Correlates of exposure to secondhand smoke (SHS) at home among non-smoking adults in Bangladesh: findings from the ITC Bangladesh survey

Secondhand smoke (SHS), the smoke generated by active smokers, remains a widespread health hazard worldwide [1, 2]. Several studies have clearly linked exposure to SHS, also known as tobacco smoke pollution (TSP) or passive smoking, to a number of health consequences in non-smokers, including lung cancer, heart disease, and asthma in children [1‐3]. The World Health Organization (WHO) has estimated that in 2004, about one-third of adults and 40% of children worldwide were exposed to SHS, and that this exposure caused 1.0% of all deaths and 0.7% of the worldwide burden of disease and disability-adjusted life years in that year [4]. Smoke-free policies for public and private environments are the most effective way to reduce SHS exposure, and therefore decrease the SHS induced health consequences [1].

In response to the growing awareness of the health risks posed by SHS, an increasing number of countries have introduced legislation to reduce or eliminate exposure to SHS in public places. Article 8 of the WHO Framework Convention on Tobacco Control (FCTC) includes guidelines for protection from TSP [5]. These guidelines recommend comprehensive bans on smoking in public places and workplaces in order to achieve 100% smoke-free environments. Households, however, are not protected under the FCTC, thus household smoking bans must be initiated voluntarily. As a result, even as smoke-free restrictions in public places are becoming more widespread, the home remains a predominant source of exposure to SHS, particularly among vulnerable children [1, 4].

During the last decade, a significant amount of research had been conducted addressing the SHS exposure at home. However, the majority of research on household exposure to SHS comes from developed or high income countries, and information about the prevalence of SHS exposure in the home and measures to reduce this exposure in developing countries is limited.

It was argued that the actual conditions of SHS exposure and the burden of the resulting health outcomes differ greatly between developed and developing countries [6, 7]. For example, factors such as the number of smokers in the home, the intensity of the smoke, construction aspects of the home (i.e. space, ventilation), and even cultural factors such as gender dynamics can all influence the experience of and the impact of SHS exposure in the home among non-smokers and children in developing countries. Information on the pattern and correlates of SHS exposure at home in developing country would guide the development of evidence-based educational and policy initiative, specific to developing countries.

With 41.1 million tobacco users in Bangladesh, including 20.9 million people who smoke, [8] exposure to SHS at home is a major public health concern. In Bangladesh, smoking is prohibited in public places and public transportation; however, designated smoking areas are permitted [9]. Data on SHS exposure in Bangladesh is limited; however some information on the prevalence of SHS exposure is available from the 2009 Global Adult Tobacco Survey (GATS) [10]. According to the GATS Bangladesh survey, 63% and 45% of all adult workers were exposed to tobacco smoke at the workplace and in public places, respectively, and 76% and 70% of non-smokers were exposed in these places [10]. In the absence of any data, it could be estimated that a significant proportion of Bangladeshi non-smokers and children are exposed to SHS at home. A 2011 study conducted among Bangladeshi households with children [11] reported a high prevalence (67%) of SHS exposure among children. In this study, only 43.5% of respondents had complete smoking restrictions in the home. This low prevalence of home smoking restrictions suggests that many non-smokers who live with at least one smoker in the home are regularly exposed to SHS.

The objective of this study was to examine the patterns of SHS exposure at home and its demographic and behavioural correlates in non-smoking adults in Bangladesh. Furthering understanding of the patterns of SHS exposure at home among adult non-smokers and identification of factors that are associated with SHS exposure would guide the development of SHS exposure reduction intervention strategies at home.

Table 1 shows the sociodemographic characteristics of the sample. Of the non-smoking respondents (N = 2813), three quarters (74%) were female and the highest percentage of respondents (38%) were 25 to 39 years of age. The vast majority of respondents were married (80%) and Muslim (84%). About one-half of respondents lived in rural areas (54%), had a moderate (1 to 8 years) level of formal education (54%) and were in the middle category for monthly household income (between 5,000 and 10,000 taka; 46%). In addition, three quarters (75%) of the respondents had a family member who smoked in the past or currently smokes and 22% of the respondents were current smokeless tobacco users.

Overall, 43% (95% CI: 36.9% – 48.5%) of Bangladeshi non-smokers are exposed to SHS. Table 2 shows that exposure to SHS varied by important demographic characteristics. Exposure to SHS was lowest among residents of Dhaka city (34%) and highest among residents of the Dhaka slums (64%). A greater percentage of women were exposed to SHS (48%) than men (34%). Younger adults were more likely to be exposed to SHS than older adults (p < 0.001). Illiterate Bangladeshis were also more likely to be exposed to SHS than the most educated (52% vs. 29%, respectively). Exposure also varied by household income (those in the highest income category were less likely to be exposed, p <0.001) and whether they had microfinance loan (those who had a loan were more likely to be exposed, p <0.02). Furthermore, non-smokers whose father smoked were more likely to be exposed to SHS (48%) than those whose father did not smoke (31%) (p <0.001). Also, SHS exposure was higher among those who had a smoking family member (47%) than those who did not have one (31%) (p <0.001). Exposure did not significantly differ by urban/rural area, marital status, religion, self-rated health, age of the youngest child in the home, or smokeless tobacco use.

As shown in Table 3, few of the attitudinal/knowledge measures were related to SHS exposure. Non-smokers who agreed (48%) or strongly agreed (42%) that smoked tobacco is addictive were significantly less likely to be exposed to SHS than those who did not agree (64%) (p < 0.05). Non-smokers who were very/extremely concerned about the harms of tobacco smoke to children were significantly less likely to be exposed to SHS (43%) than either those who were moderately concerned (61%) or unconcerned (74%) (p <0.001). Those who believed that SHS causes lung cancer in non-smokers were significantly less likely to be exposed to SHS (40%) than those who did not know this fact (62%) (p <0.001). Non-smokers who believed that SHS causes asthma in children were significantly less likely to be exposed to SHS (41%) than either those who did not believe (50%) or did not know (56%) this fact (p <0.001).

The stepwise logistic regression model (unweighted) identified eight predictors of SHS exposure: sex, age, recruitment area (Dhaka non-slum, Dhaka slum, non-tribal/non-border, tribal/border), education, income, whether father smoked, concern that tobacco smoke harms children and believing that SHS causes lung cancer in non-smokers. This model was re-estimated using a complex samples logistic regression model to estimate the effect of each predictor on SHS exposure (Table 4). Based on this model, women had 2 times greater odds of being exposed to SHS than men. The youngest Bangladeshis (aged 15–24) had 2 times greater odds of being exposed to SHS than the oldest Bangladeshis (aged 55 and older). Social gradients in exposure were also found, with illiterate and less educated Bangladeshis having higher odds of exposure compared to the more highly educated. However, Bangladeshis with a higher income (taka 5000 or more) had greater odds of being exposed to SHS compared to low income (<5000 taka) Bangladeshis. Residents other than the Dhaka non-slum areas had higher odds of SHS exposure. Bangladeshis who did not know that SHS causes lung cancer had 2 times greater odds of being exposed to SHS than those who knew this fact. Finally, concern about smoking in the presence of children was significantly associated with the odds of SHS exposure even after controlling for all other predictors: those Bangladeshis who were unconcerned/a little concerned had the 3.45 times the odds of exposure compared to those who were very/extremely concerned.

The findings of this study show that almost half (43%; 95% CI: 37%-49%) of non-smoking Bangladeshi adults are exposed to SHS at home. To our knowledge this is the first study to report SHS exposure in the home among non-smoking adults in Bangladesh. The SHS exposure rate in the current study is lower than the exposure rate in the workplace (63%) as reported in the GATS Bangladesh survey [10], but comparable to the exposure rate (45%) in public places [10]. The high SHS exposure rate in the workplace may be due to the fact that the smoking prevalence among Bangladeshi men is high and the majority of the workforces are men. Also, the partial workplace smoking restrictions implemented in 2005 are not well enforced [9].

Our findings showed that the SHS exposure prevalence at home was higher in females than males (48% versus 34%). This is consistent with findings in China [13], Korea [14] and Spain [15]. The high exposure prevalence among females is related to Bangladeshi social norms that disapprove of smoking by women; high smoking rates among males; and no or limited enforcement of household smoking bans [11]. In Bangladeshi culture, especially in rural areas, most women are housewives and spend most of their time at home, which increases their risk to SHS exposure from male smokers in the household. The dominant role of males in Bangladeshi society combined with respect to the household head (i.e. father) may also explain why father’s smoking status was a significant predictor for SHS exposure in the current study. To protect these vulnerable women from SHS exposure related morbidity and mortality, there is a need to empower women with information about SHS exposure related health effects and the measures they can take to avoid SHS exposure.

Consistent with findings in other countries [16‐18], younger age was associated with SHS exposure among non-smokers. This might be due to the way families live in Bangladesh where three generations often live together in the same home. Because smoking prevalence among the older generation (aged 45–54: 26%; aged 55 or above: 22%) was higher than among young adults (aged 15–24:3%) [9], the young adult is more likely to live with current smokers than the older generation, resulting in more SHS exposure. Another factor may be the cultural tradition in Bangladesh of respecting one’s elders, which discourages arguing against elderly members’ smoking at home.

Many nonsmoking Bangladeshis in the study did not know that SHS causes lung cancer in non-smokers and not knowing was significantly associated with SHS exposure. While this finding is consistent with studies elsewhere [16, 18], this underscores the need to increase non-smoking Bangladeshis’ knowledge of SHS exposure and its health risks. This low perception of risk might be related to the fact that many non-smoking Bangladeshis did not perceive themselves to be at risk from SHS exposure and did not take any protective measures to avoid SHS exposure. Health education campaigns [19, 20] are critical to improve people’s knowledge of and attitudes towards SHS exposure.

Consistent with previous studies [18, 20, 21], our study showed that Bangladeshis with less education were more likely than more highly educated Bangladeshis to report SHS exposure. An earlier study reported that Bangladeshis with low socioeconomic status were more likely to be exposed to SHS [11]. This educational disparity in SHS exposure among Bangladeshis underscores the need for targeted educational interventions to improve health-related knowledge among the less educated Bangladeshis and emphasize the promotion of smoke-free home policies to this disadvantaged population group. Educating smokers about smoke-free homes would also increase smokers’ likelihood of quitting smoking and decrease cigarette consumption [22].

Somewhat unexpectedly, Bangladeshis with a moderate income (5000–10,000 taka) had higher SHS exposure at home. We have further explored this and found that there is a confounding effect with residence. In urban areas, SHS exposure decreased within increasing income (p <0.001). SHS exposure did not vary by income levels in the slum areas and in rural areas; SHS exposure was highest in the middle income group (data not shown). This phenomenon can be explained by the existence of wide variation of income among urban residents, but not among rural residents. The income inequality is less pronounced in rural areas and almost nonexistent in slum areas of Bangladesh.

The study has several limitations. First, the cross-sectional design of the study limits causal inferences about the findings. Second, SHS exposure was defined as a proxy measure constructed from self-reported home smoking rules and household composition of smokers, and might thus not always reflect real exposure to SHS at home. In addition, this measure does not provide information about dimensions of exposure such as frequency and intensity. While we recognize the limitations of the exposure variable, given the nature of the ITC survey, and given that no objective measures of SHS were employed during data collection, this measure still provides an indication of the possible exposure to SHS among non-smokers in Bangladesh. , Third, the current paper focused on household SHS exposure and did not identify other sources of exposure such as workplaces or outdoors. Therefore, the findings cannot be generalizable to SHS exposure outside of home. Fourth, there is a great gender imbalance in the study (74% participants were female). The gender imbalance in this study is an artifact of the sampling procedure and the low prevalence of smoking among Bangladeshi women. With respect to factors associated with SHS exposure, we controlled for gender in our regression model, so even after controlling for the gender imbalance, we still see significant predictors of exposure. Moreover, when we stratified the bivariate associations by gender, trends and significance tests are generally similar. So we think the gender imbalance is less of a problem for factors associated with SHS exposure, especially since we control for gender in the logistic regression model.Finally, data were collected by trained interviewers who followed written interviewer guidelines. Any difference between their understanding and explanation of the questions asked could result in bias in information collected. However, such bias was minimized by the periodical observation of interviews by the senior research team members and bi-weekly meetings. Despite these limitations, this study supports the need to develop comprehensive interventions to promote SHS exposure reduction at home.

This study documents that almost half of non-smoking Bangladeshi adults are exposed to SHS in the home. Many of the predictors of SHS exposure identified in the current study are amenable to change. Policy strategies should consider these factors in the design of interventions addressing SHS exposure reduction among the public. Both population-based policies and clinical encounters are needed to improve public understanding about the health hazards of SHS exposure and to encourage the public in enacting smoking policies that promote a smoke-free home. It is also suggestive that smoke-free indoor policies should not only be enforced among non-smokers (mostly female) given the dominant role of males in Bangladeshi society. Therefore, attention should be also or especially given to (male) smokers in enforcing smoke-free indoor policies. There is evidence that smoke-free indoor policies are effective and easy to implement for the prevention of exposure to SHS, and a 100% ban on smoking in any indoor environments (i.e. at home) is the only intervention that can effectively protect people from the hazards of SHS exposure. Therefore, there is a need to implement evidence-based tobacco control strategies to promote SHS exposure reduction in the home. Interventions should deliver targeted messages to reach those in the low socioeconomic status group.