluntary smoking bans in private settings. By highlighting the role of THS awareness as a predictor of smoke-free private settings, the study suggests that integrating THS education into tobacco control strategies could help protect nonsmokers and reduce exposure in private environments where public smoke-free laws are not enforced.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12889-026-28087-5.Keywords: Thirdhand smoke, Smoking ban, Smoke-free home, Smoke-free car, Secondhand smoke, Tobacco control, Türkiye Introduction The negative effects of exposure to tobacco smoke on health have been known for a long time, and it has been proven that there is no safe threshold level for such exposure [1]. Second-hand Tobacco Smoke (SHS) is an environmental pollutant caused by both the smoke exhaled by the user and the sidestream smoke emitted from tobacco products [2]. The harmful effects of tobacco smoke continue to linger even after it is no longer visible. “Thirdhand smoke” (THS) refers to the residue formed by the persistent accumulation of tobacco smoke on indoor surfaces, dust, and textiles [3]. These residues can turn into carcinogenic substances such as nitrosamines, vaporize over time and be inhaled again [4]. Experimental and toxicological studies have shown that THS can cause biological damage, such as oxidative stress, DNA damage, and mitochondrial dysfunction. Therefore, the effects of tobacco smoke are not limited to the moment of smoking but pose a risk of long-term environmental exposure [5, 6]. THS particles can remain on surfaces for long periods and cannot be completely removed by ventilation or other measures. In addition, THS residues can be transferred through hand contact. Therefore, the only effective way to reduce exposure is to establish completely smoke-free environments [4]. Türkiye signed the World Health Organization (WHO) Framework Convention on Tobacco Control in 2004 and banned tobacco use in enclosed public places in 2009 [7]. Despite the comprehensive smoke-free legislation and other measures taken in Türkiye, which led to an initial decline in daily tobacco use (from 25.4% in 2010 to 23.2% in 2012), the prevalence has since reversed. Recent data indicate an increase, rising to 26.5% in 2016 and further to 28.3% in 2022 among the population aged 15 and over [8]. Although smoke-free policies implemented in public areas have reduced exposure to tobacco smoke, Global Adult Tobacco Survey (GATS) 2016 data from Türkiye reports that 26.7% of adults are exposed to tobacco smoke at home, whereas 17.4% are exposed in their private vehicles [9, 10]. Voluntary smoking bans in private settings such as homes and cars are measures taken by individuals to partially or completely restrict smoking on their own initiative [11]. Smoke-free private living spaces not only reduce exposure to tobacco smoke but also increase smoking cessation attempts, reduce consumption, and delay the age of first cigarette use [12]. Smoking in vehicles is at least as important as smoking indoors because of the small and enclosed nature of the space. Measurements have shown that particulate matter with a diameter of 2.5 µm or less (PM2.5) levels inside vehicles can exceed dangerous limits [13]. Evidence from studies suggests substantial variability in voluntary smoking bans in private settings. In Europe, a survey conducted in 12 countries reported that 70.2% of adults lived in completely smoke-free homes, with marked differences between countries, ranging from 44.4% in Greece to 84.5% in England [14]. Similarly, a comparative analysis of GATS 2012 and National Adult Tobacco Survey data reported that complete smoking bans were implemented in 83.7% of homes and 78.2% of private vehicles in the United States. In contrast, the corresponding figures in Türkiye were notably lower at 52.5% and 70.9%, respectively [15]. A recent national study further reported that more than half of parents in Türkiye report smoke-free homes and a majority prohibit smoking in cars [16]. It is believed that individuals' awareness and beliefs regarding the potential harms of THS are influential in decisions to ban tobacco in private living environments. A study has shown that parents' beliefs about the effects of THS on children's health are associated with indoor smoking bans [17]. In addition, an observational study investigating changes following the implementation of public smoking bans revealed an association where smokers who continued to allow smoking in their homes had low levels of knowledge about the health effects of tobacco smoke exposure [18]. In this context, it is assumed that beliefs about the long-term effects of THS may shape behaviors such as smoking bans in homes and cars to protect nonsmokers. Based on prior evidence, we hypothesized that awareness of THS may influence the implementation of smoking bans in specific areas. In this context, this study aimed to examine the frequency and determinants of voluntary bans on smoking in homes and cars among adults. Methods Study design A cross-sectional study was conducted between 1–16 June 2023 in Gazi University Faculty of Medicine Hospital, a tertiary-care university hospital in Ankara, Türkiye. All outpatient clinics, except the emergency department, were included. Adult visitors (≥ 18 years) presenting to any outpatient clinic during the data collection period were approached consecutively using a convenience sampling method. All adults who were able to communicate in Turkish and provide verbal consent were invited to participate. The questionnaire was interviewer-administered through face-to-face interviews and required approximately 7–8 min to complete. To avoid clustering, if more than one person from the same household was present, only one household member was included in the study. The sample size was calculated via the OpenEpi program on the basis of a 95% confidence interval, 50% unknown prevalence, 3% margin of error, and a design effect of 1.0. The minimum required sample size was determined to be 1049. Considering an anticipated 20% data loss, face‒to‒face surveys were planned with 1258 individuals. The participant flow diagram is presented in Fig. 1. Of the 1,258 individuals approached, 1,172 agreed to participate and provided usable data, yielding a response rate of 93%. Participants who declined or did not complete the survey (n = 86) were excluded as non-responders. Among the 1,172 completed questionnaires, there were no substantial missing data. Participant flow diagram Study questionnaire and variable definitions Data were collected using a 32-item questionnaire (Supplementary file 1). The instrument was piloted with 10 adults to ensure clarity and comprehensibility. Data collectors received standardized training from the research team prior to fieldwork. The study's dependent variable was the presence of complete smoking bans at home and in cars. Home smoking ban was assessed with the item: “What is the smoking rule inside your home (excluding open areas such as balconies that do not allow smoke to enter the house)?” Car smoking ban was assessed with the item: “If your household owns a car, what is the smoking rule inside the car?” Participants who reported having no household car were excluded only from the car-related analyses, while they remained included in all other analyses. Smoking rules were categorized as complete bans if smoking was not allowed at any time in any area, no/partial bans if smoking was allowed either in certain areas or at specific times, or permitted throughout. Car-related analyses were restricted to participants who owned a private vehicle belonging to household. Car ownership was assessed via self-report, and only car owners (n = 876) were included in the analyses examining smoking bans in cars. All participants (n = 1,172), regardless of vehicle ownership, were included in analyses related to home smoking bans. Beliefs about THS were assessed via the Beliefs About Thirdhand Smoke (BATHS) scale, developed by Haardörfer et al. in 2017, which consists of nine items covering two subdimensions: perceived health effects and environmental persistence. Each item is rated on a 5-point Likert scale. The total scores range from 9–45, with higher scores indicating stronger belief in the harmful effects of THS. The original Cronbach’s alpha for the scale was 0.91, with 0.88 for each subdimension [19]. Turkish validation was conducted by Önal et al., who reported Cronbach's alpha values of 0.68 (environmental persistence), 0.67 (health effects), and 0.71 (total) [20]. The internal consistency of the BATHS scale was assessed in our sample. Cronbach’s alpha was 0.845 for the total scale, 0.794 for the health subscale, and 0.729 for the persistence subscale, indicating good internal reliability. Beliefs about secondhand smoke (SHS) were assessed using four items evaluating participants’ perceptions of SHS-related health risks (e.g., risk of serious illness, cardiovascular disease, lung cancer, and respiratory illnesses in children). Independent variables included gender, age, education level, income, presence of children at home, smoking status, exposure to secondhand smoke, presence of smokers in the household, and BATHS scores. Participants were divided into five age groups (18–24, 25–34, 35–44, 45–54, ≥ 55). Education was grouped into three categories: low (primary school and below), moderate (high school graduate), and high (university degree or higher). Income levels were self-reported and categorized as low when income was less than expenses, moderate when income was equal to expenses, and high when income exceeded expenses. The presence of a child in the household was assessed by asking participants whether anyone under the age of 18 was living in their home. Smoking status was determined by asking participants about their history, and responses were categorized as: Current Smoker (smoking daily or occasionally within the last 30 days), Former Smoker (having smoked > 100 lifetime cigarettes but quit for > 30 days), or Never Smoker (< 100 lifetime cigarettes). Exposure to secondhand smoke during the last month was assessed overall, without restricting the question to a specific setting such as home or public places. Having a smoker in the household was assessed by asking participants whether anyone living in the household, including themselves, is a current smoker, regardless of whether they smoke inside the home. Ethics approval and consent to participate This study was performed in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Gazi University Ethics Committee, Türkiye (approval number: 2023–681). Verbal informed consent was obtained from all participants before data collection. Statistical analysis All the statistical analyses were conducted via SPSS Statistics for Windows, version 20.0. Descriptive statistics are presented as the means with standard deviations for continuous variables and frequencies with percentages for categorical variables. The normality of continuous variables such as the BATHS score was assessed via histograms and analytical methods. Parametric tests were used when the normal distribution criteria were met. Independent samples t-tests were used to compare BATHS scores according to smoking habits. Chi-square tests were used to assess associations between categorical variables and smoking bans. Table 2 presents only the variables retained in the final step of the model; therefore, some predictors included in the adjustment set (e.g., presence of a child) do not appear in the final table despite being entered into the model. Table 2Analysis of the association of the factors with complete smoking ban using a logistic regression model, Ankara, Turkey, 2023Complete Smoking Ban (Home) (n = 1172)Complete Smoking Ban (Car) (n = 876)aOR (95% CI)+aOR (95% CI)#BATHS Total1.043 (1.018–1.068)**1.044 (1.015–1.074)**Education Status Low-1.00 Moderate-1.546 (0.955–2.501) High-1.935 (1.237–3.027)**Smoking Status Current1,001.00 Former1.233 (0.799–1.901)2.510 (1.473–4.279)** Never1.574 (1.140–2.173)**2.021 (1.387–2.946)***Exposure to Secondhand Smoke Yes-1.00 No-2.155 (1.409–3.296)***Smokers in the home Yes1.001.00 No8.910 (6.202–12.801)***8.341 (5.151–13.507)***Model estimates obtained using logistic regression+OR for Complete smoking ban (home) was adjusted for age, income, smoking status, exposure to secondhand smoke, and smokers in the home#OR for Complete smoking ban (car) was adjusted for gender, age, education status, a child in the home, smoking status, exposure to secondhand smoke, and smoker in the home*p < 0.05**p < 0.01***p < 0.001 Analysis of the association of the factors with complete smoking ban using a logistic regression model, Ankara, Turkey, 2023 Model estimates obtained using logistic regression +OR for Complete smoking ban (home) was adjusted for age, income, smoking status, exposure to secondhand smoke, and smokers in the home #OR for Complete smoking ban (car) was adjusted for gender, age, education status, a child in the home, smoking status, exposure to secondhand smoke, and smoker in the home *p < 0.05 **p < 0.01 ***p < 0.001 Multivariate logistic regression analysis was performed to identify the independent factors associated with complete smoking bans at home and in cars. The model included variables that had a p-value < 0.25 in univariate analyses and/or were identified as relevant predictors in prior literature. Therefore, variables such as age, sex, education level, income status, smoking status, secondhand smoke exposure, presence of a child in the household, and having a smoker in the household were entered into the regression models. Separate models for the total BATHS score and its subdimensions were constructed to examine whether associations differed across specific domains of beliefs about THS. The statistical significance level was set at a p-value < 0.05. The completed STROBE checklist is provided as Supplementary file 2. Results A total of 1172 participants were included in the study (Fig. 1). The mean age was 43.24 (SD = 15.86) years, and 50.8% of them were male. A total of 47.0% of the participants had a higher education level, and 34.8% were current smokers. A total of 61.1% of the participants were living with at least one smoker, and 76.4% reported secondhand smoke exposure last month. Among all participants, 74.7% (n = 876) reported owning a car. Car ownership differed significantly by gender and income level, with higher ownership among men and those with higher income levels. Specifically, 71.6% of women and 78.0% of men owned a car. In comparison, car ownership was reported by 86.4% of participants with high income level, 72.0% of those with moderate income level, and 65.5% of those with low income level. In contrast, car ownership did not differ by smoking status. (p = 0.904). Table 1 presents unadjusted (crude) odds ratios from bivariate analyses. Overall, 57.4% of participants reported a complete smoking ban at home, and 63.2% of car owners reported not allowing smoking in their vehicles. Smoking bans did not differ significantly by gender in either setting. However, age was significantly associated with smoking bans: participants aged 18–24 years were more likely to report complete bans at home and in cars compared with older age groups. Higher education level was associated with a greater likelihood of implementing smoking bans in cars, whereas higher income level was associated with the presence of home smoking bans (Table 1). Table 1Association between participant characteristics and ımplementation of smoking rules at home and in cars, Ankara, Turkey, 2023n(%)Home Smoking Rulesn(%)Car Smoking RulesComplete BanOR (95% CI)Complete BanOR(95% CI)BATHS Total, Mean (SD)1172 (100.0)34.9 (5.8)1.067 (1.045–1.089)876 (100.0)34.8 (5.8)1.065 (1.039–1.091)Gender, n (%) Male577 (49.2)327 (56.7)1.00450 (51.3)271 (60.2)1.00 Female595 (50.8)346 (58.2)1.062 (0.843–1.339)426 (48.7)283 (66.4)1.307 (0.992–1.722)Age (years), n (%) 18–24200 (17.1)127 (63.5)1.00164 (18.7)109 (66.5)1.00 25–34225 (19.2)121 (53.8)0.669 (0.453- 0.987)*167 (19.1)97 (58.1)0.699 (0.447–1.093) 35–44179 (15.3)92 (51.9)0.608 (0.403–0.917)*133 (15.2)73 (54.9)0.614 (0.383–0.983)* 45–54244 (20.8)133 (54.5)0.689 (0.470–1.010)196 (22.4)124 (63.3)0.869 (0.562–1.343) ≥ 55324 (27.6)200 (61.7)0.927 (0.644–1.335)216 (24.7)151 (69.9)1.172 (0.748–1.812)Education Status, n (%) Low265 (22.6)145 (54.7)1.00158 (18.0)89 (56.3)1.00 Moderate356 (30.4)201 (56.5)1.073 (0.779–1.478)255 (29.1)151 (59.2)1.126 (0.753–1.682) High551 (47.0)327 (59.3)1.208 (0.899–1.624)463 (52.9)314 (67.8)1.634 (1.129–2.365)**Income, n (%) Low238 (20.3)122 (51.3)1.00156 (17.8)91 (58.3)1.00 Moderate604 (51.5)350 (57.9)1.310 (0.969–1.771)435 (49.7)275 (63.2)1.228 (0.845–1.783) High330 (28.2)201 (60.9)1.482 (1.058–2.075)*285 (32.5)188 (66.0)1.384 (0.926–2.069)Child in the home (≤ 18), n (%) Yes438 (37.4)249 (56.8)1.00359 (41.0)237 (66.0)1.00 No734 (62.6)424 (57.8)1.038 (0.817–1.319)517 (59.0)317 (61.3)0.816 (0.616–1.081)Smoking Status, n (%) Current408 (34.8)137 (33.6)1.00304 (34.7)115 (37.8)1.00 Former200 (17.1)136 (68.0)4.203 (2.929–6.032)***152 (17.4)120 (78.9)6.163 (3.915–9.701)*** Never564 (48.1)400 (70.9)4.825 (3.667–6.347)***420 (47.9)319 (76.0)5.191 (3.761–7.164)***Exposure to Secondhand Smoke, n (%) Yes895 (76.4)495 (55.3)1.00677 (77.3)400 (59.1)1.00 No277 (23.6)178 (64.3)1.453 (1.099- 1.920)**199 (22.7)154 (77.4)2.370 (1.644- 3.416)***Smokers in the home, n (%) Yes716 (61.1)273 (38.1)1.00542 (61.9)246 (45.4)1.00 No456 (38.9)400 (87.7)11.591 (8.436–15.926)***334 (38.1)308 (92.2)14.254 (9.230–22.011)* Total1172 (100)673 (57.4)-876 (100)554 (63.2)-SD Standart Deviation*p < 0.05**p < 0.01***p < 0.001 Association between participant characteristics and ımplementation of smoking rules at home and in cars, Ankara, Turkey, 2023 SD Standart Deviation *p < 0.05 **p < 0.01 ***p < 0.001 Never smokers were significantly more likely than current smokers to implement voluntary smoking bans both at home and in vehicles. In addition, participants who were not exposed to secondhand smoke in the past 30 days were more likely to enforce smoking bans in both settings (Table 1). Participants’ beliefs about secondhand smoke (SHS) were also assessed. Most respondents agreed that SHS causes serious illness in non-smokers (93.9%, n = 1101). Additionally, 89.2% (n = 1045) agreed that SHS can lead to cardiovascular diseases in adults, 86.2% (n = 1010) agreed that it can cause lung cancer in adults, and 95.0% (n = 1113) agreed that it can cause respiratory illnesses in children. Overall, 61.1% of participants reported smoking in their homes. The absence of smoker in the household was the strongest predictor of voluntary smoking bans, being associated with substantially higher odds of implementing bans both at home and in cars (Table 1). The mean total BATHS score was 33.99 (SD = 5.79), with mean subscale scores of 18.62 (SD = 3.57) for health effects and 15.36 (SD = 2.85) for environmental persistence. Participants who reported having complete smoking bans at home and in cars had significantly higher mean BATHS scores than those without bans. For the complete smoking ban at home, the model was adjusted for age, income, smoking status, secondhand smoke exposure, and smoker in the home. For the complete smoking ban in cars, the model was adjusted for gender, age, education level, presence of a child in the home, smoking status, secondhand smoke exposure, and smoker in the home. In multivariable logistic regression analyses, higher total BATHS scores were independently associated with increased odds of implementing smoking bans both at home and in cars (Table 2). Higher education level was associated with car smoking bans, while never smoking status and the absence of smoker in the household remained strong independent predictors of smoking bans in both settings. Former smokers and those not exposed to secondhand smoke were more likely to implement smoking bans in cars (Table 2). When examined separately, both BATHS subscales were significantly associated with smoking bans in univariate analyses. In multivariable logistic regression models, beliefs about the health effects of thirdhand smoke remained independently associated with complete smoking bans both at home and in cars. In contrast, beliefs about environmental persistence remained significantly associated with home smoking bans but not with bans in cars after adjustment for covariates (Table 3). Table 3Association between BATHS subscale scores and complete smoking bans, Ankara, Türkiye, 2023Complete Smoking Ban (Home) (n = 1172)Complete Smoking Ban (Car) (n = 876)OR (95% CI)aOR (95% CI)+OR (95% CI)aOR (95% CI)#Model 1 BATHS (Health Effect)1.082 (1.047–1.118)***1.046 (1.007–1.087)*1.103 (1.060–1.147)***1.083 (1.034–1.134)**Model 2 BATHS (Persistence in the Environment)1.151 (1.103–1.201)***1.104 (1.051–1.159)***1.110 (1.056–1.166)***1.051 (0.992–1.115)Model estimates obtained using logistic regression+OR for Complete smoking ban (home) was adjusted for age, income, smoking status, exposure to secondhand smoke, and smokers in the home#OR for Complete smoking ban (car) was adjusted for gender, age, education status, a child in the home, smoking status, exposure to secondhand smoke, and smoker in the home*p < 0.05**p < 0.01***p < 0.001 Association between BATHS subscale scores and complete smoking bans, Ankara, Türkiye, 2023 Model estimates obtained using logistic regression +OR for Complete smoking ban (home) was adjusted for age, income, smoking status, exposure to secondhand smoke, and smokers in the home #OR for Complete smoking ban (car) was adjusted for gender, age, education status, a child in the home, smoking status, exposure to secondhand smoke, and smoker in the home *p < 0.05 **p < 0.01 ***p < 0.001 Discussion In this study, over half of the participants reported having complete smoking bans in their homes (57.4%), and nearly two-thirds reported having smoking bans in their cars (63.2%). Never smoking status and the absence of smoker in the household emerged as the strongest predictors of smoke-free homes and vehicles. In addition, higher education level was associated with smoking bans in cars, and lack of exposure to secondhand smoke was related to a greater likelihood of implementing vehicle smoking bans. Stronger beliefs about thirdhand smoke were independently associated with the implementation of complete smoking bans in both settings. These findings suggest that beliefs about thirdhand smoke may play a role in voluntary decisions to implement smoke-free rules in private settings. Previous research in Türkiye on smoking bans in private settings has been limited and has largely focused on specific subgroups, particularly parents. In 2025, Odacı and Kitiş reported similar figures among Turkish parents: 57.8% enforce bans at home, and 75.8% enforce bans in cars [16]. Comparable complete home smoking ban prevalence rates have been reported in Spain (57.4%), Italy (61%), and Armenia and Georgia (72.9%), although lower rates have been reported in Thailand (38.7%) and Chongqing, China (41.7%) [21–25]. A systematic review reported that 45.2% of parents enforced a smoking ban at home, a slightly lower frequency than in this study [26]. The slightly higher frequencies reported in this study may indicate that Türkiye's smoke-free zone regulation, which has been in operation in closed public areas since 2009, has had an impact on social norms, encouraging voluntary bans. Previous study have confirmed that bans enforced in public areas reduce the social acceptability of tobacco use and decrease tobacco use in private settings [27]. Despite this, 4 out of 10 participants still live in private living spaces where smoking is permitted. Studies have shown similar results, indicating that smoking bans in private settings is a vital strategy for both protecting individuals from the immediate effects of tobacco smoke and reducing the number of tobacco users. The study highlights significant differences between subgroups in terms of smoking bans, revealing that only one-third of smokers enforce a complete ban on smoking in their homes. Similarly, a study incorporating EUREST-PLUS survey results from six European countries reported that this frequency was 26.5% [28]. We found that never smokers were 1.5 times more likely to have a completely smoke-free home than smokers were. Other studies have shown that the prevalence of households with complete indoor rules was 2.6 times greater among never smokers [21]. Jankowski et al. reported that those with more prominent odds of never smoking were 19.1 times more likely to adopt a smoke-free home rule [29]. This may be due to nonsmokers' perceptions of the risks of smoke exposure; therefore, they tend to implement voluntary bans more. Smokers may have biased perceptions that smoke exposure is not hazardous to those with whom they live [30]. These results may suggest that efforts to prevent smoking initiation not only benefit those who do not start smoking but also have a multiplier effect by protecting other nonsmokers from tobacco smoke. According to the bivariate analysis, younger participants were more likely to implement smoking bans in both settings, although this difference was not significant in the multivariate analysis. While some previous studies have reported a correlation with age, others have not [30, 31]. The enforcement of smoking bans reduces the perceived acceptability of smoking among the young population [32]. Additionally, banning tobacco use in outdoor areas has been reported to lead to the perception that smoking has become more difficult among the young population [33]. The reason why studies differ may reflect contextual and cultural factors, as well as variations in social norms and tobacco-use patterns among different age groups. Educational level was a significant predictor of car bans. Participants with higher education levels were 1.93 times more likely to prohibit smoking in their cars, echoing previous research that indicates socioeconomic factors contribute to the adoption of smoke-free policies in private domains. Shojima et al. reported that the proportion of people with higher education levels was 1.3 times greater in places with comprehensive smoking bans [34]. Other studies also reported that subjects who received higher education were more likely to adopt a complete smoking rule at home, although no significant association was found in this study [17, 35]. On the other hand, these results highlight the importance of targeting groups with low educational levels in terms of THS exposure in the context of tobacco control. A troubling finding is that there is no statistically significant difference between having a minor at home and smoking bans. Previous studies have shown that the presence of children in the home is a determinant of smoking bans and have supported the idea that parental concern encourages protective behavior [36]. The literature has shown a relationship between parents' beliefs about THS and their attitudes toward prohibiting smoking in private settings [35]. These findings suggest that when parents become informed and believe in the effects of THS, particularly for children, they are more likely to strive to create a smoke-free environment. These findings may reflect insufficient awareness among parents or adult household members regarding the harmful effects of tobacco smoke exposure on children. In a country such as Türkiye, where both tobacco use and the proportion of young people are high, this finding may indicate that families with children should be prioritized in efforts to combat the effects of THS. In line with previous research, we found that higher BATHS total and health effect scores were associated with increased odds of complete home and car smoking bans. In contrast, environmental persistence scores were associated with only complete home bans. When people have stronger beliefs that THS is harmful to health and persists in the environment, they are more likely to have voluntary rules strictly prohibiting smoking in private settings [21, 24, 35, 37]. The effect of THS belief scores was significant in all analyses examined in the multivariate models, except for the association between environmental persistence and car smoking bans. This suggests that THS belief may have a widespread effect even when adjustments are made for other explanatory variables and smoking behaviors. In this context, increasing awareness of THS exposure, among the objectives of tobacco-related health promotion programs, will help them be more successful. Those with no exposure to secondhand smoke during the previous month were more likely to have complete bans in both settings. We found that individuals with no exposure to secondhand smoke were 2.1 times more likely to use car smoking bans in multivariate analysis. It has previously been reported that nonsmokers with a greater proportion of friends who smoke are less likely to have bans [30]. This finding may be due to individuals who are exposed accepting or becoming accustomed to this situation. Not having smokers in the home is found to be a strong correlate of having a smoke-free home and car policy, similar to previous research [38]. Studies have shown that the main independent predictors of smokers reporting complete smoking bans include having other nonsmoking adults in the household [39, 40]. In the USA, in households where no one smokes, the national prevalence of smoke-free home rules was 91.4%, whereas for those with at least one adult smoker, the voluntary home smoking ban was half [41]. A notable finding of this study is that the determinants of home and car smoking bans were similar in some respects but differed in others. Complete smoking bans were slightly more frequent in cars than in homes. This may reflect the smaller and more enclosed nature of vehicles, where smoke exposure is more immediate and visible. Stronger overall THS beliefs and beliefs about the health effects of THS were associated with both home and car bans, suggesting that health-related risk perception may influence protective rules across different private settings. However, beliefs about environmental persistence remained independently associated only with home smoking bans. This difference may indicate that participants perceive THS residue and long-term contamination as more relevant to the home environment. In contrast, car smoking bans appeared to be more strongly associated with education level and recent secondhand smoke exposure, suggesting that vehicle-related rules may be shaped more by immediate exposure concerns and socioeconomic awareness than by beliefs about long-term environmental persistence. This study has several limitations. First, the sample consisted exclusively of adults who presented to outpatient clinics of a tertiary-care university hospital in Ankara. Individuals seeking care in such settings may differ from the general population in terms of socioeconomic status, chronic disease burden, health awareness, and smoking-related behaviors. The use of convenience sampling may also limit generalizability. Therefore, our findings may not reflect either the broader population of Ankara or the national Turkish population, introducing potential selection bias. Second, since car-related analyses included only vehicle owners, individuals with lower socioeconomic status were underrepresented in these models. This may introduce a form of selection bias. However, all participants, regardless of vehicle ownership, were included in analyses of home smoking bans, minimizing bias in the primary outcome of the study. Third, all data were self-reported and thus subject to social desirability bias, particularly for smoking-related behaviors. The cross-sectional design, which allows for establishing associations but not for inferring causality and reverse causation is possible. For example, stronger beliefs about THS may lead to smoking bans, but existing bans may also reinforce these beliefs. Although increasing THS awareness has been proposed as a potential intervention strategy, whether THS awareness interventions can truly modify beliefs or change behaviors remains uncertain and should be confirmed through intervention studies. Finally, unmeasured confounders—such as general health consciousness, religiosity, cultural norms, and household dynamics—may influence both THS beliefs and smoking-related behaviors. However, the relatively large sample size and validated measurement tools strengthen the internal validity of our findings. Conclusions In this study, beliefs about THS, smoking abstinence, and the lack of smokers in the home were among the most significant independent predictors of complete smoking bans in private settings. The findings suggest that incorporating beliefs about THS into health communication strategies could foster more smoke-free homes and vehicles. In cases where smokers reside in the home, creating smoke-free private living spaces is possible by enhancing smokers' awareness of the long-term effects of tobacco smoke, particularly passive smoking, to protect children and other nonsmoking adults within the home. Our findings indicate that developing interventions aimed at increasing knowledge and awareness of individuals with lower educational attainment from a sociodemographic standpoint might be advantageous. A concerning outcome of this study is the lack of a connection between the presence of children in the home and policies regarding smoking in the home or car. Raising parents' awareness of the invisible effects of smoking could contribute to the creation of more smoke-free environments. Supplementary Information Supplementary Material 1. Supplementary Material 2. Acknowledgements Not applicable. Abbreviations Third-hand smoke Beliefs About Thirdhand Smoke Environmental tobacco smoke Global Adults Tobacco Survey Authors’ contributions C.Ö. and A.U. made contributions to the conception and design of the work. C.Ö., E.G., H.T. analysed and interpreted data. C.Ö. and E.G. wrote the main manuscript text and H.T. prepared Tables 1, 2, 3. All authors reviewed the manuscript. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data availability Data that support the findings of this study have been deposited: [https://doi.org/10.5281/zenodo.15804857]. Declarations Ethics approval and consent to participate This study was performed in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Gazi University Ethics Committee, Türkiye (approval number: 2023–681). Verbal informed consent was obtained from all participants before data collection. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement Data that support the findings of this study have been deposited: [https://doi.org/10.5281/zenodo.15804857].