You are on page 1of 7

Nicotine & Tobacco Research, 2016, 750756

doi:10.1093/ntr/ntv286
Original investigation
Advance Access publication January 4, 2016

Original investigation

Prevalence and Sociodemographic


Determinants of Any Tobacco Use and Dual Use
in Six Countries of the WHO South-East Asia
Region: Findings From the Demographic and
Health Surveys
Dhirendra N.Sinha PhD1, Rizwan A.Suliankatchi MD2, RitvikAmarchand
PhD3, AnandKrishnan PhD3
1
Tobacco Free Initiative Unit, World Health Organization, Regional Office for South-East Asia, New Delhi, India;
2
Department of Community Medicine, Velammal Medical College Hospital and Research Institute, Madurai, India;
3
Centre for Community Medicine, All India Institute of Medical Sciences, New Delhi, India

Corresponding Author: Dhirendra N.Sinha, PhD, Tobacco Free Initiative Unit, World Health Organization, Regional Office for
South-East Asia, I.P. Estate, New Delhi 110002, India. Telephone: 91-11-23309501; Fax: 91-11-23705714; E-mail: sinhad@who.int

Abstract
Background: Tobacco control is an important strategy to reduce the disease burden caused by sev-
eral noncommunicable diseases. An in-depth understanding of the sociodemographic variations
in tobacco use is an important step in achieving effective tobacco control.
Aims: We aimed to estimate the age-standardized prevalence of any tobacco use and dual tobacco
use and determine their association with sociodemographic variables in six countries (Bangladesh,
Indonesia, India, Maldives, Nepal, and Timor Leste) of the WHO South-East Asia Region.
Methods: The main outcome variables any tobacco use and current dual use were created
from the latest available Demographic and Health Surveys data for each country. The prevalence
estimates were weighted using sample weights and age standardized using the WHO standard
population. Associations between the sociodemographic variables and tobacco use were calcu-
lated by performing multivariable logistic regression analysis. Analyses were performed in Stata
12 using svyset and svy commands.
Results: The highest prevalence of any tobacco use among men was in Indonesia (76.4%) and
among women in Nepal (15.7%). Also, Nepal had the highest prevalence of dual tobacco use in
both men (17.9%) and women (1.5%). With regard to sociodemographic determinants, despite the
inter-country variations, any and dual tobacco use were significantly associated with age, higher
education, greater wealth, rural residence, and ever-married marital status. The poor and unedu-
cated had a higher odds ratio for these practices.
Conclusion: Prevalence of dual tobacco use and its underlying socioeconomic disparities should
be taken into account for the planning of tobacco control activities in the region.
Implications: The dual tobacco use phenomenon is being increasingly recognized as a distinct
entity in the fight against tobacco addiction. When compared with single product users, dual users
have a greater risk of developing tobacco related diseases and are less likely to quit their habits.
However, this phenomenon has not been studied adequately in the South-East Asia region. In
this context, this study has provided a detailed and comprehensive view of dual tobacco use and

The Author 2016. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. 750
For permissions, please e-mail: journals.permissions@oup.com.
Nicotine & Tobacco Research, 2016, Vol. 18, No. 5 751

its sociodemographic determinants in six countries of the region. This study recommends that
tobacco control interventions should be targeted specifically at the disadvantaged sections of the
society, such as the poor and the uneducated, who are more likely to engage in dual as well as
any tobacco use. This study could prove as an important reference and tool for policy making in
the South-East Asia region.

Introduction used a pretested standard questionnaire, translated in appropriate


local languages. Strict quality control measures were employed to
Control of tobacco use has been considered one of the priority tar-
maintain the data quality. The methods were fairly uniform in all
gets for reducing disease burden caused by a number of disease cat-
study countries. The periods of data collection were as follows: 2007
egories such as cardiovascular diseases, peripheral vascular diseases,
in Bangladesh, 2012 in Indonesia, 20052006 in India, 2009 in
chronic respiratory diseases, and cancers.1 As a result of effective
Maldives, 2011 in Nepal and 20092010 in Timor Leste.
tobacco control policies, usage has been declining in the Western
countries in the recent decades.2 However, the developing countries
which bear the major brunt of tobacco related illnesses have not been Tobacco and Sociodemographic Variables
very successful in achieving larger magnitudes of reductions2,3 prob- The DHS used a set of four standard questions on tobacco use in all
ably because the traditional tobacco control policies do not always the countries and some country specific options on certain tobacco
necessarily target the interventions towards the vulnerable sections products. All the questions were framed with relevance to current
of the society defined by race, gender, income, and residence.4 users of tobacco only and the reference time period was not defined.
Understanding the sociodemographic variations in tobacco use For the purpose of this study, we analyzed two tobacco use vari-
is an important step in achieving effective tobacco control. Several ables. The current dual use variable included those individuals
studies have previously endeavored to examine the socioeco- who reported the use of at least one tobacco smoking product and at
nomic determinants of tobacco use.510 Although such studies have least one SLT product concurrently. The current any tobacco use
included the various forms of tobacco use, they have not examined variable included those individuals who reported the use of at least
dual tobacco use in great detail. Dual use is commonly defined as one tobacco product (smoking or SLT) at present. All these variables
the use of a smokeless tobacco (SLT) product along with tobacco were created separately for both genders. Aset of six variables col-
smoking.11 Reasons for the emergence of the dual use phenomenon lected in the DHS was used to explain the sociodemographic vari-
vary between developed and developing countries and therefore a ations in tobacco use. These included age, residential area, marital
regional analysis of dual use assumes importance. status, education, religion and wealth index. To maintain uniformity
Tobacco research has seen a gradual realization of the impor- across the surveys analysis was restricted to 1549years of age.
tance of other forms of tobacco usage such as SLT products and
dual or concomitant tobacco use.12,13 Dual users need to be given a Statistical Methods
higher priority for control because they are more prone to develop The prevalence of dual and any tobacco use was reported along with
tobacco related diseases and are less likely to quit than single prod- 95% confidence intervals (CI). All the proportions were weighted
uct users.14,15 Therefore, the aim of the present study was to estimate using sample weight provided by the DHS to account for complex
age-standardized prevalence of dual use and any tobacco use and sample design and age-standardized using the WHO 20002025
to determine their association with selected sociodemographic vari- standard population proportions to provide comparable estimates
ables in six countries of the WHO South-East Asia region, using the across countries.17 Associations between sociodemographic vari-
latest Demographic and Health Surveys (DHS) data. ables and tobacco use were calculated by performing multivariable
logistic regression to provide adjusted odds ratios (OR) for the vari-
Materials and Methods able under consideration, while adjusting for all the other variables
in the model. All analyses were performed in Stata (Stata Statistical
Data Sources Software: Release 12. College Station, TX: StataCorp LP.) using the
We used the datasets of six countries (Bangladesh, Indonesia, svyset and svy commands to account for the complex multistage
India, Maldives, Nepal, and Timor Leste) of the WHO South- stratified sampling design of the DHS.
East Asia region from the DHS program. These surveys were con-
ducted between 2006 and 2012 in the study countries by Macro Ethics Statement
International Inc and country-level research organizations and were Ethical clearance was not sought since the present study involved only a
funded by the USAID. The methodological details of DHS have been secondary analysis of publicly available survey data. However, the origi-
described elsewhere.16 Briefly, they used a multistage stratified ran- nal DHS obtained clearance from the Institutional Review Boards of all
dom sampling with Population Proportional to Size (PPS) technique. the participating agencies and the participants were informed about the
The sample was meant to be nationally representative. Although in option of voluntary participation and the confidentiality of information.
most countries all the eligible persons in the reproductive age group
(15 to 49years) were sampled, in few countries, only the currently
married persons were selected. Results
Prevalence of Current Any Tobacco Use and DualUse
Data Collection It has to be noted that all proportions were age standardized (to
Trained interviewers collected data from the members of the enable cross-country comparisons) and therefore, they do not repre-
selected households by a face-to-face interview. The interviewers sent the actual prevalence.
752 Nicotine & Tobacco Research, 2016, Vol. 18, No. 5

Among men, the highest prevalence of any tobacco use was A number of studies have been previously carried out using com-
seen in Indonesia (76.4%), followed by Timor Leste (73.2%), parable DHS datasets.6,10,18,19 However, in comparison to the cur-
Bangladesh (66.9%), India (58.1%), Nepal (55.6%), and the lowest rent study, the studies cited did not take certain important issues
was in Maldives (51.2%). Among women, the highest prevalence into accountnot analyzing dual use prevalence, not examining the
of any tobacco use was seen in Nepal (15.7%), followed by India sociodemographic determinants of dual use, and not performing the
(11.8%), Maldives (8.4%), Timor Leste (5.4%), and the lowest was age-standardization in some cases. Another study carried out using
in Indonesia (2.7%). Among men, the highest prevalence of dual Indian DHS reported only the time trends in the sociodemographic
use was observed in Nepal (17.9%), followed by India (13.4%), determinants of dual use.19
Bangladesh (9.5%), Timor Leste (2.6%), Maldives (1.5%), and the Studies on dual tobacco use carried out using other datasets (such
lowest prevalence in Indonesia (0.2%). Among women, dual use was as the Global Adult Tobacco Survey) have been published for some
low in all the countries with the highest prevalence in Nepal (1.5%) of the study countries. The previous investigation of dual use in three
and the lowest in Indonesia (0.1%; Table1, Figure1). of the present study countries (Bangladesh, Indonesia and India) by
the authors of the present study, using the Global Adult Tobacco
Association With Sociodemographic Determinants Survey data,20 showed that the sociodemographic determinants were
Among men, any tobacco use was significantly associated with age, similar to the current study, thereby adding credibility to the find-
only in India, Nepal and Timor Leste. In India, the OR increased with ings. Other authors who analyzed the Global Adult Tobacco Survey
increasing age whereas an inverted U shaped relationship was noted data have also corroborated this finding for India.21,22 However, there
in Nepal (peak in 3539years) and Timor Leste (peak in 2529years). was one contradiction with regard to the association with rural resi-
Among women also, any tobacco use was significantly associated with dence. The present study suggested that rural areas were less likely
age and the magnitude increased with increasing age in all the countries. to report dual use while the above reports suggested otherwise. The
Among men, current dual use was significantly associated with age, only reasons for this were not immediately clear and might be related
in India, Nepal, and Timor Leste. In Timor Leste, there was an inverted to differences in sample acquisition. A study conducted using the
U shaped relationship with age, with the peak OR in the 3035years WHO-Stepwise method among Bangladeshi men also arrived at con-
age group. Among women, the OR for dual use significantly increased clusions similar to the Indian studies, with regard to the dual tobacco
with age only in India and Nepal. Rural areas had a lower OR for use prevalence and its determinants.23 Other nations in the region
any tobacco use as compared to urban areas among men in India and also appeared to follow the general sociodemographic distribution
among women in Indonesia, Maldives and Timor Leste. Similarly, rural for dual use.24 The dual use phenomenon has also been explored
areas had a lower OR for dual use when compared with urban areas in the developed countries such as the United States, Norway, and
among men in India and among women in Timor Leste. In comparison Sweden.2527 In a study carried out among US men using nationally
to never-married persons, ever-married persons were found to have a representative survey data, it was seen that dual use was uncommon
higher OR for any tobacco use among men in India and Maldives and on the whole, but it was more common in young adults than in older
among women in all the countries. Similarly, ever-married persons were adults, especially in males.26 The study carried out in Norwegian
found to have a higher OR for dual use among men in India and Nepal men found a relatively small overall prevalence of dual use but a
and among women in Indonesia. The OR for any tobacco use decreased relatively greater prevalence in younger adults.27
with increasing educational status in both the genders in all the coun- With respect to the sociodemographic determinants of tobacco use,
tries (except among men in Timor Leste). In Timor Leste, men who were in general, the results of the present study were consistent with several
educated up to primary and secondary school reported the highest OR studies. The relationship between age and tobacco use was similar to
for any tobacco use as compared to men with no or higher education. that reported by various other surveys.7,9,28 The gradient of increas-
Similarly, dual use also had an inverse relationship with educational ing ORs from never married to currently married to formerly married
status in both the genders in all the countries. Asignificant inverse rela- persons has been previously stressed for tobacco smoking.12,29 The pro-
tionship between any tobacco use and wealth index was noted among tective effects of higher education and greater wealth observed in our
men in Bangladesh, Indonesia, and Nepal and among women in all study for any tobacco use were also similar to the studies carried out in
the countries. Similarly, a clear inverse relation between dual use and other regions of the world.7,9,28,30 Among the several reasons offered for
wealth index was noted among men in Indonesia and among women in this lopsided socioeconomic distribution, marketing strategies that tar-
Indonesia and Maldives (Table1). get people in the lower rungs of society have been widely blamed.31,32

Implications of the Findings


Discussion It is well known that dual tobacco users form a special risk group for
Large inter-country variations were observed in the prevalence of dual two main reasons. Firstly, the risk of oral, pharyngeal and esophageal
tobacco use and any tobacco use. The problem of dual use was very cancers and coronary heart disease was found to be higher with dual
prominent among the men of Bangladesh, India and Nepal. Men of use.14,15,33 Secondly, quit rate or intention to quit was lower and nico-
other countries and women of all the study countries reported only a tine addiction rates were higher among dual users.25,3436 Hence, one
low level of dual use. It was interesting to note that the three countries must consider them as separate entities and their sociodemographic
(Bangladesh, India, and Nepal) that had a significant burden of dual profiles must be scrutinized more meticulously. The dual use phe-
use, also had a prominent SLT use burden (data not shown). Agen- nomenon is quite complex and is influenced by a number of factors
eral pattern suggested that the dual use prevalence mirrors the preva- depending on the market forces and vulnerability of the consumers.
lence of SLT use and it is also generally low in countries dominated by Many authors acknowledge the fact that further research is required
exclusive smoking. This pattern might be related to the availability and to have a greater understanding of the dual use phenomenon in both
acceptability of smokeless products in the specific country. However, developed and developing nations.21,26,27 The drivers and determi-
such assumptions need to be verified by further longitudinal studies. nants of dual use are likely to be different for the developed and
Table1. Age-Standardized Weighted Prevalence and Sociodemographic Determinants of Current Any Tobacco Use and Current Dual Use by Gender and Country
Bangladesh Indonesia India Maldives Nepal Timor Leste

Any Tobacco Use Dual Use Any Tobacco Use Dual Use Any Tobacco Use Dual Use Any Tobacco Use Dual Use Any Tobacco Use Dual Use Any Tobacco Use Dual Use

Men
Prevalence, 66.9 (62.4, 71.1) 9.56 (7.52, 12.0) 76.4 (74.3, 78.4) 0.19 (0.11, 0.31) 58.1 (57.4, 58.8) 13.4 (12.9, 13.9) 51.2 (43.2, 59.2) 1.51 (0.97, 2.32) 55.6 (53.4, 57.8) 17.9 (16.0, 20.1) 73.2 (71.4, 74.8) 2.64 (2.00, 3.48)
% (95% CI)
OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
Age group
1519 1 1 1 1 1 1 1 1 1 1
2024 1.61 (0.50, 5.19) 1.61 (0.17, 15.1) 0.45 (0.14, 1.43) 1 2.79 (2.57, 3.03) 3.57 (3.07, 4.15) 0.98 (0.14, 6.72) 3.02 (2.22, 4.10) 5.23 (2.93, 9.33) 5.69 (4.29, 7.56) 8.12 (3.60, 18.2)
2529 1.26 (0.41, 3.85) 1.79 (0.21, 14.8) 0.41 (0.13, 1.26) 5.78 (0.64, 52.1) 3.50 (3.18, 3.85) 4.38 (3.67, 5.23) 1.05 (0.16, 6.98) 1 4.71 (3.25, 6.82) 8.99 (4.73, 17.0) 11.9 (8.28, 17.3) 20.1 (6.36, 63.6)
3034 1.67 (0.52, 5.32) 2.02 (0.23, 17.3) 0.38 (0.12, 1.15) 2.68 (0.30, 23.5) 3.69 (3.31, 4.11) 4.30 (3.55, 5.21) 1.15 (0.17, 7.73) 0.43 (0.06, 2.76) 4.20 (2.79, 6.32) 7.44 (3.82, 14.4) 9.16 (6.07, 13.8) 24.4 (5.48, 108)
3539 1.68 (0.55, 5.11) 2.30 (0.27, 19.1) 0.37 (0.12, 1.12) 3.9 (0.44, 34.5) 3.87 (3.46, 4.33) 4.33 (3.56, 5.27) 0.91 (0.13, 6.27) 1.63 (0.31, 8.57) 5.26 (3.58, 7.72) 10.0 (4.89, 20.6) 5.65 (3.82, 8.34) 11.9 (3.20, 44.6)
4044 2.18 (0.69, 6.79) 3.62 (0.42, 30.5) 0.29 (0.09, 0.91) 1.50 (0.15, 14.3) 4.06 (3.61, 4.57) 4.03 (3.28, 4.95) 0.89 (0.13, 6.15) 1.02 (0.19, 5.29) 4.75 (3.09, 7.32) 10.6 (4.86, 23.5) 5.93 (3.74, 9.41) 10.6 (2.59, 43.4)
4549 3.45 (1.08, 10.9) 6.61 (0.76, 57.4) 0.30 (0.09, 0.93) 7.70 (0.80, 74.0) 3.98 (3.54, 4.48) 3.58 (2.90, 4.42) 1.03 (0.14, 7.13) 0.98 (0.15, 6.26) 4.35 (2.67, 7.08) 8.09 (3.54, 18.5) 5.32 (3.46, 8.16) 12.7 (2.74, 58.7)
Residence
Urban 1 1 1 1 1 1 1 1 1 1 1 1
Rural 0.90 (0.69, 1.17) 0.80 (0.51, 1.24) 0.97 (0.82, 1.15) 0.70 (0.23, 2.12) 0.83 (0.78, 0.90) 0.79 (0.70, 0.89) 1.02 (0.51, 2.03) 0.40 (0.07, 2.23) 1.03 (0.81, 1.32) 1.14 (0.79, 1.63) 1.13 (0.88, 1.44) 0.95 (0.46, 1.96)
Nicotine & Tobacco Research, 2016, Vol. 18, No. 5

Marital status
Never married 1 1 1 1 1 1
Currently married 1 1 1.49 (1.38, 1.60) 1.49 (1.32, 1.69) 1 1.78 (1.39, 2.27) 1.71 (1.08, 2.72) 1.12 (0.84, 1.50) 0.73 (0.26, 2.03)
Formerly married 0.70 (0.21, 2.33) 0.16 (0.02, 1.23) 2.59 (1.99, 3.37) 3.10 (2.17, 4.42) 1.78 (0.91, 3.48) 4.95 (1.74, 14.0) 4.60 (1.26, 16.7) 1.12 (0.49, 2.58) 0.68 (0.13, 3.56)
Education
No education 1 1 1 1 1 1 1 1 1 1 1 1
Primary 0.70 (0.54, 0.91) 0.66 (0.43, 1.02) 1.68 (1.00, 2.81) 0.48 (0.08, 2.65) 0.92 (0.83, 1.01) 0.93 (0.82, 1.06) 0.76 (0.51, 1.15) 0.51 (0.18, 1.47) 0.74 (0.53, 1.03) 0.56 (0.36, 0.85) 0.91 (0.72, 1.14) 0.66 (0.35, 1.23)
Secondary 0.44 (0.32, 0.59) 0.27 (0.15, 0.46) 1.37 (0.81, 2.31) 0.16 (0.02, 0.96) 0.60 (0.55, 0.65) 0.60 (0.53, 0.67) 0.60 (0.36, 0.99) 0.16 (0.03, 0.69) 0.37 (0.27, 0.52) 0.29 (0.18, 0.46) 0.56 (0.44, 0.72) 0.30 (0.16, 0.57)
Higher 0.28 (0.20, 0.39) 0.15 (0.07, 0.31) 0.66 (0.38, 1.16) 0.33 (0.30, 0.37) 0.27 (0.22, 0.33) 0.32 (0.15, 0.69) 0.68 (0.08, 5.64) 0.21 (0.14, 0.30) 0.06 (0.03, 0.12) 0.31 (0.19, 0.48) 0.21 (0.05, 0.81)
Religion
Islam 1 1 1 1 1 1
Hinduism 0.97 (0.70, 1.34) 0.77 (0.45, 1.30) 0.89 (0.81, 0.98) 1.05 (0.90, 1.24) 0.85 (0.52, 1.38) 0.75 (0.33, 1.70)
Christianity 2.12 (0.62, 7.21) 2.98 (1.93, 4.61) 0.69 (0.58, 0.83) 0.90 (0.67, 1.21) 0.22 (0.10, 0.51) 0.23 (0.07, 0.77)
Others 0.41 (0.35, 0.49) 0.42 (0.32, 0.57) 0.64 (0.36, 1.15) 0.28 (0.11, 0.74)
Wealth index
Poorest 1 1 1 1 1 1 1 1 1 1 1 1
Poorer 0.76 (0.53, 1.08) 0.81 (0.48, 1.37) 0.89 (0.71, 1.12) 0.02 (0.00, 0.14) 0.86 (0.77, 0.95) 0.77 (0.67, 0.89) 0.91 (0.59, 1.39) 0.37 (0.11, 1.16) 0.85 (0.62, 1.17) 1.12 (0.74, 1.69) 0.91 (0.69, 1.20) 1.23 (0.57, 2.66)
Middle 0.70 (0.51, 0.96) 0.81 (0.50, 1.33) 0.78 (0.61, 0.99) 0.07 (0.02, 0.20) 0.62 (0.56, 0.69) 0.47 (0.41, 0.55) 0.76 (0.52, 1.13) 0.38 (0.11, 1.29) 0.98 (0.66, 1.45) 1.25 (0.73, 2.14) 0.88 (0.66, 1.17) 0.58 (0.27, 1.20)
Richer 0.60 (0.43, 0.84) 0.46 (0.27, 0.80) 0.55 (0.42, 0.71) 0.02 (0.00, 0.13) 0.46 (0.41, 0.52) 0.32 (0.27, 0.38) 0.66 (0.41, 1.09) 0.10 (0.02, 0.41) 0.70 (0.48, 1.03) 0.72 (0.43, 1.23) 0.83 (0.64, 1.09) 0.68 (0.31, 1.47)
Richest 0.44 (0.30, 0.64) 0.31 (0.17, 0.57) 0.44 (0.33, 0.58) 0.12 (0.02, 0.53) 0.31 (0.27, 0.35) 0.21 (0.17, 0.25) 1.11 (0.51, 2.41) 0.12 (0.01, 1.39) 0.67 (0.45, 0.99) 0.57 (0.32, 1.00) 0.67 (0.49, 0.91) 0.39 (0.16, 0.93)
Women
Prevalence, 2.65 (2.39, 2.93) 0.09 (0.05, 0.15) 11.8 (11.3, 12.3) 0.41 (0.35, 0.48) 8.42 (7.50, 9.44) 0.03 (0.01, 0.10) 15.7 (14.2, 17.3) 1.45 (1.09, 1.94) 5.45 (4.92, 6.03) 0.83 (0.66, 1.03)
% (95% CI)
Age group
1519 1 1 1 1 1 1 1 1
2024 1.53 (0.91, 2.58) 0.90 (0.14, 5.86) 1.50 (1.30, 1.72) 1.33 (0.58, 3.03) 1.11 (0.28, 4.40) 1.65 (0.99, 2.74) 1 4.07 (2.05, 8.06) 2.01 (0.62, 6.44)
2529 1.96 (1.15, 3.35) 1.56 (0.25, 9.66) 2.30 (1.98, 2.66) 1.93 (0.88, 4.24) 0.91 (0.22, 3.75) 4.49 (2.64, 7.66) 25.0 (4.46, 140) 4.86 (2.43, 9.70) 2.60 (0.78, 8.62)
3034 2.16 (1.24, 3.78) 0.59 (0.09, 3.58) 3.24 (2.78, 3.77) 2.68 (1.21, 5.90) 2.30 (0.57, 9.32) 6.18 (3.59, 10.6) 62.4 (8.36, 466) 6.94 (3.46, 13.9) 1.89 (0.57, 6.24)
3539 2.74 (1.55, 4.83) 1.05 (0.14, 7.46) 4.34 (3.73, 5.06) 5.39 (2.43, 11.9) 3.15 (0.78, 12.7) 9.63 (5.67, 16.3) 90.7 (13.0, 633) 12.1 (6.03, 24.3) 3.68 (1.06, 12.7)
4044 2.99 (1.62, 5.52) 0.56 (0.09, 3.48) 4.92 (4.21, 5.74) 4.98 (2.24, 11.0) 4.71 (1.16, 19.0) 15.9 (9.02, 28.1) 205. (28.6, 1476) 15.9 (8.08, 31.4) 4.24 (1.14, 15.7)
4549 4.02 (2.09, 7.70) 1.11 (0.14, 8.51) 5.87 (5.00, 6.88) 6.14 (2.77, 13.5) 6.59 (1.62, 26.7) 17.6 (10.2, 30.5) 161. (22.4, 1161) 18.4 (9.18, 37.1) 5.26 (1.50, 18.3)
Residence
Urban 1 1 1 1 1 1 1 1 1
Rural 0.74 (0.60, 0.91) 0.72 (0.30, 1.72) 0.88 (0.78, 0.99) 1.20 (0.81, 1.79) 0.52 (0.29, 0.91) 0.93 (0.69, 1.24) 0.56 (0.26, 1.19) 0.57 (0.42, 0.76) 0.53 (0.29, 0.95)
753
Table1. Continued
754

Bangladesh Indonesia India Maldives Nepal Timor Leste

Any Tobacco Use Dual Use Any Tobacco Use Dual Use Any Tobacco Use Dual Use Any Tobacco Use Dual Use Any Tobacco Use Dual Use Any Tobacco Use Dual Use

Marital status
Never married 1 1 1 1 1 1 1 1
Currently married 1.19 (0.79, 1.81) 6.10 (1.02, 36.2) 1.29 (1.12, 1.49) 2.03 (0.82, 5.01) 1 3.32 (2.15, 5.12) 0.47 (0.10, 2.17) 1.27 (0.89, 1.82) 1.94 (0.74, 5.05)
Formerly married 2.24 (1.43, 3.52) 13.5 (1.48, 123) 1.59 (1.35, 1.88) 2.88 (1.06, 7.81) 1.74 (1.19, 2.55) 4.16 (2.45, 7.06) 0.61 (0.10, 3.78) 1.66 (1.00, 2.75) 1.82 (0.55, 6.02)
Education
No education 1 1 1 1 1 1 1 1 1
Primary 0.40 (0.28, 0.57) 0.18 (0.04, 0.70) 0.92 (0.85, 1.00) 0.70 (0.49, 1.00) 0.62 (0.45, 0.84) 0.92 (0.76, 1.13) 0.46 (0.22, 0.94) 0.78 (0.62, 0.97) 0.53 (0.30, 0.93)
Secondary 0.50 (0.33, 0.75) 0.16 (0.03, 0.90) 0.53 (0.48, 0.59) 0.31 (0.19, 0.50) 0.16 (0.10, 0.27) 0.35 (0.25, 0.48) 0.17 (0.05, 0.56) 0.41 (0.30, 0.55) 0.33 (0.16, 0.65)
Higher 0.37 (0.22, 0.63) 0.02 (0.00, 0.37) 0.20 (0.16, 0.26) 0.29 (0.06, 1.23) 0.38 (0.06, 2.15) 0.07 (0.02, 0.19) 0.22 (0.08, 0.60)
Religion
Islam 1 1 1 1
Hinduism 0.93 (0.82, 1.05) 0.72 (0.43, 1.21) 2.02 (1.16, 3.53) 17.6 (2.06, 151)
Christianity 1.38 (1.10, 1.74) 1.85 (1.06, 3.24) 1.50 (0.62, 3.58) 19.0 (1.50, 241)
Others 0.89 (0.70, 1.14) 0.96 (0.39, 2.37) 3.76 (2.08, 6.77) 47.3 (5.39, 415)
Wealth index
Poorest 1 1 1 1 1 1 1 1 1
Poorer 0.54 (0.42, 0.70) 0.32 (0.09, 1.10) 0.64 (0.59, 0.71) 0.48 (0.32, 0.70) 0.79 (0.61, 1.03) 0.44 (0.35, 0.55) 0.53 (0.35, 0.80) 0.87 (0.66, 1.14) 0.92 (0.52, 1.64)
Middle 0.40 (0.31, 0.52) 0.41 (0.16, 1.00) 0.44 (0.39, 0.49) 0.22 (0.15, 0.33) 0.52 (0.39, 0.68) 0.29 (0.22, 0.38) 0.24 (0.13, 0.45) 0.75 (0.57, 0.98) 0.93 (0.53, 1.61)
Richer 0.44 (0.33, 0.59) 0.00 (0.00, 0.04) 0.30 (0.26, 0.33) 0.13 (0.07, 0.21) 0.36 (0.24, 0.54) 0.22 (0.16, 0.30) 0.17 (0.08, 0.35) 0.67 (0.50, 0.88) 0.64 (0.31, 1.33)
Richest 0.31 (0.22, 0.44) 0.28 (0.05, 1.40) 0.17 (0.14, 0.20) 0.07 (0.04, 0.13) 0.32 (0.15, 0.68) 0.12 (0.08, 0.18) 0.03 (0.00, 0.16) 0.64 (0.43, 0.95) 0.81 (0.37, 1.76)

CI=confidence interval; OR=odds ratio. Age group was coded into 5-year age groups from 15 to 49years. Residence was classified into urban and rural. Marital status included never married or in union, currently
married or in union and formerly married or in union (widowed, divorced and separated). Educational status included categories such as no education, primary, secondary and higher. Religion was categorized
into Islam, Hinduism, Christianity, and Others. The Others category constituted a heterogeneous group of religions that were different in different countries. The wealth index variable was categorized into five
quintiles, the lowest being the poorest and the highest being the richest.
Nicotine & Tobacco Research, 2016, Vol. 18, No. 5
Nicotine & Tobacco Research, 2016, Vol. 18, No. 5 755

Figure1. Prevalence of current exclusive tobacco smoking, current exclusive smokeless tobacco (SLT) use and current dual use by country and gender.

developing nations and therefore, require specific enquiries to avoid recognized as a distinct tobacco use indicator, especially in countries
oversimplified generalizations. Longitudinal studies are required to of the South-East Asia region where this phenomenon is prevalent.
track the exclusive use behaviors and understand the circumstances Future surveys and reports on tobacco use should report preva-
for the adoption of a second product or multiple products. The lence by exclusive categories and dual use. Future DHS should also
tobacco control policies of a country should take into account the take into consideration the limitations discussed above and make
factors driving the adoption of a second tobacco product and place the necessary modifications in the questionnaires, in order to make
a greater stress on providing cessation interventions for such users. them more useful and versatile for tobacco surveillance. The lop-
sided sociodemographic pattern of tobacco use has underscored
Strengths and Limitations the need to focus interventions specifically on the disadvantaged
The dual use phenomenon has been examined in great detail for sections of the society, such as the poor and the uneducated, similar
a number of countries in the South-East Asia region. Anumber of to smoking and SLT use. Since countries have several differences in
sociodemographic factors were analyzed together and analyses were the issues related to tobacco use, varied and localized interventions
conducted separately, for both the genders. The age standardization are needed for achieving effective tobacco control in the respective
of the prevalence estimates enabled a better and valid comparison regions.
between the countries. Other strengths of the study arise from the
strengths of the DHS itself, that is, a large nationally representative
Funding
sample with robust methodology, uniform definitions of independ-
ent and outcome variables and rigorous quality control. None declared.

A number of limitations arise when using DHS for tobacco


related data analysis and they have been discussed elsewhere.10 There
Declaration of Interests
are also a few other limitations. Firstly, the surveys were conducted
in different years in the various countries between 2005 and 2012 None declared.

but the short duration of 7years may not have influenced the results
significantly. Secondly, although the DHS were implemented using
Acknowledgments
the standard methodology in different countries, minor variations
at the country level due to variations in interview techniques and DNS and RAS conceived the study. RAS performed the data retrieval and sta-
tistical analysis. DNS, RA, AK critically modified the analysis. RAS wrote the
local language translation might have occurred. Thirdly and finally,
first draft. DNS, AK, and RA provided critical comments to the discussion. All
cross-country comparison was limited by inter-country differences in
authors read and approved the final manuscript.
eligibility criteria with regard to marital status. These points have to
be kept in context while interpreting the results.
References

1. World Health Organization. WHO Report on the Global Tobacco
Conclusions and Recommendations
Epidemic, 2008: The MPOWER Package. Geneva, Switzerland: World
This study has highlighted the need for an in-depth exploration Health Organization; 2008. www.who.int/tobacco/mpower/mpower_
of the dual use phenomenon, which needs to be increasingly report_full_2008.pdf. Accessed August 1, 2015.
756 Nicotine & Tobacco Research, 2016, Vol. 18, No. 5

2. Ng M, Freeman MK, Fleming TD, etal. Smoking prevalence and cigarette 19. Bhan N, Srivastava S, Agrawal S, et al. Are socioeconomic disparities
consumption in 187 countries, 19802012. JAMA. 2014;311(2):183192. in tobacco consumption increasing in India? A repeated cross-sectional
doi:10.1001/jama.2013.284692. multilevel analysis. BMJ Open. 2012;2(5):e001348. doi:10.1136/
3. Southeast Asia Initiative on Tobacco Tax (SITT) of the Southeast Asia bmjopen-2012-001348.
Tobacco Control Alliance (SEATCA). ASEAN Tobacco Tax Report Card 20. Palipudi K, Rizwan SA, Sinha DN, et al. Prevalence and sociodemo-
- Regional Comparisons and Trends. Bangkok, Thailand: Southeast Asia graphic determinants of tobacco use in four countries of the World
Tobacco Control Alliance (SEATCA); 2013. http://seatca.org/dmdocuments/ Health Organization: South-East Asia region: findings from the Global
ASEANTaxReportCardMay13forWEB.pdf. Accessed August 1, 2015. Adult Tobacco Survey. Indian J Cancer. 2014;51(Suppl 1):S2432.
4. Greaves L, Barr V. Filtered Policy: Women and Tobacco in Canada. British doi:10.4103/0019-509X.147446.
Columbia Centre of Excellence for Womens Health; 2000. www.cwhn.ca/ 21. Gupta PC, Ray CS, Narake SS, etal. Profile of dual tobacco users in India:
sites/default/files/FP-english-report.pdf. Accessed August 1, 2015. an analysis from Global Adult Tobacco Survey, 200910. Indian J Cancer.
5. Ciapponi A, Bardach A, Glujovsky D, Aruj P, Mazzoni A, Linetzky 2012;49(4):393400. doi:10.4103/0019-509X.107746.
B. Systematic Review of the Link Between Tobacco and Poverty. 22. Singh A, Ladusingh L. Prevalence and determinants of tobacco use in
Geneva, Switzerland: WHO; 2011. http://apps.who.int/iris/bitstr India: evidence from recent Global Adult Tobacco Survey data. PloS One.
eam/10665/44453/1/9789241500548_eng.pdf. Accessed August 1, 2015. 2014;9(12):e114073. doi:10.1371/journal.pone.0114073.
6. Ansara DL, Arnold F, Kishor S, Hsia J, Kaufmann R. Tobacco Use by Men 23. Zaman MM, Bhuiyan MR, Huq SM, Rahman MM, Sinha DN, Fernando
and Women in 49 Countries With Demographic and Health Surveys. DHS T. Dual use of tobacco among Bangladeshi men. Indian J Cancer.
Comparative Reports No. 31. Calverton, MD: ICF International; 2013. 2014;51(Suppl 1):S4649. doi:10.4103/0019-509X.147481.
www.dhsprogram.com/pubs/pdf/CR31/CR31.pdf. Accessed August 1, 2015. 24. Singh PK. Smokeless tobacco use and public health in countries of
7. Hosseinpoor AR, Parker LA, Tursan dEspaignet E, Chatterji S. South-East Asia region. Indian J Cancer. 2014;51(Suppl 1):S12.
Socioeconomic inequality in smoking in low-income and middle- doi:10.4103/0019-509X.147415.
income countries: results from the World Health Survey. PloS One. 25. McClave-Regan AK, Berkowitz J. Smokers who are also using smoke-
2012;7(8):e42843. doi:10.1371/journal.pone.0042843. less tobacco products in the US: a national assessment of characteristics,
8. Hosseinpoor AR, Parker LA, Tursan dEspaignet E, Chatterji S. Social behaviours and beliefs of dual users. Tob Control. 2011;20(3):239242.
determinants of smoking in low- and middle-income countries: results doi:10.1136/tc.2010.039115.
from the World Health Survey. PloS One. 2011;6(5):e20331. doi:10.1371/ 26. Tomar SL, Alpert HR, Connolly GN. Patterns of dual use of cigarettes and
journal.pone.0020331. smokeless tobacco among US males: findings from national surveys. Tob
9. Palipudi KM, Gupta PC, Sinha DN, etal. Social determinants of health and Control. 2010;19(2):104109. doi:10.1136/tc.2009.031070.
tobacco use in thirteen low and middle income countries: evidence from 27. Lund KE, McNeill A. Patterns of dual use of snus and cigarettes in a
Global Adult Tobacco Survey. PloS One. 2012;7(3):e33466. doi:10.1371/ mature snus market. Nicotine Tob Res. 2013;15(3):678684. doi:10.1093/
journal.pone.0033466. ntr/nts185.
10. Sreeramareddy CT, Pradhan PMS, Mir IA, Sin S. Smoking and smoke- 28. Pampel F. Tobacco use in sub-Sahara Africa: estimates from the
less tobacco use in nine South and Southeast Asian countries: prevalence demographic health surveys. Soc Sci Med. 2008;66(8):17721783.
estimates and social determinants from Demographic and Health Surveys. doi:10.1016/j.socscimed.2007.12.003.
Popul Health Metr. 2014;12:22. doi:10.1186/s12963-014-0022-0. 29. Kim S. Smoking prevalence and the association between smoking and
11. Klesges RC, Ebbert JO, Morgan GD, etal. Impact of differing definitions sociodemographic factors using the Korea National Health and Nutrition
of dual tobacco use: implications for studying dual use and a call for oper- Examination Survey data, 2008 to 2010. Tob Use Insights. 2012;5:1726.
ational definitions. Nicotine Tob Res. 2011;13(7):523531. doi:10.1093/ doi:10.4137/TUI.S9841.
ntr/ntr032. 30. Lopez AD, Collishaw NE, Piha T. A descriptive model of the cigarette
12. Lindstrm M. Social capital, economic conditions, marital status and daily epidemic in developed countries. Tob Control. 1994;3(3):242247.
smoking: a population-based study. Public Health. 2010;124(2):7177. doi:10.1136/tc.3.3.242.
doi:10.1016/j.puhe.2010.01.003. 31. Barbeau EM, Leavy-Sperounis A, Balbach ED. Smoking, social class, and
13. Mushtaq N, Williams MB, Beebe LA. Concurrent use of cigarettes and gender: what can public health learn from the tobacco industry about
smokeless tobacco among US males and females. J Environ Public Health. disparities in smoking? Tob Control. 2004;13(2):115120. doi:10.1136/
2012;2012:984561. doi:10.1155/2012/984561. tc.2003.006098.
14. Znaor A, Brennan P, Gajalakshmi V, et al. Independent and combined 32. Tuckson RV. Race, sex, economics, and tobacco advertising. J Natl Med
effects of tobacco smoking, chewing and alcohol drinking on the risk Assoc. 1989;81(11):11191124. www.ncbi.nlm.nih.gov/pmc/articles/
of oral, pharyngeal and esophageal cancers in Indian men. Int J Cancer. PMC2626111/pdf/jnma00905-0013.pdf. Accessed August 1, 2015.
2003;105(5):681686. doi:10.1002/ijc.11114. 33. Rahman MA, Zaman MM. Smoking and smokeless tobacco consump-
15. Teo KK, Ounpuu S, Hawken S, etal. Tobacco use and risk of myocardial tion: possible risk factors for coronary heart disease among young patients
infarction in 52 countries in the INTERHEART study: a case-control study. attending a tertiary care cardiac hospital in Bangladesh. Public Health.
Lancet. 2006;368(9536):647658. doi:10.1016/S0140-6736(06)69249-0. 2008;122(12):13311338. doi:10.1016/j.puhe.2008.05.015.
16. Corsi DJ, Neuman M, Finlay JE, Subramanian SV. Demographic and health 34. Kram Y, Klesges RC, Ebbert JO, Talcott W, Neilands TB, Ling PM. Dual
surveys: a profile. Int J Epidemiol. 2012;41(6):16021613. doi:10.1093/ tobacco user subtypes in the U.S. Air Force: dependence, attitudes, and
ije/dys184. other correlates of use. Nicotine Tob Res. 2014;16(9):12161223.
17. Ahmad OB, Boschi-Pinto C, Lopez AD, Murray CJL, Lozano R, Inoue M. doi:10.1093/ntr/ntu056.
Age Standardization of Rates: ANew WHO Standard. Report No.: GPE 35. Benowitz NL, Renner CC, Lanier AP, etal. Exposure to nicotine and carcin-
Discussion Paper Series: No.31. Geneva, Switzerland: WHO; 2001. www. ogens among Southwestern Alaskan Native cigarette smokers and smoke-
who.int/healthinfo/paper31.pdf. Accessed August 1, 2015. less tobacco users. Cancer Epidemiol Biomark Prev. 2012;21(6):934942.
18. Sreeramareddy CT, Ramakrishnareddy N, Harsha Kumar H, Sathian B, doi:10.1158/1055-9965.EPI-11-1178.
Arokiasamy JT. Prevalence, distribution and correlates of tobacco smoking 36. Lee S, Grana RA, Glantz SA. Electronic cigarette use among Korean
and chewing in Nepal: a secondary data analysis of Nepal Demographic adolescents: a cross-sectional study of market penetration, dual use, and
and Health Survey-2006. Subst Abuse Treat Prev Policy. 2011;6:33. relationship to quit attempts and former smoking. J Adolesc Health.
doi:10.1186/1747-597X-6-33. 2014;54(6):684690. doi:10.1136/tc.2010.039115.