A cross-sectional study of the prevalence and correlates of tobacco Use in Chennai, Delhi, and Karachi: data from the CARRS study
Berg et al. BMC Public Health
A cross-sectional study of the prevalence and correlates of tobacco Use in Chennai, Delhi, and Karachi: data from the CARRS study
Carla J Berg 0
Vamadevan S. Ajay
Mohammed K. Ali
Hassan M. Khan
Muhammad M. Kadir
KM Venkat Narayan
0 Department of Behavioral Sciences and Health Education, Emory University Rollins School of Public Health , 1518 Clifton Rd NE, 30322 Atlanta, Georgia , USA
Background: Tobacco burdens in India and Pakistan require continued efforts to quantify tobacco use and its impacts. We examined the prevalence and sociodemographic and health-related correlates of tobacco use in Delhi, Chennai (India), and Karachi (Pakistan). Methods: Analysis of representative surveys of 11,260 participants (selected through multistage cluster random sampling; stratified by gender and age) in 2011 measured socio-demographics, tobacco use history, comorbid health conditions, and salivary cotinine. We used bivariate and multivariate regression analyses to examine factors associated with tobacco use. Results: Overall, 51.8 % were females, and 61.6 % were below the age of 45 years. Lifetime (ever) tobacco use prevalence (standardized for world population) was 45.0 %, 41.3 %, and 42.5 % among males, and 7.6 %, 8.5 %, and 19.7 % among females in Chennai, Delhi, and Karachi, respectively. Past 6 month tobacco use prevalence (standardized for world population) was 38.6 %, 36.1 %, and 39.1 % among males, and 7.3 %, 7.1 %, and 18.6 % among females in Chennai, Delhi, and Karachi, respectively. In multivariable regression analyses, residing in Delhi or Karachi versus Chennai; older age; lower education; earning less income; lower BMI; were each associated with tobacco use in both sexes. In addition, semi-skilled occupation versus not working and alcohol use were associated with tobacco use in males, and having newly diagnosed dyslipidemia was associated with lower odds of tobacco use among females. Mean salivary cotinine levels were higher among tobacco users versus nonusers (235.4; CI: 187.0-283.8 vs. 29.7; CI: 4.2, 55.2, respectively). Conclusion: High prevalence of tobacco use in the South Asian region, particularly among men, highlights the urgency to address this serious public health problem. Our analyses suggest targeted prevention and cessation interventions focused on lower socioeconomic groups may be particularly important.
Tobacco use; Southeast Asia; Secondhand smoke exposure; Population studies
Tobacco use surveillance is critical, given that tobacco
use, particularly cigarette smoking, is the leading
preventable cause of mortality around the world . It
responsible for over 5 million deaths per year-more than
HIV/AIDS, tuberculosis, and malaria combined [1,2].
Smoking increases the risk of cardiovascular disease
(CVD) by 2 to 4 times, stroke by 2 to 4 times, and
diabetes mellitus by 30-40 % . In addition, smoking
interacts with other CVD risk factors, such as hypertension
and dyslipdemia, in increasing long-term CVD mortality
. Moreover, smoking causes diminished overall
heath, including self-reported poor health, increased
absenteeism from work, and increased health care
utilization and cost .
Between 1980 and 2012, the global estimated
agestandardized prevalence of daily tobacco smoking
declined by 25 % for men and by 42 % for women;
however, the population growth during this time
contributed to a 41 % increase in male daily smokers and a
7 % increase for female smokers . Moreover, over this
period, cigarette consumption worldwide increased by
26 %, indicating continued growth of the global tobacco
market . According to the World Health Organization
[WHO], if appropriate preventive measures are not
taken, the number of annual deaths will increase to 10
million per year by 2030, with 70 % of them taking place
in low-and middle-income countries . Thus,
continued efforts are needed to provide up-to-date estimates
of tobacco use, including the broad range of tobacco
products. Documenting correlates of use as well as the
impact of these products on tobacco-related health
conditions is critical for curbing the tobacco use pandemic
and the related chronic health conditions, as it will help
elucidate vulnerable populations and potential
A recent analysis of tobacco use data from nine
countries in South and Southeast Asia found high prevalence
of tobacco use in this region, ranging from 72 % in
Indonesia to 32 % in Pakistan (with Indias prevalence
being 34 %) . This study also documented the use of
tobacco in very diverse forms, particularly in India.
Another study in India from 2009 to 2011  found an
overall tobacco use prevalence of 21 %, with 40 % of
males being tobacco users compared to 5 % of females.
A 2013 study in Pakistan  documented a current
tobacco use prevalence (weighted to correspond to
ruralurban population proportions in the sample) of 45 %
among males and 6 % among females. In both India and
Pakistan, some predictors of tobacco use include older
age, male gender, low socioeconomic status, alcohol use,
and rural geographic location [10,12,13]. This high
prevalence is concerning given that South Asians have an
increased risk of tobacco-related diseases, such as CVD
and other CVD risk factors (e.g., diabetes, hypertension,
and dyslipidemia) .
Given the public health importance, it is critical to
track tobacco use within specific contexts in South Asia
and characterize tobacco use patterns in terms of
populations vulnerable to tobacco use and the types of
tobacco being used. In this manuscript, we report the
prevalence of tobacco use and its correlates among males
and females in three specific cities in South Asia Delhi,
Chennai, and Karachi using data from the Center for
Cardio-metabolic Risk Reduction in South Asia (CARRS)
Study. CARRS collects data regarding determinants (e.g.,
lifestyle factors) and occurrence of cardio-metabolic
diseases; patient-reported quality of life; and costs using
standardized tools and methods from representative samples
in these three cities.
This study was approved by the Independent Ethics
Committees of Public Health Foundation of India, All
India Institute of Medical Sciences, Madras Diabetes
Research Foundation and Aga Khan University, and Emory
Universitys Institutional Review Board. The CARRS Study
 builds on the WHO STEPS model  to
capture prevalence of risk factors, cardio-metabolic diseases
(CMDs), and their socioeconomic impact. The study sites
for the CARRS Study were metropolitan urban settings
with large, heterogeneous populations (Chennai,
population 4.68 million , Delhi, population 16.8 million ,
and Karachi, population 13 million ) that are growing
due to continued births and migration from various parts
of the country. The CARRS cohort study includes
representative cross-sectional surveys of these three cities in
South Asia as its baseline, conducted between October
2010 and December 2011 (see  for detailed methods).
Households were selected in each of the three cities
using multi-stage cluster random sampling techniques.
Each city has its own distinctive municipal sub-divisions,
encompassing municipal corporations, wards, and Census
Enumeration Blocks (CEB), which were used sequentially
as sampling frames to randomly select households. While
wards were the primary sampling units (PSUs) for Chennai
and Delhi, CEBs or clusters were the PSUs for Karachi.
STATA version 10.1 (Statacorp, TX) and data from the
most recent census (2001) were used to randomly select
the wards, CEBs, and households. To give each household
an equal chance of being selected for the study and to
identify households constructed after the last census
survey, manual listing and mapping of all households in each
CEB was done before randomly selecting them.
Two participants, one male and one female, aged
20 years or older, were selected from each household.
Those excluded from the study were pregnant women
and bed-ridden individuals. Two methods were used for
within household sampling. First, for households with
one to two adults (20 years), the sampling strategy
described in the 2002 Health Information National Trends
Study (HINTS) in the USA was used . According to
HINTS, one or both individuals (one male, one female)
were selected and enrolled into the study based on
eligibility criteria and informed consent. Second, for
households with more than two eligible adults, the Kish
method used in the WHOs STEPS surveys  was
applied. There are two main steps in KISH method.
First, all eligible participants from the household will be
ranked according to age in decreasing orders (males
followed by females). Participants are then selected using
KISH table identifying the last digit of household and
number of eligible participants . Recruitment of
participants and data collection were conducted through
three visits to each participants place of residence,
respectively. The sample size estimation, specifics on data
collection, data management efforts, and quality control
strategies have been published separately .
Correlates of tobacco Use
Sociodemographic Factors. Participants were asked to
report age; gender; education level (up to primary, high/
secondary, graduate, above graduation); occupation (not
working, professional [e.g., doctor, lawyer, large business
owner], trained [e.g., clerical, teacher, middle-level farmer],
skilled [e.g., small business owner, skilled manual labourer,
small farmer], semi-skilled [e.g., semi-skilled manual
laborer, carpenter], unskilled [e.g., landless laborer, unskilled
manual laborer); and income level per month (<10,000
Indian Rupees [INR] [<16,684 Pakistani Rupees [PKR]
or < USD$165] vs. 10,000-20,000 INR [16,684-33,368 PKR
or USD$165-330] vs. >20,000 INR [>33,368 PKR
or > USD$330]).
Health-Related Factors. Prior research has
documented lower body weight among current smokers than in
never or ex-smokers [24,25]; thus, we measured body
mass index (BMI; <18.0, 18-22.9, 23-24.99, > = 25) from
height and weight. The association between alcohol use
and tobacco use is also well documented [26,27]; thus,
participants were asked, How often do you use
alcoholic beverages? Those reporting using alcohol
occasionally or regularly and those who quit using alcohol
within the past six months were coded as current users
(given the low alcohol use in these countries compared
to western countries ); those reporting no use in the
past six months or never use were reported as nonusers.
Self-report and biological verification diabetes,
hypertension, and dyslipidemia were also included in the
current analyses, given the association of these factors
with smoking . Biological sample collection
involved drawing 15 ml of blood (in fasting state) and
collecting urine (early morning void) from each participant.
The samples were transported from field sites in cold
chain to the laboratories for analysis. Sample aliquots
were also stored in cryo-vials at - 80 degrees Celsius for
future studies. The methods of analysis and external
quality control have been standardized for all biological
samples across the study sites.
Self-reported diabetes was defined as the participant
reporting having diabetes (i.e., told by physician they have
diabetes). Newly-diagnosed diabetes was defined as not
having diagnosed diabetes but having FBG 126 mg/dl or
A1c 6.5 %. Prediabetes was defined as FBG 100-125 mg/dl
or HbA1c 5.7-6.5 %. If participants self-reported diabetes
but had FBG < 126 mg/dl or A1c < 6.5 %, they still were
considered to have self-reported diabetes, as treatment
or management may have altered FBG. Self-reported
hypertension was defined as the participants reporting
having hypertension (i.e., told by physician they have
hypertension). Newly-diagnosed hypertension was defined
as not reporting hypertension but having SBP 140 or
DBP 90 mmHg. Pre-hypertension was defined as SBP
120-139 or DBP 80-89 mmHg. Self-reported dyslipidemia
was defined as the participant reporting having high lipids
(i.e., told by physician they have high lipids).
Newlydiagnosed dyslipidemia was defined as the participant
reporting not having high lipids but having TC 200 mg/dl
or LDL 130 mg/dl. If participants self-reported
hypertension but had SBP < 140 or DBP < 90 mmHg, they
still were considered to have self-reported hypertension,
as treatment or management may have altered these
Tobacco exposure outcomes
Tobacco Use History. To assess tobacco use history,
participants were asked, Have you ever used tobacco in
any form (smoking, chewing, snuff, etc.)? and In what
forms have you consumed tobacco: In a smoking form?
In a chewed form? In any other form (snuff, toothpaste
etc)? Participants were also asked, At what age did you
first start smoking regularly? and At what age did you first
start consuming smokeless tobacco product regularly?
Current Tobacco Use. Participants were asked, Within
the past 6 months, do you currently consume tobacco:
regularly (once a week); occasionally (<once a week) or
not at all? Users were defined as those using regularly
or occasionally. Current users were asked, How often
do you use: Smoking form? Chewed form? Any other
form? with the same response options (i.e., regularly
(once a week); occasionally (<once a week) or not at all).
They were also asked to report on level of current and
past use of each of the following: tobacco smoking
options (cigarettes; beedis; cigars; hukka/chelum/pipe);
tobacco chewing options (pan with zarda; pan masala with
zarda; guthka); or other forms (snuff; others).
Secondhand Smoke (SHS) Exposure. Participants were
asked, Are you exposed to tobacco smoke from others
regularly (e.g. at home, at workplace regularly, while
travelling, any other place)? How many days a week?
How much time during a day? Those reporting at least
once a day in a week were coded as regularly exposed.
Biomarkers of Tobacco Exposure. Saliva samples were
collected from 191, 214, and 196 randomly selected
participants in Chennai, Delhi, and Karachi, respectively, to
biochemically verify self-reported tobacco use. This tool
was chosen due to greater acceptability of non-invasive,
non-stimulated salivary sampling  and the high
sensitivity of salivary cotinine in distinguishing active
tobacco use from passive smoking with lower discrepancy
between reported and measured prevalence as compared
with urine or blood . Also, Enzyme Immuno Assay
(EIA) cotinine results have shown near perfect agreement
with the reference standard GC/Mass Spectrometric
confirmation . Approximately 1000-2000 L of saliva was
collected in salivettes manufactured by SARSTEDT AG &
Co., Germany. The samples were transported from field
sites, in Cold chain, to the laboratories for analysis.
Quantitative estimation of cotinine was performed by Enzyme
immunoassay using kits from Salimetrics, PA 16803 USA.
Samples with higher cotinine levels were re-estimated
after further dilutions. The methods of analysis and
external quality control were standardized for all biological
samples across the study sites.
Stata 12.0 (College Station, Texas, USA) was used for
analysis. We used the svy technique for all analysis to
account for the complex survey design . Before any of
the survey estimation commands were used, the svyset
command was used to specify the variables that describe
the stratification, sampling weight, and primary sampling
unit variables. The age and sex standardized prevalence
(95 % CI) of ever and current tobacco use were
calculated for the three cities using World Banks estimates
for regional as well as world standard population for
comparison of prevalence. Bivariate analyses were done
to examine the relationship of sociodemographic and
health-related factors to current tobacco use (outcome)
among males and females, respectively. Multivariable
logistic regression analysis was then used to identify
independent correlates of tobacco use among males and
females, respectively, forcing the variables of interest
into the models. Because the relationship of age with
proportion of current tobacco use was curvilinear, we
included a quadratic term for age in multivariable analysis.
A p-value <0.05 was considered statistically significant.
A total of 17,274 individuals in 10,002 households were
approached in the three study sites (7,596 participants in
Chennai, 5,420 in Delhi, 4,258 in Karachi). From these, a
total of 16,288 participants (n = 7,760 men, n = 8,527
women) were recruited (the overall response rate was
94.3 % at the participant level; Chennai 90.9 %, n = 6,906;
Delhi 98.9 %, n = 5,365; and Karachi 94.3 %, n = 4,017).
Response rate for providing blood and urine were 81.8 %
(N = 13,327/16,288) and 84.3 % (N = 13,737/16,288),
respectively. The number of participants with sufficient data
retained to be able to conduct the final multivariate model
was 11,260 (n = 5,062 men; n = 6,198 women).
Overall, 51.8 % of the participants were female, 61.6 %
were below the age of 45 years, 27.5 % were in the 45-60
years age group, and 10.9 % were >60 years of age.
Additional file 1: Table S1 provides comparisons of the
CARRS sample to the regional and world population in
relation to age, indicating underrepresentation among
the younger age groups in the CARRS sample. (Note:
Comparisons to the city populations also suggested
similar underrepresentation of the younger age groups).
Lifetime (ever) tobacco use prevalence (standardized
for world population) was 45.0 %, 41.3 %, and 42.5 %
among males, and 7.6 %, 8.5 %, and 19.7 % among
females in Chennai, Delhi, and Karachi, respectively
(Table 1). The age and sex standardized estimates using
the regional population and world population are also
presented (Table 1). Past 6 month tobacco use
prevalence (standardized for world population) was 38.6 %,
36.1 %, and 39.1 % among males, and 7.3 %, 7.1 %, and
18.6 % among females in Chennai, Delhi, and Karachi,
respectively (Table 2).
Average age of initiation of tobacco smoking among
lifetime smokers was 23.4 (SD = 0.7), 24.9 (SD = 0.8), and
24.5 (SD = 1.2) years in Chennai, Delhi, and Karachi,
respectively (not shown in Tables). Average age of
initiation of smokeless tobacco among lifetime users was
29.5 (SD = 0.9), 31.8 (SD = 1.2), and 27.4 (SD = 1.6) years
in Chennai, Delhi, and Karachi, respectively. Among
males, the mean age at initiation of any smoked tobacco
use was 24.9 (SD = 0.4), 25.4 (SD = 0.4), and 24.6 (SD = 0.5)
years old in Chennai, Delhi, and Karachi, respectively. The
mean age at initiation for any smokeless tobacco use
among males was 30.4 (SD = 0.8), 29.1 (SD = 0.7), and 24.9
(SD = 0.8), respectively in Chennai, Delhi, and Karachi.
Similarly for female participants, the mean age at initiation
for any smoked tobacco use was 35.0 (SD = 0.1), 28.9
(SD = 1.7), and 28.6 (SD = 1.6) and that for any smokeless
Table 1 Ever use of tobacco products in Chennai, Delhi, and Karachi
Prevalence with 95 % CI (standardized using regional population)*
Prevalence with 95 % CI (standardized using world population)
^Prevalence data and 95 % confidence intervals adjusting for sampling weights
*Prevalence data and 95 % confidence intervals are age- and sex-standardized to the regional population. (Regional projected population by World Bank, 2010:
India's population for Delhi and Chennai and Pakistan's population for Karachi)
Prevalence data and 95 % confidence intervals are age- and sex-standardized to the worlds population
Estimates likely unreliable due to small number of subjects in this category
tobacco use was 34.3 (SD = 1.5), 32.7 (SD = 1.1), and 28.00
(SD = 0.9) in Chennai, Delhi, and Karachi, respectively.
Data averaged across all three cities, shows that the total
mean age at initiation was lower in males than females
for both smoked and smokeless tobacco use (M = 25.0,
SD = 0.3 vs. M = 28.8, SD = 1.1, respectively for smoked
tobacco and M = 28.2, SD = 0.6 vs. M = 30.9, SD = 0.7,
respectively for smokeless tobacco). Among both males and
females, age of initiation of tobacco smoking versus
smokeless tobacco use was younger except in Chennai,
where the mean age of initiation of smokeless tobacco use
was lower than that for the smoked form among females
(M = 34.3, SD = 1.5 vs. M = 35, SD = 0.1, respectively).
Table 2 presents data of different types of tobacco use
among males and females. Males in Chennai most
commonly reported use of cigarettes in Chennai (22.7 %)
and Karachi (20.8 %) and beedis (15.5 %) in Delhi.
Females most commonly reported use of chewed tobacco
in Chennai (3.1 %) and Dehli (2.5 %) and pan with zarda
in Karachi (11.0 %). Table 2 also shows that mean
salivary cotinine values (ng/mL) were significantly higher in
current tobacco users as compared to participants who
reported no current tobacco use. Interestingly, there
were no differences in cotinine levels in relation to SHS
exposure versus no exposure among participants who
reported no tobacco use.
Bivariate correlates of male tobacco use included: age
(p < 0.001); lower education qualifications (p < 0.001);
occupation (p < 0.001); lower income (p < 0.001); lower
BMI (p < 0.001); current alcohol use (p < 0.001); and
status of diabetes (p < 0.001), hypertension (p < 0.001), and
dyslipidemia (p = 0.001) (Table 3). City of residence was
not significantly associated with tobacco usage (p = 0.454).
Factors associated with tobacco use in females were
similar to that in males except that tobacco use was more
prevalent in older age groups and a lack of an association
with dyslipidemia status. Also, among females, city of
residence was a significant determinant of tobacco usage
(p = 0.001), with the highest prevalence seen in Karachi.
In the multivariate logistic regression analyses (Table 4),
correlates of tobacco use among both males and females
included residing in Karachi or Delhi versus Chennai;
older age; no formal or up to primary education; earning
less income; and lower BMI. Additional correlates among
males were being a semi-skilled laborer versus not and
current alcohol use, whereas additional correlates among
females included professionals versus not working and
lower odds of having a known diagnosis of dyslipidemia.
(Note that we explored including BMI and income as
continuous variables and found similar results.) Multiple
imputation was done for missing data, and the model was
rerun, demonstrating similar results.
We found a high prevalence of tobacco use among
males in Chennai, Delhi, and Karachi, while the
prevalence of tobacco use among females is much lower.
Specifically, lifetime tobacco use ranged from 41.3 % in
Delhi to 45.0 % in Chennai among males and from 7.6 %
in Chennai to 19.7 % in Karachi among females. Past six
month use ranged from 36.1 % in Delhi to 39.1 % in
Karachi among males and from 7.1 % in Delhi to 18.6 %
Table 3 Bivariate analyses examining tobacco use among males and females in Chennai, Delhi, and Karachi (Continued)
Newly diagnosed dyslipidemia
Note: P-values yielded by Pearson Chi-square test
in Karachi among females. Our estimates are in line with
previous studies [10,11,33,34]. However, compared to
our estimates in Delhi and Chennai, the Global Adult
Tobacco Survey (GATS) found a higher current tobacco
use prevalence in India, specifically 34.6 % (47.9 %
among males, 20.3 % among females) . Additionally,
our estimates of current tobacco use among males and
females in Karachi were lower than previously
documented in Pakistan . These differences in estimates
may be related to the fact that rural males (not included
in this study) have been shown to have higher tobacco
use [10,12,13] or to variation in measures used. For
example, many studies including the GATS assess current
tobacco use by asking if participants currently use
tobacco daily, less than daily, or not at all without stating
a time frame, whereas the current study asked
participants to report whether they had used these products in
the past six months, which is more specific. To put these
data in a global context, a 2012 study of 16 countries
participating in GATS  found that 48.6 % of men
and 11.3 % of women were tobacco users, with 40.7 % of
men and 5.0 % of women using a combustible tobacco
product . As such, the tobacco use prevalence in
these three cities is lower for males than in the countries
included in the GATS but more similar to the females
included in the GATS surveys .
The current study also documented a wide range of
tobacco products used in this population, which is
similar to findings in prior research . Among males, use
of smoked tobacco was highest in Chennai and lowest in
Karachi, whereas chew tobacco rates were highest in
Karachi and lowest in Chennai. Use of smoked or chewed
tobacco among females was highest in Karachi. The most
common tobacco product used in Chennai and Karachi
was cigarettes, whereas beedis were most commonly used
in Delhi. Other common tobacco products included
tobacco chew, particularly in Chennai and Delhi, and Pan
with Zarda, particularly in Karachi. The reasons for the
differences in tobacco products used in these cities need
to be explored.
In multivariable regression, factors associated with
tobacco use among each males and females included
residing in Karachi or Delhi versus Chennai. The reasons
for these city differences are unclear; however, these
findings may be partially attributed to lower price of
tobacco products in Pakistan in comparison with India
. An additional finding indicated that, while Karachi
had the highest tobacco use prevalence among the three
cities for both men and women, lifetime tobacco use
among males across the three cities were quite similar,
which may suggest lower tobacco prices in Pakistan
encouraging experimentation with tobacco. In addition,
older age, lower education, earning less income, and
lower BMI were correlates of tobacco use among both
males and females, which aligns with prior research
[10,12,13]. Another correlate of tobacco use among males
was alcohol use. This is in line with well-established
research in other countries documenting the connection
between other substance use, particularly alcohol use, and
smoking, both in this region and outside this region
[10,12,13,37,38]. For females, another correlate of tobacco
use was lower odds of having newly diagnosed
dyslipidemia, which has not been documented previously.
The sociodemographic factors associated with tobacco
use among the genders reflected one very important
difference males who were semi-skilled laborers versus
not working were at greater risk for being a tobacco
user, whereas females who were professionals versus not
working were at greater risk for use. This may reflect the
global trend of the tobacco industry targeting females in
developing countries to increase their total market,
particularly by targeting females who are educated and in
urban settings .
An important strength of the current study was
validation of self-reported tobacco use by estimating salivary
cotinine level in a random subsample. As expected, the
mean cotinine levels were higher among current tobacco
users compared to nonusers. However, average cotinine
levels were higher among both users and nonusers than
shown in prior research [42,43]. For example, a 2000
study of 222 tobacco users and 97 nonusers found that
mean salivary cotinine was 166 ng/ml in tobacco users
and 6.3 ng/ml in nonusers. This study indicated an
optimal cut-off to discriminate users from nonusers is
between 7 and 13 ng/ml. The nonusers in the CARRS
sample had higher cotinine levels than this cut-off point
(range 18.8 to 41.6). This prior study indicated that
smoking status of significant others was associated with
higher cotinine levels among nonusers ; perhaps the
high prevalence of tobacco use in these three cities
implies high levels of SHS exposure impacting the cotinine
levels of nonusers. However, another interesting finding
Table 4 Multivariate regression predicting tobacco use among males and females in Chennai, Delhi, and Karachi (Continued)
Known diagnosis dyslipidemia
Newly diagnosed dyslipidemia
Adjusted Wald test for all parameters
F(24,69) = 24.10, p < 0.0001
F(24,68) = 10.71, p < 0.0001
from our study is that cotinine levels were not different
between nonusers exposed to SHS versus not exposed,
which warrants further examination.
The current study has important implications for
research and practice. In terms of research, this study
suggests the need for more longitudinal research regarding
correlates of tobacco use among individuals in India and
Pakistan, given the relatively limited scope of factors
included in this data set. There is specific interest in the
relationship between tobacco use and comorbid
conditions like hypertension, diabetes, and dyslipidemia which
cannot be adequately explored in a cross-sectional
analysis like ours. We will have the opportunity to explore
this as our cohort study matures. In addition, the social
norms, tobacco control policies, and potential exposure
to tobacco marketing should be examined in these
differing contexts to determine the impact of these
sociocontextual factors that impact tobacco use initiation and
maintenance among males and females. Regarding
practice, policies involved in the Framework Convention on
Tobacco Control, particularly those impacting the social
norms of tobacco use (e.g., public smoke-free policies,
regulation of tobacco advertising) and systems to aid in
cessation, must be supported in order to influence
tobacco use initiation and maintenance among this
population. In terms of practice implications, it appears from
our analysis that people with known risks for heart
disease and stroke (e.g., diabetes, hypertension) are less
likely to smoke. Again, more longitudinal exploration of
these relationships will be helpful, but regardless,
practitioners must continue to address tobacco use in the
clinical setting, particularly among patients with medical
A limitation of the CARRS model is that the study setting
is urban and does not include the larger rural population;
this sample also is limited due to an underrepresentation
of the younger age groups [20,44]. Second, because of the
cross-sectional nature of this study, we cannot determine
the directionality of the relationships documented.
However, the longitudinal nature of the CARRS study will
allow us to address this limitation in future research.
Moreover, the multivariate model allows us to determine
the amount of variance in tobacco use accounted for by
the other factors assessed in this study. Another limitation
is that, because the primary aim of this study was not to
explore all dimensions of tobacco use, several important
factors potentially related to tobacco use (e.g., social
norms, exposure to marketing) were not assessed.
Additionally, our time-frame for tobacco use (i.e., past
6 month use) is somewhat unconventional but was used
to capture the relatively low overall use of tobacco
products among women in this region. This has implications
for comparability to findings from other studies using
other assessment approaches and other time frames.
Finally, due to the multiple tobacco products assessed and
the variability in number of products used, frequency of
use, and variability in nicotine content across tobacco
products, our current analysis of cotinine levels was
mainly aimed at examining cotinine levels among current
tobacco users versus nonusers, which confirmed
differences between users and nonusers.
Tobacco use prevalence is high, particularly among men,
in Chennai, Delhi, and Karachi. Moreover, there is a
broad range of tobacco products being used and
differences in use prevalence of these products within these
specific cities. Thus, tobacco control policy
implementation is critical to address tobacco-related morbidity and
mortality. Future research should examine psychosocial
and contextual factors influencing tobacco use among
those living in India and Pakistan. Specifically, factors
impacting differential prevalence of tobacco use among
males and females, the social norms of tobacco and
other substance use, and the impact of health problems
on cessation should be examined further. In addition,
interventions and policies that might impact attitudes
toward tobacco and social norms regarding tobacco use
should be investigated and considered.
Additional file 1: Table S1. World and regional population versus
BMI: Body mass index; CARRS: Center for Cardiometabolic Risk Reduction
in South Asia; CEB: Census enumeration blocks; CI: Confidence interval;
CMD: Cardiometabolic diseases; CPD: Cigarettes per day; CVD: Cardiovascular
disease; DBP: Diastolic blood pressure; FBG: Fasting blood glucose;
FCTC: World Health Organization Framework Convention on Tobacco
Control; HINTS: Health Information National Trends Study; HIV/AIDS: Human
Immunodeficiency Virus/Acquired Immune Deficiency Syndrome; INR: Indian
Rupees; LDL: Lowdensity lipoprotein cholesterol; M: Mean; PSU: Primary
sampling units; PKR: Pakistani Rupees; SBP: Systolic blood pressure;
SD: Standard deviation; SHS: Secondhand smoke; TC: Total cholesterol;
USD: United States Dollars.
The authors declare that they have no competing interests.
CJB conceptualized the analyses and wrote the manuscript. VSA conceptualized
the study, contributed to study operation, is part of the coordinating center,
and aided in manuscript development. MKA conceptualized the study,
contributed to study operation, contributed to measure development, and
aided in manuscript development. DK contributed to study operation, lead data
analysis, and manuscript development. HMK contributed to study operation,
data collection, and manuscript development. RS contributed to study
operation and manuscript development. RP contributed to measure
development and aided in manuscript development. DM contributed to study
operation and manuscript development. ZF contributed to study operation and
manuscript development. MMK conceptualized the study, contributed to study
operation, and aided in manuscript development. NT conceptualized the study,
contributed to study operation, is part of the coordinating center, contributed
to measure development, and aided in manuscript development. MV
contributed to study operation and manuscript development. KMVN
contributed to study conceptualization, study operation, data collection, and
manuscript development. DP conceptualized the study, contributed to study
operation, is part of the coordinating center, contributed to measure
development, and aided in manuscript development. All authors read and
approved the final manuscript.
This study is coordinated by Centre of Excellence Centre for Cardiometabolic
Risk Reduction in South Asia (CoECARRS) based at Public Health Foundation of
India (PHFI), New Delhi, India in collaboration with Centre for Chronic Disease
Control (CCDC), New Delhi, Emory University, Atlanta, U.S.A, All India Institute of
Medical Sciences (AIIMS), New Delhi, Madras Diabetes Research Foundation
(MDRF), Chennai, India and Aga Khan University, Karachi, Pakistan. We hereby,
acknowledge the contributions of the field and research staff of the CARRS
Surveillance Investigators Group (a list of all members is included below).
This project is funded in whole or in part by the National Heart, Lung, and
Blood Institute, National Institutes of Health (NIH), Department of Health and
Human Services, under Contract No. HHSN268200900026C, and the United
Health Group, Minneapolis, Mn, USA.
Several members of the research team at PHFI, Emory University, and CCDC
were/are supported by the Fogarty International Clinical Research Scholars
Fellows programme (FICRSF) through Grant Number 5R24TW007988 from
NIH, Fogarty International Center (FIC) through Vanderbilt University, Emorys
Global Health Institute, and D43 NCDs in India Training Program through
Award Number 1D43HD05249 from the Eunice Kennedy Shriver National
Institute Of Child Health & Human Development (NICHD) and FIC. However,
the contents of this paper are solely the responsibility of the writing group
(listed below) and do not necessarily represent the official views of FIC,
Vanderbilt University, Emory University, PHFI, NICHD, or the NIH.
COECARRS Surveillance Investigators Group.
Steering Committee: Dorairaj Prabhakaran, K. M. Venkat Narayan, K Srinath
Reddy, Nikhil Tandon, V. Mohan, Muhammed M. Kadir, Mohammed K. Ali,
Vamadevan S Ajay.
Operations: Dorairaj Prabhakaran, Nikhil Tandon, K. M. Venkat Narayan,
Mohammed K Ali, Muhammed M. Kadir, S. Roopa, Hassan M. Khan, R. Anjana RM,
Pradeepa, M. Deepa, Vamadevan S Ajay, Dimple Kondal, Ruby Gupta, Pragya
Coordinating Centre (Delhi): Dorairaj Prabhakaran, Nikhil Tandon, S. Roopa,
Vamadevan S Ajay, Manisha Nair, Nivedita Devasenapathy, Divya Pillai
Development of questionnaires and manual of operations: Dorairaj
Prabhakaran, Nikhil Tandon, K. M. Venkat Nararayan, Mohammed K. Ali,
Manisha Nair, Nivedita Devasenapathy, R. Pradeepa , Ed Gregg, Anwar
Merchant, Romaina Iqbal.
Data management and statistical team: Dimple Kondal, Shivam Pandey,
Laboratory: Lakshmy Ramakrishnan, Ruby Gupta, Savita.
Information Technology: Ramanathan K, Ansel J DCruz, Gnanashekaran K.
Online data entry software: Mahesh Dorairaj.
Data collection teams:
Chennai: Field supervisor: Rahul T; Field interviewers: Alagarsamy, Anthony
JV, Arul Dass.A, Arul Pitchai.S, Ashok Kumar, Balaji V, Dhanasekar L, KalaiVani
D, Kumar M, Nandhakumar, Prathiban K, Sampath, Saravana Kumar P,
SaravananR, Senthil RajaR, ShenbagaValliE, SivamanikandanK, SureshT, Uma
Sankari G; Laboratory assistants: Geetha Priya L, Gowri, Irin Jayakumari A,
Padmapriya, Ramakrishnan R, Revathy, Satish Raj S, Sudha M, Suresh, Vijay
Baskar S; Data entry operators: Narayanan, Nirmala.
Delhi: Field supervisor: Liladhar Dorlikar; Field interviewers: Parag Jyoti Das,
Kulwant Kaur, Sweta Kumari, Meena Thakur, Garima Rautela, Avijeet Malik,
Anita Yadav, Makhan, Rishi Garg, Arun; Laboratory assistants: Priyanka
Nautiyal, Sunil Dogra, Geetha; Data entry operators: Naveen Kaushik, Avnish.
Karachi: Field supervisor: Mehboob John Samuel; Field interviewers &
laboratory assistants: Yousuf Sadiq, Shukrat Khan, Shahirah Ziarat Khan,
Nadia Khan, Noureen Khan, Naseem Sehar, Asif Shabaz, Fakhrah Perveen,
Karan Inayat, Tajir Hussain, Tariq Hussain, Nasreen Khan; Data entry
operator: Sayed Arif Hussain Kazmi.
1. Mendez D , Alshanqeety O , Warner KE . The potential impact of smoking control policies on future global smoking trends . Tob Control . 2013 ; 22 ( 1 ): 46 - 51 .
2. Lozano R , Murray CJL , Lopez A. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the global burden of disease study 2010 . Lancet. 2012 ; 380 ( 9859 ): 2095 - 128 .
3. Sempos CT , Durazo-Arvizu R , McGee DL , Cooper RS , Prewitt TE . The influence of cigarette smoking on the association between body weight and mortality. The Framingham heart study revisited . Ann Epidemiol . 1998 ; 8 ( 5 ): 289 - 300 .
4. U.S. Department of Health and Human Services . The Health Consequences of Smoking-50 Years of Progress: A Report of the Surgeon General . 2014 , U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health: Atlanta , GA.
5. Multiple Risk Factor Intervention Trial Research Group . Multiple risk factor intervention trial . Risk factor changes and mortality results . JAMA . 1982 ; 248 ( 12 ): 1465 - 77 .
6. Browner WS , Du Chene AG , Hulley SB . Effects of the multiple risk factor intervention trial smoking cessation program on pulmonary function. A randomized controlled trial . West J Med . 1992 ; 157 ( 5 ): 534 - 8 .
7. Ng M , Freeman MK , Fleming TD , Robinson M , Dwyer-Lindgren L , Thomson B , et al. Smoking prevalence and cigarette consumption in 187 countries , 1980 - 2012 . JAMA. 2014 ; 311 ( 2 ): 183 - 92 .
8. World Health Organization . WHO Report on the Global Tobacco Epidemic , 2011 . 2011 , World Health Organization: Geneva , Switzerland
9. Sreeramareddy CT , Pradhan PMS , Mir IA , Sin S. Smoking and smokeless tobacco use in nine south and southeast Asian countries: prevalence estimates and social determinants from demographic and health surveys . Popul Health Metr . 2014 ; 12 : 22 .
10. Chockalingam K , Vedhachalam C , Rangasamy S , Sekar G , Adinarayanan S , Swaminathan S , et al. Prevalence of tobacco use in urban, semi urban and rural areas in and around Chennai city . India PLoS One . 2013 ; 8 ( 10 ), e76005.
11. National Institute of Population Studies (NIPS) [Pakistan] and ICF International. Pakistan Demographic and Health Survey 2012 - 2013 . 2013 , National Institute of Population Studies; ICF International: Islamabad , Pakistan; Calverton, Maryland, USA
12. Asthma Epidemiology Study Group. Tobacco smoking in India: prevalence, quit-rates and respiratory morbidity . Indian J Chest Dis Allied Sci . 2006 ; 48 ( 1 ): 37 - 42 .
13. Ahmad K , Jafary F , Jehan I , Hatcher J , Khan AQ , Chaturvedi N , et al. Prevalence and predictors of smoking in Pakistan: results of the National Health Survey of Pakistan . Eur J Cardiovasc Prev Rehabil . 2005 ; 12 ( 3 ): 203 - 8 .
14. Wilkinson P , Sayer J , Laji K , Grundy C , Marchant B , Kopelman P , et al. Comparison of case fatality in south Asian and white patients after acute myocardial infarction: observational study . BMJ . 1996 ; 312 ( 7042 ): 1330 - 3 .
15. UK Prospective Diabetes Study Group. ethnicity and cardiovascular disease . The incidence of myocardial infarction in white, South Asian, and AfroCaribbean patients with type II diabetes . Diabetes Care . 1998 ; 21 : 1271 - 7 .
16. Anand SS , Yusuf S , Vuksan V , Devanesen S , Teo KK , Montague PA , et al. Differences in risk factors, atherosclerosis and cardiovascular disease between ethnic groups in Canada: the study of health assessment and risk in ethnic groups (SHARE) . Indian Heart J . 2000 ; 52 ( 7 Suppl): S35 - 43 .
17. Pais P , Pogue J , Gerstein H , Zachariah E , Savitha D , Jayprakash S , et al. Risk factors for acute myocardial infarction in Indians: a case-control study . Lancet . 1996 ; 348 ( 9024 ): 358 - 63 .
18. Nair M , Ali MK , Ajay VS , Shivashankar R , Mohan V , Pradeepa R , et al. CARRS surveillance study: design and methods to assess burdens from multiple perspectives . BMC Public Health . 2012 ; 12 : 701 .
19. World Health Organization. STEPwise approach to surveillance (STEPS) . 2014 ; Available from: http://www.who.int/chp/steps/en/.
20. Office of the Registrar General & Census Commissioner India . Census of India 2011 . 2011 , Ministry of Home Affairs, Government of India. Available from: http://censusindia.gov.in/: New Delhi.
21. Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat . World Population Prospects: The 2006 Revision and World Urbanization Prospects: The 2007 Revision. 2007 , Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat . Available from: http://www.un.org/en/ development/desa/population/.
22. Rizzo L , Brick J , Park I. A minimally intrusive method for sampling persons in Random Digit Dial surveys . Public Opin Q . 2004 ; 68 ( 2 ): 267 - 74 .
23. World Health Organization, STEPwise approach to surveillance (STEPS). 2008, World Health Organization . Available from: http://www.who.int/chp/steps/ manual/en/index.html.
24. Klesges RC , Meyers AW , Klesges LM , La Vasque ME. Smoking, body weight, and their effects on smoking behavior: a comprehensive review of the literature . Psychol Bull . 1989 ; 106 ( 2 ): 204 - 30 .
25. Munafo MR , Tilling K , Ben-Shlomo Y. Smoking status and body mass index: a longitudinal study . Nicotine Tob Res . 2009 ; 11 ( 6 ): 765 - 71 .
26. Bobo JK , Husten C. Sociocultural influences on smoking and drinking . Alcohol Res Health . 2000 ; 24 ( 4 ): 225 - 32 .
27. Istvan J , Matarazzo JD . Tobacco, alcohol, and caffeine use: a review of their interrelationships . Psychol Bull . 1984 ; 95 ( 2 ): 301 - 26 .
28. Patra J , Jha P , Rehm J , Suraweera W. Tobacco smoking, alcohol drinking, diabetes, low body mass index and the risk of self-reported symptoms of active tuberculosis: individual participant data (IPD) meta-analyses of 72,684 individuals in 14 high tuberculosis burden countries . PLoS One . 2014 ; 9 ( 5 ), e96433.
29. Binnie V , McHugh S , Macpherson L , Borland B , Moir K , Malik K. The validation of self-reported smoking status by analysing cotinine levels in stimulated and unstimulated saliva, serum and urine . Oral Dis . 2004 ; 10 ( 5 ): 287 - 93 .
30. Connor Gorber S , Schofield-Hurwitz S , Hardt J , Levasseur G , Tremblay M. The accuracy of self-reported smoking: a systematic review of the relationship between self-reported and cotinine-assessed smoking status . Nicotine Tob Res . 2009 ; 11 ( 1 ): 12 - 24 .
31. Alterman AI , Gariti P , Niedbala RS . Varying results for immunoassay screening kits for cotinine level . Psychol Addict Behav . 2002 ; 16 ( 3 ): 256 - 9 .
32. Lehtonen R , Pahkinen E. Practical Methods for Design and Analysis of Complex Surveys . 2nd ed . West Sussex, England: John Wiley & Sons, LTD.; 2003 .
33. Gilani SI , Leon DA . Prevalence and sociodemographic determinants of tobacco use among adults in Pakistan: findings of a nationwide survey conducted in 2012 . Popul Health Metr . 2013 ; 11 ( 1 ): 16 .
34. International Institute for Population Sciences. Global Adult Tobacco Survey: Fact Sheet: India , 2009 - 2010 . 2010 , International Institute for Population Sciences: Mumbai, India.
35. Giovino GA , Mirza SA , Samet JM , Gupta PC , Jarvis MJ , Bhala N , et al. Tobacco use in 3 billion individuals from 16 countries: an analysis of nationally representative cross-sectional household surveys . Lancet . 2012 ; 380 ( 9842 ): 668 - 79 .
36. Blecher E , van Walbeek C. An Analysis of Cigarette Affordability , in International Union Against Tuberculosis and Lung Disease . 2008 : Paris, France.
37. Pinsker EA , Berg CJ , Nehl EJ , Prokhorov AV , Buchanan TS , Ahluwalia JS . Intentions to quit smoking among daily smokers and native and converted nondaily college student smokers . Health Ed Res . 2013 ; 28 ( 2 ): 313 - 25 .
38. Sutfin EL , McCoy TP , Berg CJ , Champion H , Helme DW , O'Brien MC , et al. Tobacco use among college students: a comparison of daily and nondaily smokers . Am J Health Behav . 2012 ; 36 ( 2 ): 218 - 29 .
39. Mackay J , Crofton J. Tobacco and the developing world . Br Med Bull . 1996 ; 52 ( 1 ): 206 - 21 .
40. Pampel FC , Denney JT . Cross-national sources of health inequality: education and tobacco use in the world health survey . Demography . 2011 ; 48 ( 2 ): 653 - 74 .
41. Sharma V , Kerr SH , Kawar Z , Kerr DJ . Challenges of cancer control in developing countries: current status and future perspective . Future Oncol . 2011 ; 7 ( 10 ): 1213 - 22 .
42. Etter JF , Vu Duc T , Perneger TV . Saliva cotinine levels in smokers and nonsmokers . Am J Epidemiol . 2000 ; 151 ( 3 ): 251 - 8 .
43. Wall MA , Johnson J , Jacob P , Benowitz NL . Cotinine in the serum, saliva, and urine of nonsmokers, passive smokers, and active smokers . Am J Public Health . 1988 ; 78 ( 6 ): 699 - 701 .
44. National Institute of Population Studies (NIPS) [Pakistan] aMII. Pakistan Demographic and Health Survey 2006 - 07 . 2008 : Islamabad, Pakistan: National Institute of Population Studies and Macro International Inc.