Prevalence, Sources, and Correlates of Second-hand Smoke Exposure Among Non-smoking Pregnant Women in India

Data Source and Study Population

We utilised data from the second round of GATS (2016–17) in India, which surveyed a nationally representative sample of 74 037 adults aged 15 and older. This sample was selected using a multistage, stratified cluster sampling method and included 1403 pregnant women, 9 of whom were smokers at the time of the survey. For the purposes of this study, only non-smoking pregnant women were considered, resulting in an effective sample size of 1394 for assessing SHS exposure at home. Pregnant women who are employed outside the home or who frequent public places are also at risk of SHS exposure. Among our sample, 982 pregnant women either worked outside their homes or visited public places in the previous month.

Measures

Statistical Analysis

Descriptive analysis was employed to determine the prevalence of SHS exposure among pregnant women both at home and in external environments. To assess SHS exposure outside the home, we calculated exposures in the workplace and in each public place visited. The numerator consisted of pregnant women exposed to SHS, while the denominator encompassed all pregnant women who had visited those settings in the previous 30 days. The proportion of SHS exposure at each location was then multiplied by the proportion of pregnant women who visited that location, to ascertain the proportion of pregnant women exposed to SHS outside their homes.

Logistic regression models were developed to identify the correlates of SHS exposure both at home and in other settings. The initial model involved a univariate analysis using simple logistic regression to explore the relationship between the independent variables and SHS exposure, aiming to calculate the unadjusted odds ratios (ORs). Subsequently, a multivariable logistic regression model was employed to adjust for the effects of variables other than the primary variable of interest, with the results presented as adjusted odds ratios (aORs). In both descriptive and inferential analyses, survey weights were applied; however, the sample count (n) reported is unweighted. The analyses were conducted using Stata version 14.0 (StataCorp., College Station, TX, USA).

Ethics Statement

The study received approval from the Ethics Committees of Tata Institute of Social Sciences (TISS-IRB/2015-16/05/GATS-INDIA-II).

Descriptive Analysis

Table 1 presents the descriptive statistics of the study sample. Table 2 shows the weighted prevalence of SHS exposure both at home and in external environments, categorised by demographic characteristics. A weighted estimate indicates that the overall prevalence of SHS exposure at home was approximately 36.9%. The highest prevalence of exposure at home occurred in the 15–24 age group, at 43.0%. A gradient was observed in educational levels, with exposure rates decreasing from 51.4% among those without formal education to 22.3% among those with higher education. Unemployed individuals showed a lower prevalence of 36.3%, compared to 39.5% among those who are employed. Among the caste categories, women from scheduled tribes reported the highest prevalence (54.9%). Regionally, the Northeast exhibited the highest prevalence at 48.7%, followed by the Central and East regions (44.5 and 40.1%, respectively). Rural areas had a higher prevalence of 41.1%, compared to 27.2% in urban settings. In terms of economic status, the lower 3 quintiles experienced a prevalence of 42.0% to 42.8%, while the top 2 quintiles had a prevalence of 25.8% and 27.1%, respectively.

Table 1

Descriptive statistics for sample of non-smoking pregnant respondents in the GATS (2016–17), India (n=1394)

Table 2

Prevalence (%) of SHS exposure among non-smoking pregnant respondents in the GATS (2016–17), India

The prevalence of SHS exposure outside the home was 26.5%. Among different age groups, the highest prevalence was observed in individuals aged 35 and above (29.7%). The groups with the highest prevalence included illiterate and employed individuals (30.1 and 35.5%, respectively), scheduled tribes (40.1%), and residents of the North region (40.4%). Urban areas showed a higher prevalence of 31.9%, compared to 23.6% in rural areas. The highest exposure rate (40.3%), was found in the topmost economic quintile.

Table 3 shows the contribution of various sources to exposure. Column (A) represents the percentage of women who were affected by a particular source of exposure in the last 30 days. Column (B) shows the percentage of SHS exposure among pregnant non-smoking women who visited their workplace or a specific public place. The highest proportion of women exposed to SHS in a particular public place occurred in private offices (39.1%), followed by public transport (27.2%), and government buildings (25.9%). The lowest proportion was observed in bars/nightclubs (8.4%). Column (C), calculated as a product of columns (A) and (B), shows the proportions of pregnant non-smoking women exposed to SHS in public places. Only 8.8% of women visited private offices in the previous 30 days, with 39.1% of them reporting SHS exposure. Therefore, only 3.4% of all pregnant women experienced SHS exposure at private offices. In contrast, 41.9% of pregnant women used public transport during the same period, with 27.2% encountering SHS exposure. Therefore, 10.0% of all pregnant women were exposed to SHS in public transport.

Table 3

Contribution of various sources to SHS exposure among pregnant non-smoking women

Logistic Regression

At-home SHS exposure

The ORs and aORs for SHS exposure at home, along with their 95% confidence intervals (CIs), are presented in Table 4 using both simple and multivariable logistic regression analyses. We observed a significant association between geographic region and home exposure. In the unadjusted model, the odds of exposure were significantly higher in the Northeast, Central, North, and East regions compared to the South. The aORs also indicated higher risks in the North (aOR, 5.33; 95% CI, 2.45 to 11.60), Central (aOR, 4.46; 95% CI, 1.98 to 10.02), Northeast (aOR, 4.18; 95% CI, 1.78 to 9.81), and East (aOR, 3.59; 95% CI, 1.58 to 8.15). Conversely, the unadjusted model revealed a significantly lower risk of exposure in urban areas (aOR, 0.53; 95% CI, 0.32 to 0.88), a finding not replicated in the adjusted model. Pregnant women aged 25–34 (aOR, 0.61; 95% CI, 0.39 to 0.93) and those aged 35 and older (aOR, 0.48; 95% CI, 0.27 to 0.86) exhibited lower risks of SHS exposure at home compared to those aged 15–24. Both models demonstrated significant associations between education level and SHS exposure, with decreased likelihoods of exposure among respondents with secondary (OR, 0.50; 95% CI, 0.30 to 0.82) or higher education (OR, 0.27; 95% CI, 0.16 to 0.47). Pregnant women from scheduled tribes faced a higher risk of exposure (OR, 2.49; 95% CI, 1.25 to 4.99) in the unadjusted model, though this effect was not evident in the adjusted model. Additionally, home exposure was not associated with economic quintiles in the adjusted analysis, nor was it linked to employment or religion in either the unadjusted or adjusted models.

Table 4

Logistic regression results for SHS exposure at home and outside the home

Second-hand Smoke at Home

The overall prevalence of SHS exposure among Indian pregnant women was approximately 36.9% at home during 2016–2017. Among non-smoking women, SHS exposure rates varied from 15% in Mexico to 69% in Vietnam during 2008–2010 across 14 low-and middle-income countries (LMICs) [15]. A recent study examining the prevalence of SHS exposure at home in 57 LMICs from 2010 to 2018 utilised the Demographic and Health Survey (DHS) databases [16]. Only 7 of these countries reported lower levels of SHS exposure at home among pregnant women compared to India. High exposure levels in a densely populated country like India imply that a significant number of pregnant women are affected by SHS, which is a major cause for concern. Overall, the exposure rate was 24.4% among pregnant women and 22.8% among other women. In India, the exposure rate for all adult non-smoking women was 38%, which is similar to that of pregnant women [9].

Similar to findings from a previous study [17], SHS exposure at home was higher among young pregnant women than among older women in this study. Child marriage significantly increases women’s risk of SHS exposure at home compared to those married as adults [18]. As women age, they may become more assertive and gain greater household power [19,20], which can enhance their ability to control their exposure and thus may serve as a protective factor against SHS exposure at home. It is also possible that in socioeconomic groups where smoking is more prevalent, the age at marriage tends to be lower. Studies conducted by Aurrekoetxea et al. [5] in Spain, Alghamdi et al. [21] and Wahabi et al. [22] in Saudi Arabia, Kelly et al. [23] in India, and Nazar et al. [24] across 15 LMICs support our finding that a lower education level is associated with increased SHS exposure.

Awareness of the consequences of SHS exposure is significantly associated with lower levels of education, which may in turn increase the risk of SHS exposure [25]. A strong correlation exists between the educational status of women and that of their husbands. The prevalence of smoking among men decreases with higher levels of education, potentially protecting their wives. Additionally, educated women may enhance their ability to negotiate with smoking members of the household.

Regional variations in SHS exposure can be attributed to differences in smoking prevalence across various regions in India [26]. Additionally, disparities in fertility patterns and gender inequalities between regions may play a role. The northern region exhibited higher SHS exposures both at home and in outdoor settings.

Second-hand Smoke Outside the Home

The overall prevalence of SHS exposure outside the home was approximately 26.5% among Indian pregnant women in 2016–2017. The GATS report indicated that SHS exposure at the workplace decreased from 18.9% to 17.7% (a relative change of 6.3%) among non-smoking women from 2009–2010 to 2016–2017, despite a 23.6% decrease in smoking prevalence (from 14.0% in 2009–2010 to 10.7% in 2016–2017) [9]. The current study reveals that 24.8% of employed pregnant women experienced SHS exposure at workplace. These findings suggest a weak implementation of COPTA in India. Latin American countries reported lower workplace SHS exposure compared to other regions, while Asian countries exhibited higher rates of workplace SHS exposure [15].

The scheduled tribe exhibited a significantly higher risk of exposure to SHS outside the home in this analysis. A study conducted in the United States found that Hispanic women were less likely than white and African-American women to work in environments with an official workplace smoking policy [27]. This disparity can be attributed to the ‘ethno-racial stratification of occupations,’ where Hispanic women are less frequently employed in professional roles that typically enforce a smoking policy [28]. Similarly, the association between scheduled tribes and increased SHS exposure outside the home in India can be partially explained by their higher likelihood of employment in the informal sector, which often lacks strict adherence to smoke-free workplace policies [29]. Additionally, the variation in SHS exposure among disadvantaged castes may stem from several factors, including their limited knowledge and awareness of the dangers of SHS, broader structural issues such as their reduced ability to negotiate for tobacco smoke-free environments due to their social disadvantage, which often places them at greater risk of living in adverse conditions [30], and a general lack of consideration for these disparities in the formulation of tobacco control policies [31].

The findings of this study highlight the need for a risk awareness campaign targeted at active smokers living in households with women of reproductive age, emphasising the dangers of SHS exposure during pregnancy. Additionally, there is a need to mobilise community support, particularly among younger and less educated women, to enhance their understanding of the adverse effects of SHS exposure. A more stringent smoke-free policy in public transport is also essential, as this setting is frequented by a significant number of pregnant women, making it the location where the highest percentage of non-smoking pregnant women are exposed to SHS among all non-home environments. Furthermore, the findings reveal a heightened risk of SHS exposure in the North region outdoors, underscoring the urgency to reinforce policies that prohibit smoking in public places, particularly in this region.

National and regional smoking levels significantly influence SHS exposure both at home and in public spaces. Interventions aimed at reducing the prevalence of smoking within a society are crucial for minimising SHS exposure. Protecting the health of non-smokers can be achieved through stringent laws and robust governance in workplaces and public areas. However, banning smoking in public places may not significantly change the overall smoking habits of individuals. Such bans might inadvertently push smokers from regulated public areas to unregulated private settings, such as homes. This shift can expose non-smoking individuals to SHS, particularly in situations where women may not have control over their smoking spouses. Therefore, while legal enforcement is necessary, it must consider existing gender inequities that could lead to unintended negative consequences. Although this specific issue was not confirmed by data from the current study, and the implementation of COTPA remains weak in India, evidence from other countries suggests that interventions should address these potential negative spillover effects.

The exposure to SHS at home for non-smoking women is predominantly caused by men family members who smoke [32]. Therefore, reducing smoking prevalence among men is crucial for decreasing SHS exposure in domestic settings. Studies have shown that behavioural interventions aimed at men smokers to promote smoke-free homes are effective [33,34]. In India, educational campaigns have been conducted to highlight the harmful effects of SHS. However, the GATS does not capture data on the proportion of men smokers who refrain from smoking inside their homes, nor the impact of national educational campaigns on their smoking behaviours. Nonetheless, these campaigns have influenced some men to quit or at least avoid or reduce smoking at home. Additionally, there is a trend toward nuclear family structures in India [35], which may decrease SHS exposure by reducing the number of men smokers in a household. The influence of family structure on SHS exposure represents another area for research. While these research topics originate from Indian data, they hold significance globally. It is essential to monitor SHS exposure among pregnant women worldwide, which can be achieved through the DHS and GATS.

The study has certain limitations. The dataset does not include information on whether men living in the household smoke. Additionally, it lacks details regarding the type of workplace (government or private) and occupation. The number of pregnant women within each primary sampling unit was small, which precluded the possibility of conducting multi-level analysis. Such analysis could have helped identify community and regional factors influencing exposure. Despite these limitations, the paper successfully highlights the high levels of SHS exposure among pregnant women and identifies groups where the exposure is significantly higher.