INTRODUCTION
Exposure to environmental tobacco smoke (ETS) describes any tobacco smoke exposure other than active smoking and comprises secondhand smoke (SHS) and thirdhand smoke (THS)1. It was identified as a public health problem in the 1986 US Surgeon General’s Report on the adverse health effects of involuntary smoking2. Numerous studies have shown that SHS or ETS exposure has many adverse health consequences2-5. Consequently, governments around the world have implemented laws to prohibit smoking in public and work places3,6. In addition to reducing cigarette smoking, smoking restrictions in public places and houses protected people from the health risks of SHS exposure and were a powerful stimulus to adopt voluntary smoke-free policies in homes and cars7,8. In China, Shanghai was the first city to legally limit indoor smoking in certain public places within the city since March 2010. However, the average daily tobacco consumption in urban areas of Shanghai in 2015 was not significantly different from that in 2010 (14.3±9.0 vs 15.3±28.2 cigarettes/day, p>0.05)9. Therefore, SHS or ETS exposure in private places, such as homes, is still an important issue affecting the health of non-smokers.
THS refers to tobacco smoke toxicants that settle on indoor surfaces, fabrics and dust. It lingers for a long time, well after tobacco smoking has taken place10,11. It can also be re-emitted into the gas phase and undergo chemical transformations as it reacts with ozone12 and nitrous acid13 gases that are commonly present in houses14 and cars7,13. The chemical transformations may yield secondary highly carcinogenic contaminants such as: formaldehyde15, tobacco-specific nitrosamines, 4-(methylnitrosamino)-4-(3-pyridyl)butanal, and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone13,16 as well as tobacco-related toxicants, including volatile N-nitrosamines, aromatic amides, polycyclic aromatic hydrocarbons, and volatile carbonyls16,17.
Previous studies have demonstrated the harmful effects of THS in cells, animal models, and people including children18. THS collected from smokers’ homes contained high levels of nicotine19. THS exposure caused functional alterations and cytotoxicity in both animal and human cells20, including mitochondrial stress, dysregulations of gene expression21 and DNA damage20-23. THS exposure in mice and fetal rats caused changes in liver, lung, skin tissue and behavior24-27. Several studies demonstrated that infants and children are at a higher risk of THS exposure than adults because they breathe faster, have thinner skin, and stay longer in homes and on the floor, where dust is deposited, disturbed, and resuspended in the air18.
One study found that compared to 65% of non-smokers, only 43% of smokers agreed that THS harms children. Moreover, strict prohibition of smoking in homes was more prevalent among non-smokers28. Consequently, parental awareness and beliefs on the impact of THS on children’s health can affect their behavior directly, and determine whether children avoid THS. Therefore, it is necessary to investigate the belief scale of THS among parents. Our study aims to understand the beliefs about THS among parents of primary school children in Shanghai. Our study contributes to promoting smoke-free policies at home.
METHODS
Participants and procedures
We performed a cross-sectional survey in Changjiang Road Primary School in June 2019. The Changjiang Road Primary School is located in Songnan town, which is a medium-sized economic area in Baoshan District, Shanghai. There were 885 children in the primary school, half were non-Shanghai residents, which represents the general situation of primary schools in Baoshan District. The paper-based survey questionnaires were distributed to the pupils by their teachers and were taken home to their parents. If both parents smoked or neither parent smoked, either person could fill in the questionnaire. If only one of the parents smoked, we encouraged the person who smoked to fill in the questionnaire. Our aim was to recruit more smokers in order to better understand their perceptions of THS. All participants were required to sign the informed consent form, which contained information such as the purpose of the survey, the content of the study, the inclusion and exclusion criteria, the possible risks and benefits, and privacy protection. It also stated that participants can withdraw from the survey at any time. The questionnaires were returned to the teachers the next day and then mailed to the research team. The information collection and data import and analysis were completed by two independent researchers, who could not contact the participants. The survey was completely anonymous. The study was approved by the Ethics Committee of School of Public Health, Shanghai Jiao Tong University School of Medicine.
In total, we received 843 questionnaires, a response rate of 95.25%. The correlation coefficient in social psychology was about 0.21 according to the review of general psychology studies29. Based on that, we used the website ‘Understanding Statistical Power and Significance Testing’ (https://rpsychologist.com/d3/nhst/) to calculate the minimal sample size. The significance level was set as α=0.05, power at 1-β=0.90 and effect size Cohen’s d=0.21. The sample size of 843 met the required minimum of 238.
The ‘beliefs about thirdhand smoke’ or BATHS scale
We investigated the participants’ beliefs about THS using the ‘beliefs about thirdhand smoke (BATHS)’ scale30 (Supplementary file Table S1). We translated the BATHS scale into Chinese and carried out the survey in the Chinese population. The scale assesses the THS persistence in the environment (Factor 1) and THS impact on health (Factor 2). Factor 1 includes items describing THS in the building environment, capturing persistence of smoke particles, accumulation of THS, and ineffectiveness of THS reduction by means other than not smoking in the house. Factor 2 includes health impact of THS and transmission of THS through means other than the air30. Participants were asked whether they strongly disagreed, disagreed, not sure, agreed, or strongly agreed, with the statements coded on a scale 1–5.
Smoking behaviors
Several questions were asked to the participants: 1) ‘Do you smoke?’. A smoker was defined as someone who has consumed tobacco at least once in the past year; 2) ‘Do you smoke in front of children?’; and 3) ‘How many people smoke in your family?’.
Information about children
We also asked the participants several questions about their children: 1) ‘Is your child a boy or a girl?’; 2) ‘How old is he or she?’; and 3) ‘Has your child suffered from a respiratory disease in the past 6 months, including cold, pneumonia, bronchitis, asthma, tracheitis, laryngitis or rhinitis?’.
Data analyses plan
We had checked the data and found that <10% of the data were missing. We eliminated rows with missing data when performing the data analyses. Data were verified for normality of distribution and equality of variances by SPSS version 22.0. Descriptive statistics for participant demographics were calculated. The quantitative variables are presented as mean (± SD) and qualitative data are described as frequency and percentage. We performed t-test/ANOVA (normal distribution) or Mann-Whitney U/Kruskal-Wallis test (non-normal distribution) to assess the difference between scale scores by participant characteristics. We then conducted multivariate analysis to explore the factors influencing the BATHS scale and subscale, using the generalized linear model. Independent variables included demographics and variables identified by univariate analysis that had a statistically significant association with the BATHS score. Odds ratios, adjusted for parent gender, parent age, parent education level and family income, were calculated for each dependent variable.
We conducted the exploratory factor analysis to assess the fit of the two-factor solution through principal component analysis (PCA) and calculated Cronbach’s alpha using SPSS 22.0. Significance test was bilateral and the level of statistical significance was set at p<0.05 for all analyses.
RESULTS
Characteristics of participants
Demographics of respondents are shown in Table 1. The mean age of the participants was 39.42 (SD=7.46) years. Over half of the participants were male (54.10%). The majority of participants were married (93.59%), aged <40 years (62.28%), had a high school or higher education level (83.21%), and 80.77% reported an average annual income of ≥50000 RMB (100 Chinese Renminbi about 15 US$). Most participants lived in their own new house (65.09%). More than half of their children reported respiratory diseases in the past six months (60.48%). There were 359 smokers, accounting for 42.86% of the participants.
Table 1
Characteristics of participants and differences in beliefs about thirdhand smoke (THS) scale and subscale scores among families of primary school children in Shanghai, 2019 (N=843)
BATHS scale assessment
The reliability of the 9-item scale measured by Cronbach’s alpha was >0.90 (raw 0.907, standardized 0.912) and the reliability of the subscales was strong (raw/standardized Cronbach’s alpha=0.791/0.807 for Factor 1 THS persistent, 0.877/0.880 for Factor 2 THS health) (Table 2).
Table 2
Reliability assessment and factor analysis of beliefs about thirdhand smoke (BATHS) scale among families of primary school children in Shanghai, 2019 (N=843)
[i] The 9-item scale’s reliability as measured with Cronbach’s alpha was greater than 0.90 (raw 0.907, standardized 0.912) and strong reliability in the subscales (raw/standardized Cronbach’s alpha=0.791/0.807 for Factor 1, and 0.877/0.880 for Factor 2). Factor 1 includes four items related to THS persistence in the environment and Factor 2 includes five items related to THS impact on health.
Univariate analysis for BATHS scale and subscale
First, we performed univariate analysis for BATHS scale assessment (Table 1). Female (female 4.10±0.68, male 3.97±0.70, p=0.006), younger participants (<40 years 4.09±0.71, 40–65 years 3.96±0.64, >65 years 3.18±0.94, p<0.001), participants with higher education (≤ junior high school 3.81±0.76, senior high school 3.96±0.69, undergraduate 4.14±0.65, ≥ Master’s 4.38±0.63, p<0.001), new-house owners (new house 4.09±0.68, secondhand house with redecoration 3.98±0.69, secondhand house without redecoration 3.83±0.73, p<0.001), and non-smokers (smokers 3.91±0.65, non-smokers 4.12±0.72, p<0.001) were more likely to obtain higher scores in the BATHS scale. The annual income of participants (given per person in units of 10000 RMB) also influenced their BATHS scale scores significantly (≤5: 3.89±0.72; >5 and ≤7: 4.03±0.65; >7 and ≤11: 3.97±0.71; >11: 4.21±0.66; p=0.001). Participants whose children suffered from respiratory diseases in the past six months had higher scores (suffered respiratory diseases 4.08±0.70, otherwise 3.96±0.68, p=0.006). The results also indicated that when more smokers lived together they obtained lower scores in the BATHS scale (p<0.001).
We also performed univariate analysis for the BATHS subscale and found that the results of THS impact were almost the same as that of the BATHS scale (Table 1). However, there was no significant difference by sex in THS persistence in the environment.
Multivariate analysis for BATHS scale and subscale
We performed multivariable analysis using a generalized linear model to predict the factors influencing the score of the BATHS scale (Table 3). Model included the following variables: sex, age, education level, smoking status, house situation, numbers of smokers living together, health status of children, and annual income. Regarding beliefs about THS, the overall model was significant (p<0.05). The model illustrated that the BATHS scale scores of participants aged >65 years were lower than for participants aged <40 years (OR=0.476; 95% CI: 0.311–0.728, p=0.001). The BATHS scale scores of participants with a Bachelor’s degree (OR=1.190; 95% CI: 1.020–1.388, p=0.027) and Master’s degree or better (OR=1.449; 95% CI: 1.102–1.906, p=0.008) were higher than for those who had junior high school education or lower. In addition, the results indicated that the scores of residents living in a secondhand house with redecoration (OR=0.882; 95% CI: 0.782–0.995, p=0.041) and secondhand house without redecoration (OR=0.801; 95% CI: 0.698–0.919, p=0.002) were lower compared with those of new-house owners. The results also showed that the scores for participants whose children suffered respiratory diseases in the past six months (OR=1.104; 95% CI: 1.003–1.216, p=0.043) were higher than those whose children had no respiratory diseases.
Table 3
Analysis the factors influencing the score of beliefs about thirdhand smoke (BATHS) using generalized linear model among families of primary school children in Shanghai, 2019 (N=843)
| Variable | Categories | β | Wald χ2 | OR* | 95% CI | p |
|---|---|---|---|---|---|---|
| Sex | Male | 1 | ||||
| Female | 0.052 | 0.876 | 1.053 | 0.945–1.175 | 0.349 | |
| Age (years) | <40 | 1 | ||||
| 40–65 | -0.081 | 2.456 | 0.922 | 0.833–1.021 | 0.117 | |
| >65 | -0.743 | 11.714 | 0.476 | 0.311–0.728 | 0.001 | |
| Education level | ≤Junior high school | 1 | ||||
| Senior high school | 0.042 | -0.306 | 1.042 | 1.102–1.906 | 0.580 | |
| Undergraduate | 0.174 | 4.902 | 1.190 | 1.020–1.388 | 0.027 | |
| ≥Master’s | 0.371 | 7.030 | 1.449 | 1.102–1.906 | 0.008 | |
| Average annual income/person (10000 RMB) | ≤5 | 1 | ||||
| >5 and ≤7 | 0.094 | 1.521 | 1.098 | 0.946–1.274 | 0.217 | |
| >7 and ≤11 | -0.027 | 0.140 | 0.973 | 0.843–1.123 | 0.708 | |
| >11 | 0.135 | 2.982 | 1.145 | 0.982–1.335 | 0.084 | |
| Smoking status | Smoker | 1 | ||||
| Non-smoker | 0.057 | 0.523 | 1.509 | 0.907–1.236 | 0.470 | |
| House situation | Owned new house | 1 | ||||
| Secondhand house with redecoration | -0.126 | 4.187 | 0.882 | 0.782–0.995 | 0.041 | |
| Secondhand house without redecoration | -0.222 | 9.987 | 0.801 | 0.698–0.919 | 0.002 | |
| Number of smokers living together | 0 | 1 | ||||
| 1 | -0.110 | 2.312 | 0.896 | 0.778–1.032 | 0.128 | |
| >1 | -0.036 | 0.123 | 0.964 | 0.787–1.181 | 0.725 | |
| Health status of child | No respiratory diseases | 1 | ||||
| Suffered from respiratory diseases | 0.099 | 4.088 | 1.104 | 1.003–1.216 | 0.043 |
Multivariable analysis for environmental persistence factor of THS in the BATHS subscale (Table 4) revealed that average scores of persons aged >65 years were lower than those aged <40 years (OR=0.506; 95% CI: 0.319–0.801, p=0.004). Scores of THS persistence were increased by 23.1% and 44.7%, respectively, in participants with college education, or Master’s or better, in comparison with those with junior high school education (undergraduate OR=1.231; 95% CI: 1.044–1.453, p=0.014, and ≥Master’s OR=1.447; 95% CI: 1.077–1.945, p=0.014). Simultaneously, scores assessing THS persistence in participants with average annual income >110000 RMB were higher than those with ≤50000 RMB (OR=1.199; 95% CI: 1.017–1.414, p=0.031). Participants living in a secondhand house without redecoration obtained lower average THS health scores than those of a redecorated house (OR=0.786; 95% CI: 0.678–0.912, p=0.002). The model also showed that average persistence in the environment scores for respondents whose children suffered from respiratory diseases in the past 6 months were higher than those whose children did not (OR=1.124; 95% CI: 1.013–1.246, p=0.028).
Table 4
Analysis of the factors influencing the beliefs about thirdhand smoke (BATHS) subscale score of persistence using generalized linear model among families of primary school children in Shanghai, 2019 (N=843)
| Variable | Categories | β | Wald χ2 | OR* | 95% CI | p |
|---|---|---|---|---|---|---|
| Sex | Male | 1 | ||||
| Female | -0.031 | 0.267 | 0.969 | 0.862–1.091 | 0.606 | |
| Age (years) | <40 | 1 | ||||
| 40–65 | -0.093 | 2.793 | 0.911 | 0.816–1.016 | 0.095 | |
| >65 | -0.682 | 8.446 | 0.506 | 0.319–0.801 | 0.004 | |
| Education level | ≤Junior high school | 1 | ||||
| Senior high school | 0.068 | 0.707 | 1.070 | 0.914–1.253 | 0.400 | |
| Undergraduate | 0.208 | 6.082 | 1.231 | 1.044–1.453 | 0.014 | |
| ≥Master’s | 0.370 | 6.000 | 1.447 | 1.077–1.945 | 0.014 | |
| Average annual income/person (10000 RMB) | ≤5 | 1 | ||||
| >5 and ≤7 | 0.116 | 2.022 | 1.123 | 0.957–1.318 | 0.155 | |
| >7 and ≤11 | 0.006 | 0.005 | 1.006 | 0.862–1.173 | 0.944 | |
| >11 | 0.182 | 4.662 | 1.199 | 1.017–1.414 | 0.031 | |
| Smoking status | Smoker | 1 | ||||
| Non-smoker | 0.117 | 1.874 | 1.124 | 0.951–1.328 | 0.171 | |
| House situation | Owned new house | 1 | ||||
| Secondhand house with redecoration | -0.124 | 3.561 | 0.883 | 0.776–1.005 | 0.059 | |
| Secondhand house without redecoration | -0.240 | 10.069 | 0.786 | 0.678–0.912 | 0.002 | |
| Number of smokers living together | 0 | 1 | ||||
| 1 | -0.100 | 1.653 | 0.905 | 0.777–1.054 | 0.199 | |
| >1 | -0.023 | 0.042 | 0.977 | 0.785–1.216 | 0.837 | |
| Health status of child | No respiratory diseases | 1 | ||||
| Suffered from respiratory diseases | 0.116 | 4.846 | 1.124 | 1.013–1.246 | 0.028 |
Generalized linear model evaluating the THS impact on the health factor in the BATHS subscale (Table 5) indicated that the average THS health scores in women were higher than those in men (OR=1.125; 95% CI: 1.004–1.260, p=0.042). Scores of the older people were lower than those of younger people (OR=0.453; 95% CI: 0.290–0.706, p<0.001). Participants with a Master’s, or higher, education level obtained higher scores than those with junior high school education or below (OR=1.445; 95% CI: 1.089–1.918, p=0.011). Participants living in a secondhand house without redecoration had lower scores than owners of a redecorated house (OR=0.817; 95 % CI: 0.807–0.943, p =0.006).
Table 5
Analysis of the factors influencing the beliefs about thirdhand smoke (BATHS) subscale score of health using generalized linear model among the families of primary school children in Shanghai, 2019 (N=843)
| Variable | Categories | β | Wald χ2 | OR* | 95% CI | p |
|---|---|---|---|---|---|---|
| Sex | Male | 1 | ||||
| Female | 0.118 | 4.125 | 1.125 | 1.004–1.260 | 0.042 | |
| Age (years) | <40 | 1 | ||||
| 40–65 | -0.072 | 1.768 | 0.931 | 0.837–1.035 | 0.184 | |
| >65 | -0.793 | 12.198 | 0.453 | 0.290–0.706 | <0.001 | |
| Education level | ≤Junior high school | 1 | ||||
| Senior high school | 0.028 | 0.126 | 1.028 | 0.882–1.198 | 0.723 | |
| Undergraduate | 0.154 | 3.557 | 1.167 | 0.994–1.369 | 0.059 | |
| ≥Master’s | 0.368 | 6.498 | 1.445 | 1.089–1.918 | 0.011 | |
| Average annual income/person (10000 RMB) | ≤5 | 1 | ||||
| >5 and ≤7 | 0.078 | 0.966 | 1.081 | 0.926–1.261 | 0.326 | |
| >7 and ≤11 | -0.058 | 0.577 | 0.944 | 0.813–1.096 | 0.447 | |
| >11 | 0.092 | 1.283 | 1.097 | 0.935–1.287 | 0.257 | |
| Smoking status | Smoker | 1 | ||||
| Non-smoker | 0.008 | 0.009 | 1.008 | 0.858–1.184 | 0.925 | |
| House situation | Owned new house | 1 | ||||
| Secondhand house with redecoration | -0.202 | 3.773 | 0.883 | 0.779–1.001 | 0.052 | |
| Secondhand house without redecoration | -0.124 | 7.651 | 0.817 | 0.708–0.943 | 0.006 | |
| Number of smokers living together | 0 | 1 | ||||
| 1 | -0.114 | 2.283 | 0.893 | 0.770–1.034 | 0.131 | |
| >1 | -0.044 | 0.167 | 0.957 | 0.774–1.183 | 0.683 | |
| Health status of child | No respiratory diseases | 1 | ||||
| Suffered from respiratory diseases | 0.086 | 2.854 | 1.090 | 0.986–1.205 | 0.091 |
DISCUSSION
We investigated the beliefs about THS among parents or grandparents of primary school children in Shanghai in order to provide an evidence base for incorporation of THS actions into tobacco control interventions, in the hope of promoting smoke-free homes. This study indicates that younger people and those who received higher education were more likely to believe that THS would persist in the environment and impact children’ s health, as reported in previous studies31,32. Moreover, it is interesting that people who lived in a new house, compared to those living in rented houses without redecoration, were more likely to believe that THS can persist in the environment and influence children’s health. We also found that participants whose children suffered from respiratory diseases believed that THS can persist in the environment for a long time. However, they were not sure about the health impact of THS. A similar result was observed in the high-income group. Our analysis showed that females were more likely to believe THS impacts the health of their children but not the environmental persistence of THS. The BATHS scale scores were not different between smokers and non-smokers, which was inconsistent with another study that indicated that both current and former smokers disagreed with the adverse impacts of THS on children’s health33.
Strengths and limitations
Our study involved a large number of participants to produce robust results. However, there are several limitations. First, this study was carried out in one primary school in Baoshan District, Shanghai, for convenience. The survey was conducted among families and the smokers in the family were encouraged to take part in the questionnaire, therefore the percentage of smokers in our study does not reflect the national rates. Second, we found that the proportion of men (54.1%) in our study was higher than the national (51.3%), according to the sixth national census34. This could partially explain why the smoking rate in our study was higher compared with the national. Third, cross-sectional data precludes the inference of causality. It is unclear whether these findings could be adapted to other geographical areas or to adults without school-age children. Additionally, the reliance on parent self-report could lead to response biases.
Considerations for the future
In 2007, the Framework Convention on Tobacco Control required a total ban on smoking in public places, including all indoor public places, indoor workplaces, public transport and other outdoor areas in China. However, smoking still occurs in many households. Even in the absence of children, smoking can be harmful because toxic contaminants, generated by smoking, can settle on the surfaces of furniture, on skin, hair and the clothing of family members11-13. Smoke-free homes are defined as homes where no one is allowed to smoke inside, but smoke-free policies in multi-unit housing do not force smokers to use smoke-free facilities, they simply prevent smokers from smoking in settings where SHS affects others through infiltration35. Previous research has indicated that as much as 60% of airflow in multi-unit housing facilities can come from other units36. Therefore, how to carry out a smoke ban in families, exploring family THS exposure, intervening in smoking by family members, and promoting infants’ health by reducing smoke exposure, are vital to tobacco control programs. Our findings provide details and reflections for future improvement and implementation of tobacco control programs.
CONCLUSIONS
Our study shows that older people, males, low-income groups, less educated men, and those who rent houses are less aware of the adverse impacts of THS. Through understanding the status quo of THS beliefs among family members, targeted education can be carried on the risks of THS and family members can be encouraged to change their smoking behavior. These actions will help to establish healthy concepts, reduce the harm of tobacco, and eventually help to achieve a smoke-free family environment.
