Tobacco Control Issue: Original Article
Prevalence and sociodemographic determinants of tobacco use in four countries of the World Health Organization: South‑East Asia region: Findings from the Global Adult Tobacco Survey Palipudi K, Rizwan SA1, Sinha DN2, Andes LJ, Amarchand R1, Krishnan A3, Asma S Centers for Disease Control and Prevention, Atlanta, USA, 1The INCLEN Trust International, New Delhi, 2South East Asia Regional Office, WHO, New Delhi, 3Centre for Community Medicine, All India Institute of Medical Sciences, India Correspondence to: Dr. Krishna M Palipudi, E‑mail: [email protected]
Abstract
INTRODUCTION: Tobacco use is a leading cause of deaths and Disability Adjusted Life Years lost worldwide, particularly in South‑East Asia.
Health risks associated with exclusive use of one form of tobacco alone has a different health risk profile when compared to dual use. In order to tease out specific profiles of mutually exclusive categories of tobacco use, we carried out this analysis. METHODS: The Global Adult Tobacco Survey (GATS) data was used to describe the profiles of three mutually exclusive tobacco use categories (“Current smoking only,” “Current smokeless tobacco [SLT] use only,” and “Dual use”) in four World Health Organization South‑East Asia Region countries, namely Bangladesh, India, Indonesia and Thailand. GATS was a nationally representative household‑based survey that used a stratified multistage cluster sampling design proportional to population size. Prevalence of different forms of usage were described as proportions. Logistics regression analyses was performed to calculate odds ratios (OR) with 95% confidence intervals. All analyses were weighted, accounted for the complex sampling design and conducted using SPSS version 18. RESULTS: The prevalence of different forms of tobacco use varied across countries. Current tobacco use ranged from 27.2% in Thailand to 43.3% in Bangladesh. Exclusively smoking was more common in Indonesia (34.0%) and Thailand (23.4%) and less common in Bangladesh (16.1%) and India (8.7%). Exclusively using SLT was more common in Bangladesh (20.3%) and India (20.6%) and less common on Indonesia (0.9%) and Thailand (3.5%). Dual use of smoking and SLT was found in Bangladesh (6.8%) and India (5.3%), but was negligible in Indonesia (0.8) and Thailand (0.4%). Gender, age, education and wealth had significant effects on the OR for most forms of tobacco use across all four countries with the exceptions of SLT use in Indonesia and dual use in both Indonesia and Thailand. In general, the different forms of tobacco use increased among males and with increasing age; and decreased with higher education and wealth. The results for urban versus rural residence were mixed and frequently not significant once controlling for the other demographic factors. CONCLUSION: This study addressed the socioeconomic disparities, which underlie health inequities due to tobacco use. Tobacco control activities in these countries should take in account local cultural, social and demographic factors for successful implementation. Key Words: Global Adult Tobacco Survey dual use, prevalence, South‑East Asia Region, socioeconomic determinants, tobacco use
Introduction Tobacco use is responsible for six million deaths and 6.3% of Disability Adjusted Life Years (DALY) lost worldwide. [1] Despite decreases in tobacco smoking in the high‑income regions, the global estimates remain stable due to compensating increase in the South‑East Asia Region (SEAR).[2] The burden due to all forms of tobacco use (which also includes smokeless tobacco [SLT] use) would be much higher than these estimates. Majority of this burden due to tobacco use is borne by Low‑ and Middle‑Income Countries, because of their large population size and widely prevalent tobacco use.[1,3] Dual tobacco use, popularly defined as the concurrent use of tobacco smoking and at least one other form of SLT,[4] has been perceived as a threat to global tobacco control efforts in recent times. This particular type of use has emerged at least partly as the tobacco industry’s response to overwhelming scientific evidence of harm caused to human life by smoking.[5] SLT products, which are cheaper and lack the disadvantage of violating smoking laws such as “use in public places” and which purportedly help in quitting smoking. They are likely to be co‑adopted by those who are already smoking, in Access this article online Quick Response Code:
Website: www.indianjcancer.com DOI: 10.4103/0019-509X.147446 PMID: *******
S24
order to quit smoking or as a means of continued nicotine supply in public places. The question of dual tobacco use causing greater harm when compared to single product use is still open to debate.[6‑9] Although many studies [10‑15] have described the sociodemographic profile of tobacco use in SEAR, very few have tried to examine exclusive categories of tobacco use such as smoking only, SLT use only and dual use. The need for such analyses arises from the fact that use of one type alone has a different health risk profile as compared to dual use. To tease out the specific profile of these exclusive categories we needed to analyze them separately. In order to effectively plan for future tobacco control strategies we need to understand dual use and its determinants as opposed to studies, which focus on all tobacco use. The Global Adult Tobacco Survey (GATS) provided an excellent opportunity to examine the prevalence and determinants of exclusive categories of tobacco use. Methods We used the GATS data to describe the profile of tobacco users in four World Health Organization (WHO) SEAR countries, namely Bangladesh, India, Indonesia and Thailand. Detailed methodology of GATS in each country has been described elsewhere, [16‑19] but briefly GATS was a nationally representative household‑based survey, designed to obtain data on tobacco use behaviors of civilian noninstitutionalised individuals aged ≥15 years. Each country used a stratified multistage cluster sampling design to produce nationally representative samples. The important Indian Journal of Cancer | Dec 2014 | Volume 51 | Supplement 1
Palipudi, et al.: Tobacco use in the SEAR
survey details are given in Table 1. A minimum sample size of 8000 households was taken in each country, except India where 8000 households were chosen from each of six regions. Households were selected at random and in each selected household face‑to‑face interviews were done in the participant’s local language. Efforts were taken to maintain confidentiality of information obtained. Ethical clearance was obtained from country specific health ministries and implementing institutions.
followed by Thailand (23.4%), Bangladesh (16.1%) and India (8.7%). Prevalence of current SLT use was similar in India (20.6%) and Bangladesh (20.3%) followed by Thailand (3.5%) and Indonesia (0.9%). Prevalence of dual use was highest in Bangladesh (6.8%), followed by India (5.3%), Indonesia (0.8%) and Thailand (0.4%).
We defined three mutually exclusive groups of tobacco use: “Current smoking only,” “Current SLT use only,” and “Dual use” [Box 1].
Current tobacco use in males was highest in Indonesia (67.1%) and Bangladesh (58.0%) while the highest prevalence for females were in Bangladesh (28.7%) and India (20.3%). In all countries, a larger proportion of males used tobacco than females. Current tobacco use was also higher in rural areas (ranging from 29.2% in Thailand to 45.1% in Bangladesh) when compared to urban areas (ranging from 22.9% in Thailand to 38.1% in Bangladesh) in all four countries. There was a trend toward increasing prevalence with increasing age in all countries except in Indonesia, where the prevalence was slightly lower in the ≥ 65 years age group. Current tobacco use was lower with increasing education (except in Bangladesh) and with increasing wealth status (except in Indonesia) [Table 2].
The prevalence of different forms of usage were described as proportions with 95% confidence intervals [CI] and also within categories of sociodemographic variables such as gender (male, female), age group (15–24, 25–44, 45–64, 65 or older), residence (rural, urban), education (no formal education, primary, secondary, college/university or above) and wealth index quintiles (lowest, low, middle, high, highest). Association between different forms of usage and sociodemographic variables were calculated by performing logistic regression and strength of association described in terms of odds ratios (OR) with 95% confidence interval (CI). Where a particular cell contained zero, adjacent categories of explanatory variables were combined to allow the model to converge. All analyses were weighted, accounted for the complex sampling design and conducted using Statistical Package for the Social Sciences (SPSS) version 18. Results Prevalence of different forms of tobacco use [Figure 1]
The overall prevalence of current tobacco use was highest in Bangladesh (43.3%), followed by Indonesia (35.7%), India (34.6%) and lowest in Thailand (27.2%). Prevalence of current smoking only was highest in Indonesia (34%) Box 1: Definition of tobacco related terms used in GATS Definition Current tobacco use: Currently smoking tobacco in any form or using SLT or both Current smoking only: Currently smoking any tobacco product on a daily or less than daily basis exclusively Current SLT only: Currently using any SLT product on a daily or less than daily basis exclusively Dual use: Currently smoking any form of tobacco together with any form of SLT on a daily or less than daily basis
Prevalence of different forms of tobacco use by sociodemographic characteristics Current tobacco use
Current smoking only
Prevalence of current smoking only was very high in males (ranging from 15.0% in India to 65.6% in Indonesia) as compared to females (ranging from 0.8% in Bangladesh to 2.8% in Thailand) in all four countries. With increasing age, the prevalence also increased reaching a peak in 45–64 years age group and declining thereafter, except in Thailand where the peak was earlier in the age group
1RWREDFFRXVH 'XDOXVH &XUUHQW6/7RQO\ &XUUHQWVPRNLQJRQO\
%DQJODGHVK
,QGLD
,QGRQHVLD
7KDLODQG
Figure 1: Prevalence of different forms of tobacco use
SLT=Smokeless tobacco; GATS=Global adult tobacco survey
Table 1: Methodological details of the GATS survey by country Country
Survey dates
Un‑weighted numbers Sampled men
Sampled women
Overall response rate (%)
Coverage of population (%)
Bangladesh India Indonesia
July to August 2009 June 2009 to January 2010 October to November 2011
4468 33,767 3948
5161 35,529 4357
93.6 91.8 94.3
95.5 99.9 99.9
Thailand
February to May 2009
10,052
10,514
94.2
98.3
GATS=Global adult tobacco survey
Indian Journal of Cancer | Dec 2014 | Volume 51 | Supplement 1
S25
S26
OR
India (95% CI) Prevalence
OR
OR
Indonesia (95% CI) Prevalence
28.1 (25.1, 31.2)
1.0
17.2 (15.8, 18.6)
*P
Prevalence and sociodemographic determinants of tobacco use in four countries of the World Health Organization: South‑East Asia region: Findings from the Global Adult Tobacco Survey Palipudi K, Rizwan SA1, Sinha DN2, Andes LJ, Amarchand R1, Krishnan A3, Asma S Centers for Disease Control and Prevention, Atlanta, USA, 1The INCLEN Trust International, New Delhi, 2South East Asia Regional Office, WHO, New Delhi, 3Centre for Community Medicine, All India Institute of Medical Sciences, India Correspondence to: Dr. Krishna M Palipudi, E‑mail: [email protected]
Abstract
INTRODUCTION: Tobacco use is a leading cause of deaths and Disability Adjusted Life Years lost worldwide, particularly in South‑East Asia.
Health risks associated with exclusive use of one form of tobacco alone has a different health risk profile when compared to dual use. In order to tease out specific profiles of mutually exclusive categories of tobacco use, we carried out this analysis. METHODS: The Global Adult Tobacco Survey (GATS) data was used to describe the profiles of three mutually exclusive tobacco use categories (“Current smoking only,” “Current smokeless tobacco [SLT] use only,” and “Dual use”) in four World Health Organization South‑East Asia Region countries, namely Bangladesh, India, Indonesia and Thailand. GATS was a nationally representative household‑based survey that used a stratified multistage cluster sampling design proportional to population size. Prevalence of different forms of usage were described as proportions. Logistics regression analyses was performed to calculate odds ratios (OR) with 95% confidence intervals. All analyses were weighted, accounted for the complex sampling design and conducted using SPSS version 18. RESULTS: The prevalence of different forms of tobacco use varied across countries. Current tobacco use ranged from 27.2% in Thailand to 43.3% in Bangladesh. Exclusively smoking was more common in Indonesia (34.0%) and Thailand (23.4%) and less common in Bangladesh (16.1%) and India (8.7%). Exclusively using SLT was more common in Bangladesh (20.3%) and India (20.6%) and less common on Indonesia (0.9%) and Thailand (3.5%). Dual use of smoking and SLT was found in Bangladesh (6.8%) and India (5.3%), but was negligible in Indonesia (0.8) and Thailand (0.4%). Gender, age, education and wealth had significant effects on the OR for most forms of tobacco use across all four countries with the exceptions of SLT use in Indonesia and dual use in both Indonesia and Thailand. In general, the different forms of tobacco use increased among males and with increasing age; and decreased with higher education and wealth. The results for urban versus rural residence were mixed and frequently not significant once controlling for the other demographic factors. CONCLUSION: This study addressed the socioeconomic disparities, which underlie health inequities due to tobacco use. Tobacco control activities in these countries should take in account local cultural, social and demographic factors for successful implementation. Key Words: Global Adult Tobacco Survey dual use, prevalence, South‑East Asia Region, socioeconomic determinants, tobacco use
Introduction Tobacco use is responsible for six million deaths and 6.3% of Disability Adjusted Life Years (DALY) lost worldwide. [1] Despite decreases in tobacco smoking in the high‑income regions, the global estimates remain stable due to compensating increase in the South‑East Asia Region (SEAR).[2] The burden due to all forms of tobacco use (which also includes smokeless tobacco [SLT] use) would be much higher than these estimates. Majority of this burden due to tobacco use is borne by Low‑ and Middle‑Income Countries, because of their large population size and widely prevalent tobacco use.[1,3] Dual tobacco use, popularly defined as the concurrent use of tobacco smoking and at least one other form of SLT,[4] has been perceived as a threat to global tobacco control efforts in recent times. This particular type of use has emerged at least partly as the tobacco industry’s response to overwhelming scientific evidence of harm caused to human life by smoking.[5] SLT products, which are cheaper and lack the disadvantage of violating smoking laws such as “use in public places” and which purportedly help in quitting smoking. They are likely to be co‑adopted by those who are already smoking, in Access this article online Quick Response Code:
Website: www.indianjcancer.com DOI: 10.4103/0019-509X.147446 PMID: *******
S24
order to quit smoking or as a means of continued nicotine supply in public places. The question of dual tobacco use causing greater harm when compared to single product use is still open to debate.[6‑9] Although many studies [10‑15] have described the sociodemographic profile of tobacco use in SEAR, very few have tried to examine exclusive categories of tobacco use such as smoking only, SLT use only and dual use. The need for such analyses arises from the fact that use of one type alone has a different health risk profile as compared to dual use. To tease out the specific profile of these exclusive categories we needed to analyze them separately. In order to effectively plan for future tobacco control strategies we need to understand dual use and its determinants as opposed to studies, which focus on all tobacco use. The Global Adult Tobacco Survey (GATS) provided an excellent opportunity to examine the prevalence and determinants of exclusive categories of tobacco use. Methods We used the GATS data to describe the profile of tobacco users in four World Health Organization (WHO) SEAR countries, namely Bangladesh, India, Indonesia and Thailand. Detailed methodology of GATS in each country has been described elsewhere, [16‑19] but briefly GATS was a nationally representative household‑based survey, designed to obtain data on tobacco use behaviors of civilian noninstitutionalised individuals aged ≥15 years. Each country used a stratified multistage cluster sampling design to produce nationally representative samples. The important Indian Journal of Cancer | Dec 2014 | Volume 51 | Supplement 1
Palipudi, et al.: Tobacco use in the SEAR
survey details are given in Table 1. A minimum sample size of 8000 households was taken in each country, except India where 8000 households were chosen from each of six regions. Households were selected at random and in each selected household face‑to‑face interviews were done in the participant’s local language. Efforts were taken to maintain confidentiality of information obtained. Ethical clearance was obtained from country specific health ministries and implementing institutions.
followed by Thailand (23.4%), Bangladesh (16.1%) and India (8.7%). Prevalence of current SLT use was similar in India (20.6%) and Bangladesh (20.3%) followed by Thailand (3.5%) and Indonesia (0.9%). Prevalence of dual use was highest in Bangladesh (6.8%), followed by India (5.3%), Indonesia (0.8%) and Thailand (0.4%).
We defined three mutually exclusive groups of tobacco use: “Current smoking only,” “Current SLT use only,” and “Dual use” [Box 1].
Current tobacco use in males was highest in Indonesia (67.1%) and Bangladesh (58.0%) while the highest prevalence for females were in Bangladesh (28.7%) and India (20.3%). In all countries, a larger proportion of males used tobacco than females. Current tobacco use was also higher in rural areas (ranging from 29.2% in Thailand to 45.1% in Bangladesh) when compared to urban areas (ranging from 22.9% in Thailand to 38.1% in Bangladesh) in all four countries. There was a trend toward increasing prevalence with increasing age in all countries except in Indonesia, where the prevalence was slightly lower in the ≥ 65 years age group. Current tobacco use was lower with increasing education (except in Bangladesh) and with increasing wealth status (except in Indonesia) [Table 2].
The prevalence of different forms of usage were described as proportions with 95% confidence intervals [CI] and also within categories of sociodemographic variables such as gender (male, female), age group (15–24, 25–44, 45–64, 65 or older), residence (rural, urban), education (no formal education, primary, secondary, college/university or above) and wealth index quintiles (lowest, low, middle, high, highest). Association between different forms of usage and sociodemographic variables were calculated by performing logistic regression and strength of association described in terms of odds ratios (OR) with 95% confidence interval (CI). Where a particular cell contained zero, adjacent categories of explanatory variables were combined to allow the model to converge. All analyses were weighted, accounted for the complex sampling design and conducted using Statistical Package for the Social Sciences (SPSS) version 18. Results Prevalence of different forms of tobacco use [Figure 1]
The overall prevalence of current tobacco use was highest in Bangladesh (43.3%), followed by Indonesia (35.7%), India (34.6%) and lowest in Thailand (27.2%). Prevalence of current smoking only was highest in Indonesia (34%) Box 1: Definition of tobacco related terms used in GATS Definition Current tobacco use: Currently smoking tobacco in any form or using SLT or both Current smoking only: Currently smoking any tobacco product on a daily or less than daily basis exclusively Current SLT only: Currently using any SLT product on a daily or less than daily basis exclusively Dual use: Currently smoking any form of tobacco together with any form of SLT on a daily or less than daily basis
Prevalence of different forms of tobacco use by sociodemographic characteristics Current tobacco use
Current smoking only
Prevalence of current smoking only was very high in males (ranging from 15.0% in India to 65.6% in Indonesia) as compared to females (ranging from 0.8% in Bangladesh to 2.8% in Thailand) in all four countries. With increasing age, the prevalence also increased reaching a peak in 45–64 years age group and declining thereafter, except in Thailand where the peak was earlier in the age group
1RWREDFFRXVH 'XDOXVH &XUUHQW6/7RQO\ &XUUHQWVPRNLQJRQO\
%DQJODGHVK
,QGLD
,QGRQHVLD
7KDLODQG
Figure 1: Prevalence of different forms of tobacco use
SLT=Smokeless tobacco; GATS=Global adult tobacco survey
Table 1: Methodological details of the GATS survey by country Country
Survey dates
Un‑weighted numbers Sampled men
Sampled women
Overall response rate (%)
Coverage of population (%)
Bangladesh India Indonesia
July to August 2009 June 2009 to January 2010 October to November 2011
4468 33,767 3948
5161 35,529 4357
93.6 91.8 94.3
95.5 99.9 99.9
Thailand
February to May 2009
10,052
10,514
94.2
98.3
GATS=Global adult tobacco survey
Indian Journal of Cancer | Dec 2014 | Volume 51 | Supplement 1
S25
S26
OR
India (95% CI) Prevalence
OR
OR
Indonesia (95% CI) Prevalence
28.1 (25.1, 31.2)
1.0
17.2 (15.8, 18.6)
*P
