INTRODUCTION
Tobacco consumption is considered as one of the major behavioral risk factors and an underlying cause of adverse health, disability and preventable deaths1. More than 8 million people die each year due to tobacco-related disease worldwide2, and these numbers are likely to increase over the coming decades3. Most of the deaths projected to occur in low- and middle-income countries2,3. India was the second main tobacco consumer after China in 20164. However, the trend in tobacco consumption slightly declined5 due to effective population-level strategies and law, price increase and restrictions on the availability of tobacco6. Although a recent estimate has revealed that 266.8 million Indian adults consume any form of tobacco, and over one million deaths are attributed to tobacco-related diseases every year. The prevalence of tobacco use is highest in northeastern states of the country4. Tobacco use is significantly associated with a large and growing cause of premature mortality, morbidity and disability for all ages in the country3. The annual economic burden from tobacco constitutes approximately 1.04% of India’s GDP, and the direct health expenditure on treating tobacco-related diseases accounts for 5.3% of the total private and public health expenditures in India per year7. Therefore, reducing the burden of health and economic consequences of tobacco is obviously of paramount importance in the field of public health.
Evidence identified tobacco cessation as the most cost-effective intervention to reduce tobacco-related disease and premature mortality than other tobacco control programs1,8,9. Evidence also suggests that older age, higher level of education, health problems and counselling from health professionals contribute to tobacco cessation10-12. India has endorsed MPOWER (Monitor, Protect, Offer help, Warn, Enforce and Raise taxes) as a tool to help implement the World Health Organization’s Framework Convention on Tobacco Control in 200413. These tools include reducing affordability through taxation on tobacco products, passing smoke-free laws, mandating health warnings about the dangers of tobacco on the packaging, banning tobacco advertising, and offering help for the cessation of tobacco use, promotion, and sponsorship to control the demand and supply of tobacco14. Article 12 of the FCTC requires signatories to promote and strengthen public awareness of tobacco control issues using available communication tools, including mass media campaigns1. Research from developed countries has shown that mass media can be successful in disseminating harmful information regarding tobacco consumption and play an important role in discouraging all forms of tobacco8,15,16. Mass-media interventions consist of communication through television, radio, newspaper, billboards, and print media to encourage or motivate tobacco cessation. Mass media campaigns legitimize community action on tobacco control and trigger other interventions over a large population15.
Initial evidence from India has found that advertising related to tobacco and smoking is positively associated with a higher smoking rate due to brand name or celebrity smoking, especially among youth populations17,18. However, on the other hand, recent evidence shows that the anti-smoking message delivered through mass media have encouraged tobacco cessation in India12,19. In addition, Pierce and Gilpin20 found that information regarding smoking and health in news media are associated with smoking cessation but not initiation. Kar et al.21 reported that Indian adults exposed to anti-tobacco messages in newspapers/magazines and cinemas and those who belong to the northeastern region are more willing to quit than their counterparts.
Little is known about role of mass media on quitting tobacco use in North Eastern States of India. This study provides insights into the current use of mass media among tobacco products users to design and implement effective tobacco legislation or to design successful intervention programs to combat increasing tobacco use. We aimed to test the hypotheses that adults exposed to more anti-tobacco advertising through mass media will be more likely to stop smoking successfully. Moreover, the primary goal of the current study was to evaluate the pattern of quitting smoking and smokeless tobacco use among Indian adults and to assess the impact of mass media on tobacco cessation.
METHODS
Study design and participants
The present secondary analysis was based on the data from the second round of Global Adult Tobacco Survey India (GATS 2 India) 2016–2017. This survey was conducted by the Tata Institute of Social Sciences (TISS), under the stewardship of the Ministry of Health and Family Welfare (MoHFW), New Delhi. Its specific objective was to obtain reliable estimates for various dimensions of tobacco use in different regions of the country4. The survey covered 30 states of the country and two union territories. GATS-2 was conducted with 84047 households and 74037 individuals aged ≥15 years.
Outcome variables
Though GATS 2 collected data on tobacco use of any form, this study considered the individuals who were either users of smoked tobacco or smokeless tobacco.
In the survey, six specific questions pertaining to both smoked and smokeless tobacco use were used in each section. Initially, a screening question (Question B04) was asked to identify current daily smokers, asking: ‘How old were you when you first started smoking tobacco daily?’. For respondents who were uncertain about the age at which they began daily smoking, a follow-up question (Question B05) was asked: ‘How many years ago did you first start smoking tobacco daily?’. Similar sets of questions were posed to individuals who smoked tobacco less than daily and those who were former smokers (Questions B08, B09, B11, and B12).
Parallel questions were asked to smokeless tobacco users. Question C04 asked: ‘How old were you when you first started using smokeless tobacco daily?’. In cases where respondents did not recall the exact age, a subsequent question (Question C05) asked: ‘How many years ago did you first start using smokeless tobacco daily?’ Additional questions were similarly directed at those who used smokeless tobacco less than daily or were former users (Questions C08, C09, C11, and C12).
To convert these responses into time-to-event variables, the duration from initiation to cessation was calculated for former tobacco users, for whom the age of quitting tobacco use was available. For current users of either smoked or smokeless tobacco, where quitting age was not available, these individuals were treated as censored observations in the time-to-event analysis.
Independent variables
The independent variables considered under the current study were sex (male, female), marital status (unmarried, married, other), residence (rural, urban), education level (no education, primary education, secondary education, higher education), occupation (working/non-working), and media (exposed/unexposed). Household wealth index was estimated as assets of the household, and accordingly, principal component analysis (PCA) was employed for the construction of wealth index where each household was assigned a total score based on the availability of specific assets. Households were then ranked according to these scores, and the wealth index was divided into quintiles. The first quintile represents the individuals belonging to 20% of households with lowest wealth score up to the top representing 20% of the highest wealth score22.
Statistical analysis
A univariate analysis was performed to assess the distribution of the study population. A log-rank test was conducted and the average quitting time of both smoked tobacco users and smokeless tobacco users, by background characteristics, was captured separately. Age at initiation of tobacco and age at quitting were used to compute the time variable. The censored cases in the study were those who were continuing tobacco use till the date of survey, for this study. In order to estimate the hazard ratio for quitting tobacco, the Cox23 proportional hazard model was used based on the formula:
where λ(t|z) is the hazard rate at time t, zp is a predictor variable, λ0(t) is the baseline hazard (when all predictors are zero), and β1,...,βp are coefficients representing the contribution of each predictor variable. This model was used to explain the effect of independent variables with time until the occurrence of the event, which in our case is the quitting of smoked and smokeless tobacco.
Kaplan-Meier survival analysis was conducted to investigate the relationship of the survival distribution to the covariates under study24. The graph of Kaplan-Meier analysis was obtained by sex, residence, occupation, education level, wealth index and mass media exposure for smoked tobacco users and smoked tobacco users separately. However, in the current study an attempt has been made to capture the progression of quitting tobacco in all the northeastern states of the country using Kaplan-Meier survival analysis and Cox’s proportional hazard ratio25. Results are presented with 95% confidence intervals, after adjusting for GATS sampling weight. All analyses was done with SPSS 12.
RESULTS
Table 1 presents the average quitting point in time of smoked tobacco users along with the log-rank test, by background characteristics. The average quitting point in time of smoked tobacco was 66.7 years for females and 54.2 years for males. The average quitting time was higher among the respondents belonging to rural areas. Considering education level, the average quitting time of smoked tobacco was inversely proportional to education level. That is, the average quitting time of smoked tobacco among those with no education is 63.9 years compared to those with higher education level, which is 43.9 years. The average quitting time of smoked tobacco among the poorest is 46.7 years whereas among the richest it is 66.8 years. Working respondents had an average quitting time of 60.6 years, whereas for non-workers it is 54.0 years. The average quitting time of smoked tobacco among the married and the single was more or less the same. The average quitting time of smoked tobacco is more among the respondents who were not exposed to mass media. Apart from this, results from log-rank test for all the background characteristics suggested that there is significant difference in mean quitting time of smoked tobacco by education level and wealth index.
Table 1
Log-rank test and average quitting time of the smoked tobacco users, by background characteristics, GATS-2, India
Table 2 presents the average quitting time of smokeless tobacco users along with the log-rank test, by background characteristics under study. Findings suggested that the average quitting time of smokeless tobacco was 76 years for females and 63 years for males. The average quitting time was higher among the urban dwellers. In addition, the average quitting time for smokeless tobacco was 70.4 years for respondents with no education and 51.9 years for the those with higher education. Considering the wealth index, it was found that the average quitting time for smokeless tobacco is highest among respondents from the poorer section, followed by those from the richer section. The average quitting time of smokeless tobacco was higher among the working respondents. The maximum average quitting time of smokeless tobacco was among the married respondents. Apart from this, the average quitting time of smokeless tobacco was higher among the respondents who were exposed to mass media. However, results from the log-rank test for all the background characteristics suggested that there is a significant difference in the mean quitting time of smoked tobacco by sex, education level, wealth index, marital status, and exposure to mass media.
Table 2
Log-rank test and average quitting time of the smokeless tobacco users, by background characteristics GATS-2, India
Table 3 presents the Cox proportional hazard ratio of quitting smoked and smokeless tobacco by background characteristics under study. Findings from the current study documented the education level, occupation, marital status and mass media exposure of a person as significant determinants of quitting smoked tobacco among the users. Considering education level, it was found that compared to the uneducated, those with primary, secondary, and higher education, were 1.54 times (p<0.05), 1.79 times (p<0.01), and 2.84 times (p<0.01) more likely to quit smoking tobacco, respectively. However, education level showed a significant positive association with quitting smoked tobacco. Working (OR=0.68; p<0.01) and married (OR=0.67; p<0.10) respondents had lower odds of quitting smoked tobacco compared to the non-working and those who were single.
Table 3
Cox proportional hazard ratio of quitting tobacco among smoked and smokeless tobacco users by background characteristics under study GATS-2, India
| Characteristics | Quitting smoked tobacco | Quitting smokeless tobacco | ||
|---|---|---|---|---|
| HR | 95% CI | HR | 95% CI | |
| Sex | ||||
| Male | 1 | |||
| Female | 0.77 | 0.52–1.16 | 0.60*** | 0.42–0.89 |
| Residence | ||||
| Urban | 1 | |||
| Rural | 1.04 | 0.77–1.41 | 1.09 | 0.75–1.61 |
| Education level | ||||
| No education | 1 | |||
| Primary | 1.54** | 1.08–2.21 | 1.06 | 0.69–1.67 |
| Secondary | 1.79*** | 1.21–2.69 | 1.89*** | 1.20–3.00 |
| Higher | 2.84*** | 1.53–5.30 | 1.88* | 0.9–3.97 |
| Wealth index | ||||
| Poorest | 1 | |||
| Poorer | 1.09 | 0.69–1.75 | 0.74 | 0.44–1.29 |
| Middle | 1.11 | 0.68–1.82 | 0.84 | 0.47–1.50 |
| Richer | 1.00 | 0.63–1.62 | 0.70 | 0.40–1.25 |
| Richest | 0.91 | 0.54–1.52 | 0.74 | 0.41–1.35 |
| Occupation | ||||
| Working | 0.67*** | 0.51–0.90 | 0.82** | 0.57–1.192 |
| Non-working | 1 | |||
| Marital status | ||||
| Single | 1 | |||
| Married | 0.66* | 0.42–0.95 | 0.86** | 0.47–0.96 |
| Other | 0.84 | 0.48–1.50 | 1.09 | 0.51–2.32 |
| Media | ||||
| Exposed | 1.01** | 1.01–1.44 | 1.21** | 1.07–1.70 |
| Non-exposed | 1 | |||
Considering smokeless tobacco users, this study identified that sex, education level, occupation, marital status and mass-media exposure of a person were significant determinants of quitting smokeless tobacco. Female respondents were 0.61 (p<0.01) times less likely to quit smokeless tobacco compared to males. Respondents with secondary and higher education were 1.8 times more likely to quit smokeless tobacco compared to thethose who are uneducated. Respondents who were either working (OR=0.83; p<0.05) or married (OR=0.86; p<0.05) were less likely to quit smokeless tobacco compared to the non-working and the single.
However, when considering exposure to mass media, findings suggested that respondents exposed to mass media were 1.01 and 1.22 times more likely to quit smoked and smokeless tobacco compared to the non-exposed, respectively.
Supplementary file Figure 1 depicts the Kaplan-Meier survival curve showing the quitting time of smoked tobacco by different background characteristics. The survival curve for education level shows that the quitting time of smoked tobacco among highly educated users, was lower than that of the uneducated. Considering mass-media exposure, the quitting of smoked tobacco was earlier among those exposed to mass media.
Supplementary file Figure 2 depicts the Kaplan-Meier survival curve showing the quitting time of smokeless tobacco users by different background characteristics. The survival curve for education level showed that the quitting time of smokeless tobacco was lower among the highly educated users, compared to the uneducated. In addition, the quitting time of smokeless tobacco was earlier among those exposed to mass media.
DISCUSSION
The growth of the population, coupled with strategic initiatives by the tobacco industry, has driven the rise of tobacco consumption, particularly in low-income nations. These factors have led to millions of people becoming fatally addicted to tobacco every year. However, recent studies reveal that the prevalence of smoking is highest in the northeastern states of India, while the use of smokeless tobacco products is most common in the Empowered Action Group (EAG) states of the country14. The World Health Organization, in its convention on Tobacco control, constructs a framework to increase awareness among citizens about health hazards caused by tobacco consumption, exposure to tobacco smoke and the health benefits of living a tobacco-free life14.
The findings presented in the study provide valuable insights into the factors affecting tobacco cessation among different demographic groups. Our findings revealed that females tend to quit smoked tobacco at an average age of 66.7 years, whereas males quit at 54.2 years. This indicates a significant difference in quitting patterns between genders. In prospective observational studies, it has been observed that women are less inclined to quit smoking compared to men. Consequently, women are at a greater risk of experiencing the long-term consequences of smoking. Alarmingly, healthcare practitioners might not be fully aware of women’s heightened vulnerability to smoking-related diseases. Addressing this issue is crucial because smoking cessation interventions should be customized for women, considering the distinct reasons behind their smoking habits and the unique obstacles they face in successfully quitting26,27. Similar gender disparities exist in case of smokeless tobacco as well, with females quitting at an average of 76 years and males at 63 years. This suggests that females, in general, tend to quit tobacco habits later than males, regardless of the type of tobacco used. Information gathered from treatment studies often indicates that women are less successful in quitting smoking compared to men. However, these findings have been challenged, primarily due to conflicting data from epidemiological studies. The aim of this review was to resolve this discrepancy. By examining evidence from both efficacy and effectiveness trials, along with prospective observational studies on relapse, it became evident that women encounter greater challenges in sustaining long-term abstinence from smoking compared to men28.
The current study found that an inverse relationship between education and quitting time. Individuals with no education quit at 63.9 years, while those with higher education quit at 43.9 years, hence higher education significantly shortens the quitting time. Educated individuals quit smokeless tobacco earlier than those with no education. This highlights the importance of education in tobacco cessation campaigns. An analysis of national health survey data in Finland, using an adjusted regression model, revealed that individuals with a higher level of education were more likely to quit smoking compared to those with basic education29.
Quitting time is significantly different based on wealth index, with the poorest quitting at 46.7 years and the richest at 66.8 years. Financial disparities influence tobacco cessation rates significantly. Surprisingly, the quitting time is highest among the poorer section, indicating other factors might be at play, potentially access to healthcare and awareness; these findings align with previous studies that have studied smoking and smokeless tobacco cessation rates and revealed significant disparities. Individuals with low wealth index exhibited higher chances of quitting compared to those with a high asset index. Despite the well-documented health risks associated with smoking, the economic implications have been understudied.
Although a direct causal link cannot be established, it appears that smokers fund their tobacco expenses from income that non-smokers can save. Consequently, reducing smoking rates could assist in household incomes, particularly among the economically disadvantaged, underscoring the potential economic benefits of anti-smoking initiatives30,31. Working individuals have a slightly higher quitting time compared to non-workers (60.6 years vs 54.0 years). Working individuals tend to quit smokeless tobacco later. Previous research has underscored the significant health and economic implications of smoking for both individuals and society as a whole. For male office workers who currently smoke, it was found that working more than 52 hours per week is linked to a reduced intention to quit smoking. To gain a more comprehensive understanding, future studies should explore additional work-related factors, including job-related stress and physical workload32,33. Married and single individuals have similar quitting times, indicating that marital status might not be a significant factor in smoked tobacco cessation.
Lastly, the current study disclosed that mass media plays a vital role in raising public awareness, especially in diverse social classes and castes, regarding public health issues. Our study focused on the impact of mass-media exposure, particularly anti-tobacco advertisements in India, on adults. It revealed a positive influence, leading to early tobacco cessation. The constant media barrage, including television screens, billboards, and mobile internet access, significantly shapes beliefs and behaviors, notably concerning smoking. Online platforms offer cost-effective experimentation for communication techniques. Research indicates that exposure to outdoor tobacco ads and magazine information prompts smoking initiation. Behavioral change aligns closely with recent media exposure, emphasizing the need for consistent and frequent tobacco control media campaigns to effectively reduce smoking prevalence among adults8,34,35.
Limitations
Global Adult Tobacco Survey India is a cross-sectional survey rather than a longitudinal design, so the survival analysis conducted in the present study only takes into account individuals present at the time of the survey that excludes those who may have quit tobacco long ago and are no longer affected by tobacco-related issues. Moreover, this study does not capture follow-up information on individuals who reported quitting and have remained tobacco-free over time, and regarding their quit attempts across time. In addition, all data on tobacco use and cessation are self-reported, potentially introducing recall bias and reputational bias, as the timing of tobacco initiation and cessation is based on retrospective self-reported data in a cross-sectional survey, which may affect the accuracy of the time-to-event analysis.
CONCLUSIONS
This study underscores the intricate relationship between sociodemographic factors and tobacco cessation. Addressing these complexities is essential for desinging effective, targeted interventions that can reduce tobacco use across varied communities. The study also provides important insights into the role of mass media in influencing tobacco consumption in India, while examining the socio-economic factors linked to this behavior. It is vital for the government to plan and enforce comprehensive policies to regulate tobacco consumption and its widespread prevalence in the country. Education level, occupation, marital status, and mass-media exposure significantly influence quitting both smoked and smokeless tobacco. Higher education, non-working status, and exposure to mass media increase the likelihood of quitting.
