OUP user menu

Home smoking bans in Finland and the association with child smoking

Susanna U. Rainio, Arja H. Rimpelä
DOI: http://dx.doi.org/10.1093/eurpub/ckm098 306-311 First published online: 9 October 2007

Abstract

Background: Few studies in Europe have investigated home smoking bans and their association with child smoking. Methods: A nationwide survey of 12 to 18-year olds in 2005 (n = 6503, response rate 66%) was used to study home smoking bans in Finland. Multinomial logistic regression analyses were used to study association of home smoking bans and child smoking. Results: Of the respondents, 58% reported a total ban, 27% a partial ban, 4% reported no ban and 10% chose the option ‘I cannot say’. The lack of total ban was strongly associated with living in non-intact families, parents’ lower educational level, parental smoking and parents’ permissive attitude towards child smoking. Moreover, partial or no ban increased the likelihood of being a daily smoker. In the multinomial logistic regression model, the odds ratios (ORs) for children's daily smoking, adjusted for sociodemographic factors, parental smoking and their permissive attitude, were OR 2.9 [95% confidence interval (CI) 2.3–3.6] for partial ban and OR 14.3 (8.6–23.7) for no ban. In families where both parents smoked, the adjusted ORs were correspondingly 1.5 (95% CI 0.7–3.0) and 2.9 (95% CI 1.1–7.8). Conclusions: Home smoking bans will contribute towards a reduced risk of child smoking even when parents smoke. Tobacco control legislation needs to be enhanced with measures promoting awareness of the benefits achievable through strict home smoking bans. Families characterized by lower socioeconomic status and smoking parents are particular target groups.

  • children
  • home smoking ban
  • smoking

Finland has a long history in national smoke-free policy. Since the first Tobacco Act in 1977,1 legislation has been a powerful tool to reduce environmental tobacco smoke exposure in public indoor places.2 In 1995, Tobacco Act was amended to protect workers from environmental tobacco smoke, and in 2000, it was extended to restaurants and cafeterias.1 Today, smoke-free public environments are a norm widely accepted by the public.1,,3 Exposure to tobacco smoke in private homes has not been controlled by legislation, but emerging evidence suggests that enforcement of smoke-free legislation can have an indirect effect on smoking restrictions in the homes, too.4

The prevalence and trends of home smoking bans or smoke-free homes have mainly been studied in the USA,5–7 Canada,8 the UK,9 Australia10,,11 and New Zealand.12,,13 There is a paucity of data on smoking restrictions in Finnish homes as well as elsewhere in Europe, particularly as seen through the eyes of children. However, given the strong national tobacco control legislation, public acceptance of smoking restrictions in public places and workplaces and the increasing awareness of the health hazards of passive smoking, we might expect to see comprehensive smoking bans adopted by an overwhelming majority of the Finnish homes.

Home smoking bans represent a promising and relatively new approach to smoking prevention for children. Research has shown that strict home smoking bans contribute to preventing and reducing children's smoking (decreased smoking initiation and daily smoking, increased smoking cessation).14–22 In Finland, prevention of adolescent smoking has mainly focused on legislative measures, mass communication campaigns and health education in schools. The role of family has been largely overlooked.23 A national health policy statement in Finland from 2001 set goals to reduce smoking rates among adolescents to <15% by the year 2015.24 Despite the encouraging decline in adolescent smoking in the 2000s, daily smoking prevalence continues to remain at an unacceptably high level; 22% of 14–18-year-old Finns smoke daily.3

Besides contributing to child smoking, home smoking bans have been associated with several sociodemographic and tobacco-related family characteristics. The likelihood of having such bans or a smoke-free home has proved to be distinctly greater in families whose parents have higher educational level, do not smoke and are both the child's biological parents.12,,25–28

The present study focuses on children's own reporting of home smoking bans. The purpose is to study the prevalence of home smoking bans and its variation by the sociodemographic background and tobacco-related factors in their family. Moreover, we will examine whether home smoking bans are associated with experimental and daily smoking of the child.

Material and methods

Sample

We used nationwide data from the Adolescent Health and Lifestyle Survey of 2005. A self-administered, 12-page structured questionnaire was mailed to 9918 adolescents in February 2005, with two re-inquiries to non-respondents. The filling instructions included invitation to parents to familiarize themselves with the questions before their child's answering. The sample of 12-, 14-, 16- and 18-year-olds was drawn from the National Population Register Centre on the basis of particular dates of birth, so that all Finns born on sample days were included. The Ethical Committee of the Department of Public Health, University of Helsinki accepted the study protocol. Of them, 6503 youths responded, the response rate was 66%. The number of respondents and response rates in boys by age were: 395 (67%) of 12-year-olds, 1092 (64%) of 14-year-olds, 806 (57%) of 16-year-olds and 681 (49%) of 18-year-olds. In girls, the corresponding rates were: 418 (75%) of 12-year-olds, 1189 (75%) of 14-year-olds, 985 (76%) of 16-year-olds and 937 (68%) of 18-year-olds.

Measures

Home smoking ban

Home smoking ban variable was based on the question: ‘Is smoking permitted at your home? ‘with options: (i) no one is permitted to smoke anywhere; (ii) smoking is permitted in some places only or sometimes; (iii) smoking is permitted freely at my home and (iv) I cannot say. Those reporting smoking was not permitted anywhere were assumed to have a ‘total ban’ on smoking in their home. Those who reported smoking was permitted in some places only/sometimes were assumed to have a ‘partial ban’, and those permitted to smoke freely in their homes were assumed to have ‘no ban’. A missing data rate in this variable was 1% (n = 86).

Sociodemographic and tobacco-related factors

Respondents were characterized by several sociodemographic and two tobacco-related factors according to child self-reports. Sociodemographic factors included family structure, father's education, mother's education and urbanization level of residence. Family structure was categorized as intact family (child living with his/her own mother and own father) and other type of family. Father's and mother's education was divided into three categories: (i) low (9 years or less in education), (ii) middle (9–12 years) and (iii) high (over 12 years). Urbanization level of residence was defined by the population density: (i) capital city area (the capital Helsinki and the adjoining towns), (ii) large cities (population over 100 000), (iii) other towns, (iv) rural municipalities (densely populated areas, such as small towns/village in rural areas) and (v) other sparsely populated rural areas (isolated homesteads in rural areas).

Tobacco-related factors were described by variables of parental smoking and parental permissiveness toward child smoking at home. Parental smoking was based on child's self-reported responses to the separate questions regarding mother's and father's smoking and classified into five categories: (i) two smoking parents, (ii) smoking father, (iii) smoking mother, (iv) both currently non-smokers but one/both ex-smokers and (v) two never-smoking parents. Exclusion criteria for this variable comprised cases where neither parent's (0.5% of respondents, n = 34) or only one parent's smoking status was reported and cases where either parent's status was reported as having no mother/father/do not know (n = 700).

Parental permissiveness toward child smoking at home as experienced by the child was based on the question: ‘Do your parents allow, or would they allow if you smoked, you to smoke at home?’ with alternatives no/yes/do not know.

Child smoking

Child smoking was divided into daily/experimental/never. Daily smoking was defined as having smoked over 50 cigarettes in all, having smoked during the past week and at least once per day or more often. Respondents reporting that they smoked at least once per week or more often but not daily were also defined as daily smokers. Tobacco experimenters had smoked at least one cigarette in their lifetime but were not daily smokers. Never-smokers were defined as never having smoked any cigarette.

Analysis of non-respondents

The respondents were categorized into three groups based on how promptly they returned the questionnaire (original, first re-inquiry and second re-inquiry). It was assumed that the later the person answered the more he/she resembled a non-respondent. There were less children reporting a total ban in the second (57%) and third wave (52%) than in the first wave (60%) (P = 0.001), suggesting that the later the child answered the less likely it was to have a total smoking ban in the home. Child's daily and experimental smoking varied systematically between early and late respondents (P < 0.001 for both smoking groups) so that the later the child answered the more common it was for him/her to report either daily or experimental smoking.

Data analysis

Results were tested by the Pearson's chi-squared tests. A two-sided P-value of <0.001 was considered statistically significant.

A multinomial logistic regression modelling was used to examine the association between home smoking bans and child daily and experimental smoking. Never-smoking children were used as a reference category for both smoking groups. The analysis was divided into three stages. First, the model was adjusted for age and sex. Second, the model was further adjusted for parental smoking. In the third stage, sociodemographic variables (family structure, father's education, mother's education and urbanization level of residence) and the variable of parental permissiveness toward child smoking were adjusted for and entered into the model together with age, sex and parental smoking. The frequency of missing values in variables under study varied between 0.4 and 11.5%. At each stage, variables were entered simultaneously into the model. In order to see whether the smoking ban affects child smoking even when the parents smoke, we studied the association between home smoking ban and child smoking by restricting the analysis to those with two smoking parents.

Results are presented as odds ratios (ORs) with 95% confidence intervals (95%CIs). The statistical analyses were performed using SPSS version 11.5. for Windows software.

Results

Of the respondents, 58% reported total ban, 27% partial ban, 4% no ban and 10% (n = 661) chose the option ‘I cannot say’. Gender differences were small (P = 0.028). The number of total bans reported decreased by age (P < 0.001). (table 1)

View this table:
Table 1

Distribution (%) of home smoking bans by age and sex

The degree of home smoking ban varied significantly by family structure. Children from intact families reported more often total bans than did children from other type of families. The percentage of total bans systematically increased when the level of parental education increased. The degree of home smoking ban did not show any meaningful differences by urbanization level, but the highest differences were found in the ‘I cannot say’ category. (table 2)

View this table:
Table 2

Distribution (%) of home smoking bans by respondents’ sociodemographic characteristics

Respondents who answered with ‘I cannot say’ were more often those whose parents had low level of education, who lived in less urbanized areas (smaller towns, rural regions) and in other than intact families.

Parental smoking was strongly related with home smoking bans. Respondents with former or never-smoking parents reported most often total bans (67% and 83%, respectively). The data also indicated a much greater percentage of total bans among children who thought their parents would not allow them to smoke at home (70%) compared to children who thought smoking at home would be permitted by their parents (15%). However, a notably high proportion of respondents who did not know whether their parents would allow them to smoke at home or not reported either total or partial bans. (table 3)

View this table:
Table 3

Distribution (%) of home smoking bans by two tobacco-related factors in a family

Children who responded with ‘I cannot say’ were most often those whose parent(s) smoked or who did not know their parents’ attitude toward their smoking at home.

When comparing home smoking bans with child's own smoking status, we found that 71% of the never-smoking children, 58% of the experimental smokers and 29% of the daily smoking children reported total bans in their homes. When analysed by age, the distribution remained fairly similar. Of 18-year-old daily smokers, 26% reported total ban and of 14- and 16-year-olds 33% and 30%, respectively, reported total ban.

Association between home smoking ban and child smoking

After controlling for age and sex, both partial ban and no ban were strongly associated with child daily smoking as compared to those who had a total ban and were never smokers. When the data were further adjusted for parental smoking, these strong and statistically significant associations persisted. The adjustment of several other factors somewhat reduced their strength, but the relationships with child daily smoking continued to remain significant. The associations were stronger for daily smoking than for experimental smoking. In the ‘I cannot say’ category, the likelihood of being either a daily smoker or an experimental smoker was increased. (table 4)

View this table:
Table 4

Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for the child daily and experimental smoking by the degree of home smoking ban among 12-18-year-old Finns. In the multinomial logistic regression model, the reference group for daily and experimental smoking was never-smoking child

The association between home smoking ban and child smoking was further studied for families, where both parents were smokers. After adjusting for the variables of age, sex, family structure, father's education, mother's education, urbanization level of residence and parental permissiveness toward child smoking, the ORs for daily smoking were OR = 1.5 (95%CI 0.7–3.0) for partial ban, OR = 2.9 (95%CI 1.1–7.8) (P < 0.05) for no ban and OR = 2.8 (95%CI 1.2–6.5) (P < 0.05) for ‘I cannot say’; and for experimental smoking were OR = 1.1 (95%CI 0.7–1.9), OR = 1.1 (95%CI 0.5–3.0) and OR = 0.73 (95%CI 0.35–1.52), respectively.

Discussion

Our study showed that only slightly more than a half (58%) of the 12 to 18-year-olds reported a total smoking ban in their home. In view of the effective national smoke-free policy over the years, combined with a public opinion in favour of smoking restrictions in public places and workplaces alike,2,,3 it is rather unexpected that home smoking restrictions are as scarce as shown by this study. Living in non-intact families, parents’ lower educational level, parental smoking and parents’ permissive attitude toward child smoking were strongly associated with the lack of a total ban. Children from homes where smoking was not totally banned had a high likelihood of experimenting with smoking and becoming daily smokers. On the other hand, banning smoking in homes where both parents smoked reduced the odds of child daily smoking.

Since little research has been conducted among children, it is difficult to make comparisons with other studies. However, when compared with a similar study conducted in Australia, where 75% of adolescents surveyed indicated that smoking was banned in their homes,22 the percentage observed in our study can be considered low. The median percentage from 14 states in the United States, although based on adult reporting only, indicated that 73.7% of homes were completely smoke-free.29

In Finland, few other than legislative actions have been taken to tackle the problems related to environmental tobacco smoke. Specific campaigns to promote the adoption of home smoking bans have been non-existent for years. In comparison, e.g. in Australia, a large ‘Car and home: smoke free zone’ campaign was launched.30 Another factor complicating the comparison of the figures may be the question used to measure home smoking bans. This study offered a response alternative ‘I cannot say’, an option largely disregarded in previous studies, which may have affected the reporting and consequently our findings.

On the role of sociodemographic and tobacco-related factors in the adoption of total ban in homes, our findings broadly correspond with previous studies.26,,27,31 It has been shown that smoking is marginalizing lower socioeconomic groups.32 This might explain why home smoking restrictions are less common among those population groups. The differences between parents’ educational level, parental smoking and family structure with respect to home smoking bans were clear. Home smoking bans did not substantially differ by urbanization level of residence.

A strong association between the lack of a home smoking ban and child daily smoking was found. A similar pattern was seen for experimental smoking, yet the associations were much weaker than for daily smoking. Similar to other studies18,,22 adjustment for parental smoking somewhat decreased the effect of home smoking bans. On the other hand, our study showed that when the ban was total, children were substantially protected from smoking even in homes in which parents themselves smoked. Although we were able to adjust for several factors in the analysis, some other important factors, e.g. factors related to cultural norms and smoking visitors in homes might also have been considered, but the information was not available for this study. These factors have not received much attention in the literature and would warrant further research to determine how they influence adoption of home smoking bans.

The study has some limitations. The overall response rate was fairly good although the rates were lower in boys than in girls and in the older age groups. Our indirect analysis of non-respondents suggested total bans to be less prevalent among non-respondents, probably leading to a slight overestimation of total bans in this study. We cannot exclude the possibility that child and parental smoking were under- or over-reported, but other studies have shown that adolescents in general give valid reports of their own tobacco use33 as well as of their parents’ smoking.34 Concerning child smoking, true prevalence rates are likely to be somewhat higher than obtained in the survey. Overall, these biases are not likely to change the overall picture significantly.

Another weakness is that the accuracy of self-reported responses to the question of home smoking ban remains uncertain. It is not known how the children understood the question and whether the distribution of certain responses would have been similar if asked from adults. Furthermore, our findings might have been different if we had chosen a different measure of home smoking ban. Dropping ‘I cannot say’ option from the question would have led to lost information.

Responses to home smoking ban can be assumed to be based on two viewpoints: whether the children believed or whether they were aware that there are smoking bans in their homes. Knowledge or belief that such smoking bans exist can be expected to reduce the frequency of ‘I cannot say’ responses. Of the respondents, 10% chose the option ‘I cannot say’. This response may be interpreted as ‘I don't know’ indicating that the issue of home smoking policy has not been discussed within the family. Moreover, the question may not be clear enough for children because, for example, rules for smoking at home are not always congruent. Whether home smoking bans were applied also to smoking visitors was not inquired. Since it is likely that rules for visitors are different, the issue would merit further studies to determine the actual degree of home smoking bans.

To date, studies of home smoking bans have been conducted nearly exclusively by parents or other adult household members. Asking parents to report home smoking bans may modify the results; Mumford et al.35 found discrepancies among household members’ reports of strict home smoking bans, especially in homes with smokers. They concluded that analyses should not solely rely on individual reports, especially in households with smokers and children. Therefore, the importance of obtaining the children's own perspective of home smoking bans is obvious.

Our study provides an overview of home smoking bans in Finland. Despite strong national smoke-free policies, a remarkable number of homes are still without a total smoking ban. Since a majority of these households can be characterized by a lower socioeconomic status, serious consideration is necessary. Moreover, this study shows clearly that the lack of home smoking bans contributes significantly to child smoking, even when adjusted for parental smoking and several other factors. When considering homes with two smoking parents, the total ban can still reduce the likelihood of the child to smoke regularly. Tobacco control programmes primarily designed to minimize exposure to environmental tobacco smoke should also advocate the promotion and strengthening of public awareness of an added benefit achievable through home smoking restrictions, which seem to effectively contribute to child smoking prevention.

Acknowledgements

This study was supported by the Ministry of Social Affairs and Health (the §27 Appropriation of the Tobacco Act), the Competitive Research Funding of the Pirkanmaa Hospital District and the Juho Vainio Foundation. We thank Marja Vajaranta for language revision and Lasse Pere for data management, both from the School of Public Health, University of Tampere.

Conflicts of interest: None declared.

Key points

  • Few studies in Europe have investigated home smoking restrictions, particularly as seen through the eyes of children, and their association with child smoking.

  • Although emerging evidence suggests that enforcement of national smoke-free legislation can have an indirect effect on smoking restrictions in private homes, our results show that in Finland, despite 30 years experience in strong national tobacco legislation, only half of families with adolescent children have adopted total smoking bans in their homes.

  • A noteworthy finding was that children's lower socioeconomic background was strongly associated with a lack of total ban.

  • Children from homes where smoking was not totally banned had a high likelihood of experimenting with smoking and becoming daily smokers. On the other hand, banning smoking in homes where both parents smoked reduced the odds of child daily smoking.

  • Tobacco control programmes primarily designed to minimize exposure to environmental tobacco smoke should also advocate the promotion and strengthening of public awareness of an added benefit achievable through home smoking restrictions, which seem to effectively contribute to child smoking prevention.

References

View Abstract