Atherosclerosis 234 (2014) 454e460

Contents lists available at ScienceDirect

Atherosclerosis
journal homepage: www.elsevier.com/locate/atherosclerosis

Lifetime cumulative exposure to waterpipe smoking is associated with

coronary artery disease
Abla M. Sibai a, Rania A. Tohme a, b, Mohamad M. Almedawar c, d, Taha Itani e,
Sara I. Yassine a, Eden A. Nohra f, Hussain A. Isma’eel c, d, *
a
Department of Epidemiology & Population Health, Faculty of Health Sciences, American University of Beirut, Beirut, Lebanon
b
Centers for Disease Control and Prevention, Atlanta, GA 30333, USA
c
Department of Internal Medicine, American University of Beirut Medical Center, Beirut, Lebanon
d
Vascular Medicine Program, American University of Beirut Medical Center, Beirut, Lebanon
e
Department of Public Health Medicine, School of Public Health, University of Bielefeld, Bielefeld, Germany
f
Department of Surgery, American University of Beirut Medical Center, Beirut, Lebanon

a r t i c l e i n f o a b s t r a c t

Article history: Objective: Globally, waterpipe (WP) smoking is becoming a more prevalent form of tobacco consump-
Received 22 November 2013 tion. Whilst research so far has demonstrated a significant link between WP use and a number of health
Received in revised form outcomes, little is known of its association with heart disease. We examine in this study the association
20 March 2014
of WP smoking with angiographically confirmed coronary artery disease (CAD).
Accepted 31 March 2014
Available online 15 April 2014
Methods: A total of 1210 patients, aged 40 years and over and free from smoking-associated illnesses or
history of cardiovascular procedures, admitted for coronary angiography at four major hospitals in
Lebanon, were included. The extent of CAD was summarized in two ways, firstly as diseased (
50% and
Keywords:
Waterpipe smoking

70% occlusion in at least one main coronary artery) versus non-diseased (entirely normal coronaries),
Coronary artery disease and secondly, as CAD cumulative score based on Duke CAD Prognostic Index. A score of WP-years,
Heart disease capturing intensity and lifetime duration of exposure, was estimated for each individual.
Caseecontrol study Results: Lifetime exposure exceeding 40 WP-years was associated with a threefold significant increase in
the odds of having severe stenosis (
70%) compared to non-smokers (OR ¼ 2.94, 95% CI 1.04e8.33) as
well as with the CAD Index (b ¼ 7.835, p-value ¼ 0.027), net of the effect of socio-demographic char-
acteristics, health behaviors and co-morbidity. A doseeresponse relationship between WP-years and
percent stenosis was also established. WP smoking status (never, past and current) did not associate with
CAD.
Conclusions: Cumulative exposure to WP smoking is significantly associated with severe CAD. There is a
need to monitor WP use among cardiac patients and include this information in their medical charts in
the same manner cigarettes smoking is documented. This is likely to increase awareness of the hazards of
WP smoking and prompt physicians to target WP tobacco control by providing advice to their patients on
WP smoking cessation.
Ó 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction contributing to an annual mortality rate that exceeds six million

More than 3 billion people worldwide currently smoke tobacco
in its various forms, and the majority (80%) live in low and middle-
income countries. Tobacco use has already been established as the
number one modifiable risk factor for cardiovascular diseases,
* Corresponding author. American University of Beirut Medical Center, Cairo
Street, P.O.Box: 11-0236, Riad El Solh, Beirut 1107-2020, Lebanon. Tel.: þ961 1
350000x5785; fax: þ961 1 370814.
E-mail address: hi09@aub.edu.lb (H.A. Isma’eel).
individuals [1]. Whist smoking tobacco using waterpipe (WP) has
been most commonly associated with the Eastern Mediterranean
Region, it is increasingly becoming a popular method of tobacco
consumption worldwide [2]. Data from several national and
regional studies suggest that the epidemic of WP smoking is on the
rise reaching a prevalence of 20e30% among adults and increasing
significantly among younger cohorts [3]. With the accumulated
evidence of the epidemiology of this behavior and its health con-
sequences, the American Lung Association addresses WP smoking
as a ‘growing threat to public health’ and an ‘emerging deadly
trend’ [4].

http://dx.doi.org/10.1016/j.atherosclerosis.2014.03.036
0021-9150/Ó 2014 Elsevier Ireland Ltd. All rights reserved.
 A.M. Sibai et al. / Atherosclerosis 234 (2014) 454e460
Research has so far demonstrated a significant link between WP
smoking and a number of diseases including lung, oral and bladder
cancers, respiratory illnesses and adverse pregnancy outcomes [5].
Some studies have also shown associations with surrogate markers
of coronary disease such as heart rate variability [6], heart rate and
systolic blood pressure [7]. Though anticipated to share the coro-
nary disease hazards of cigarettes smoking [8], epidemiological
studies addressing the long-term effect of WP smoking on heart
disease remains largely lacking [5]. We examine in this study the
association of WP smoking with heart disease as manifested by
angiographically determined coronary artery disease (CAD).
2. Materials and methods
2.1. Study population
The study sample included consecutive admissions of patients
for coronary angiography at four major tertiary hospitals in Beirut,
the capital city of Lebanon, and its suburbs, during the period
extending from July 2007 until August 2008. Patients were
excluded from the study if they: 1) were younger than 40 years old;
2) had undergone a previous coronary angiography, coronary artery
bypass surgery, or percutaneous coronary intervention; 3) had a
history of heart disease (myocardial infarction, angina, valvular
heart disease, congenital heart disease), peripheral vascular dis-
ease, or stroke; 4) had chronic or obstructive pulmonary disease; or
5) any type of cancer associated with smoking (lung, gum, mouth,
throat, pancreas, bladder and cervix).
During the study period, a total of 2525 consecutive patients
were admitted for cardiac catheterization. Of these, 625 were
admitted on an out-patient basis and could not be contacted, 632
did not meet the inclusion criteria, 32 refused to participate in the
study, and 26 had missing data on the main exposure variables,
yielding a total of 1210 study subjects with complete interviews.
2.2. Outcome measure and classification
Extent of CAD, our outcome variable, was based on percent
maximal stenosis at any location (ostium, proximal, mid and distal)
in the coronary vessels (left main, left anterior descending, first and
second diagonals, left circumflex, first, second, and third obtuse
marginals, right coronary artery, and posterior descending artery).
In vessels that had more than one stenosis, the most severe stenosis
was recorded. For this study, the extent of CAD was summarized
and examined in two ways: firstly, as diseased versus non-diseased
(dichotomous outcome) using the caseecontrol study design and
corresponding method of analyses, and secondly, as detailed
characterization of the extent of coronary disease severity
(continuous outcome) using the cross-section approach and cor-
responding method of analyses.
In the caseecontrol design, subjects with mild stenosis (1e49%,
n ¼ 247) were excluded from the analysis. This was made for a
precise separation between diseased and non-diseased subjects
and for better delineation of risk factors for angiographically
defined coronary atherosclerosis [9]. Hence, persons with entirely
normal coronaries (0% stenosis, n ¼ 382) were considered as con-
trols, and following the American College of Cardiology/American
Heart Association guidelines [10], we defined two groups of cases:
patients with
50% stenosis in any coronary artery (Group I cases,
n ¼ 581) and patients with
70% stenosis in any coronary artery
(Group II cases, n ¼ 327).
For the second classification of the outcome, the atherosclerotic
burden of CAD was summarized for the total sample (n ¼ 1210) as a
continuous measure based on the Duke Coronary Artery Disease
Prognostic Index [11]. This CAD Index takes into consideration not
455
only the percent of coronary lesions but also the number of
diseased vessels and the site, thus providing a weight for various
combinations. For example, a one vessel disease with 75% stenosis
was given a score of 23, while a three-vessel disease with
95% in
at least one vessel was given a score of 63, and a left main disease of
75% stenosis was given a score of 82. This yielded a stenosis Index
for each patient ranging from 0 (completely normal arteries) to
100% (left main
95%).
2.3. Interviews and measures of exposure
Interviewers were recruited at each hospital for the face-to-face
interviews using a structured questionnaire. In order to avoid
reporting bias, patients were interviewed either prior to the per-
formance of the catheterization (37%) or prior to their knowledge of
the results of their cardiac catheterization (57%).
The questionnaire items on WP smoking habits were adapted
with minor modifications from Maziak and colleagues standardized
interview schedule [12] eliciting information on WP smoking status
(never, past and current), duration (age at initiation and age at
stopping for the past smokers), frequency (number of days of
smoking per week) and intensity (average number of WPs
consumed on each occasion). Consequently, ‘waterpipe-years’ (WP-
years) were estimated for each individual by multiplying the average
number of WPs smoked per day by the number of years they have
been smoking. This yielded a continuous score ranging from 0 for the
never WP smokers to a maximum of 164 WP-years. Thus, for a
person scoring 20 WP-years for example, this meant that the subject
had smoked an average of two WPs per day for a total of 10 years or
one WP per day for 20 years, and so on. The score, capturing lifetime
WP smoking exposure, was then grouped into four categories (0, 1e
20, 21e40 and
41 WP-years). Information on cigarette smoking
was similarly reported and assessed according to status (never, past
and current) and cigarette pack-year. The pack-year was grouped
into five categories (0, 1e20, 21e40, 41e60, and
61 pack-years).
Data on socio-demographic characteristics (gender, age, edu-
cation, work and marital status) and the classical, behavioral and
clinical risk factors for CAD (physical activity, alcohol consumption,
diabetes, hypertension, hyperlipidemia, and family history of CAD)
were also collected. Physical activity was assessed by including
both habitual work and leisure physical activity [13]. Alcohol
drinking was categorized as “never/rare” (none or <1 glass per
week), “occasional” (1e2 glasses per week), and “frequent” (
3
glasses per week) [14]. Diabetes, hypertension and hyperlipidemia
were considered present if the subject was diagnosed of having any
of the conditions or if they were on a diet and/or were taking
medications for these conditions. Family history of CAD was
restricted to early CAD onset in first degree relatives (<50 years for
males and <55 years for females).
The Institutional Review Board of the American University of
Beirut approved the study protocol which complies with the
Declaration of Helsinki, and participants signed an informed con-
sent prior to the interview. All data were treated in a confidential
manner and the anonymity of respondents was maintained.
2.4. Data analysis
Frequencies and means with standard errors (SE) were used to
describe the sample, and differences in baseline characteristics and
potential confounders across categories of the WP-years were
examined using chi-square test and t-test, as appropriate. Associ-
ations between WP and cigarette smoking with angiographically
determined CAD were analyzed in two ways: considering the
outcome firstly as a dichotomous variable (cases vs controls) and
secondly as continuous (CAD index). Two case control comparisons
456 A.M. Sibai et al. / Atherosclerosis 234 (2014) 454e460

Table 1
Distribution of cases and controls by water pipe and cigarette smoking behavior.

Smoking status Total sample Controls (n ¼ 382) Group I Group II P-valuea P-valueb
(n ¼ 1210) cases
50% cases
70%
(n ¼ 581) (n ¼ 327)

n % n % n % n %

Waterpipe smoking status

Never 975 80.6 307 80.4 463 79.7 257 78.6 0.844 0.580
Past smoker 36 3.0 10 2.6 19 3.3 13 4.0
Current smoker 199 16.4 65 17.0 99 17.0 57 17.4
Waterpipe-years
Never 975 80.6 307 80.4 463 79.7 257 78.6 0.041 0.267
1e20 140 11.6 57 14.9 66 11.4 43 13.1
21e40 49 4.0 10 2.6 25 4.3 15 4.6
41þ 46 3.8 8 2.1 27 4.6 12 3.7
Mean  SE 5.2  0.6 3.3  0.8 5.9  0.9 5.2  1.1 0.026 0.159
Cigarette smoking status
Never 445 36.8 160 41.9 200 34.4 106 32.4 0.064 0.034
Past smoker 221 18.3 66 17.3 114 19.6 66 20.2
Current smoker 544 45.0 156 40.8 267 46.0 155 47.4
Pack years
Never 445 36.8 160 42.0 200 34.7 106 32.9 0.004 0.002
1e20 164 13.6 68 17.8 72 12.5 37 11.5
21e40 231 19.1 64 16.8 123 21.4 77 23.9
41e60 150 12.4 35 9.2 70 12.2 38 11.8
61þ 211 17.4 54 14.2 111 19.3 64 19.9
Mean  SE 30.4  1.1 24.3  1.7 33.2  1.6 34.3  2.2 <0.001 <0.001
a
Comparing Group I vs controls.
b
Comparing Group II vs control.

(controls vs group I:
50% and controls vs group II:
70% stenosis) terms of their WP smoking behavior (status and WP-years), a
were evaluated. Because associations with CAD were more evident significantly higher proportion of current and past cigarette
for WP-years than WP smoking status, the analysis presented smokers were found among cases than controls (p-value ¼ 0.034),
thereafter focused on WP-years. Three multiple logistic regression with cases significantly accumulating more cigarette pack-years (p-
equations and three multiple linear regression equations were
modeled to determine the association of WP-years with CAD. The
first model presented the unadjusted association of WP-years with
CAD, the second model controlled for socio-demographic charac-
teristics (age, gender, education and occupation), and the third
model controlled additionally for the classical CAD risk factors
(cigarette smoking, alcohol consumption, physical activity, dia-
betes, hypertension, hyperlipidemia, and family history of CAD).
Variables entered in these models were either significant at the
bivariate level (p-value <0.2) or were of conceptual importance as
established risk factors. Odds ratios (ORs) and their corresponding
95% confidence intervals (CIs) were calculated. The data were
initially entered on CSPro 3.3, and then imported to SPSS (version
17.0 Chicago, IL) and STATA (STATA MP, release 14) for analysis. A p-
value <0.05 was considered significant.

3. Results

Of the 1210 study subjects, 235 (19.4%) were ever WP smokers,

765 (63.3%) were ever cigarette smokers, 894 (73.9%) ever smoked
either of the two forms of tobacco, and 57(4.7%) were current
smokers of both types of tobacco. Also, 140 (11.6%) accumulated a
lifetime WP smoking between 1 and 20 WP-years, 49 (4.0%) be-
tween 21 and 40 WP-years, and 46 (3.8%) more than 40 WP-years.
Further details on smoking habits by caseecontrol status and CAD
score are presented in Table 1. Compared to the controls, group I
cases with
50% stenosis did not vary significantly with respect to
their WP or cigarette smoking status (never, past, current); how-
ever, cases and controls differed significantly by their lifetime
exposure to tobacco use with cases accumulating more WP-years
and more cigarette pack-years than controls (5.9  0.9 vs.
3.3  0.8, p-value ¼ 0.026; and 33.2.0  1.6 vs. 24.3  1.7, p- Fig. 1. a Extent of atherosclerotic burden as a function of WP smoking status and WP-
value  0.001, respectively). Whilst there was no significant dif- years. b Extent of atherosclerotic burden as a function of cigarette smoking status and
ference between group II cases (
70% stenosis) and controls in pack years.
 A.M. Sibai et al. / Atherosclerosis 234 (2014) 454e460 457

value<0.001). Results for the CAD Index by waterpipe and cigarette
smoking behavior are illustrated graphically in Fig. 1aeb. With the
never smokers treated as the referent category, the only signifi-
cance difference was noted in the case of exposure to
41 WP-years
(mean CAD scores 33.5 vs 22.8, respectively; p-value ¼ 0.003).
The distribution and associations between potential cofounders
and our main exposure, WP-years, are examined in Table 2. Lifetime
WP smoking did not vary significantly by gender. However, higher
WP-years accrued with increasing age. WP-years was significantly
associated with education and work status, with those consuming
more than 40 WP-years being more likely to be illiterate and to
include never workers. In terms of health related characteristics,
WP smoking was significantly more common among the never
cigarette smoking group. Furthermore, physical activity, diabetes,
hypertension and hyperlipidemia were significantly associated
with WP-years. Patients with family history of CAD were more
likely to include WP smokers but this did not reach statistical
significance.
Tables 3 and 4 present the findings of the logistic and linear
regression analyses, respectively. Overall, the magnitude of the
associations was stronger when comparing controls to group II
cases having
70% stenosis than controls to group I cases having

50% stenosis. The odds was strongest and significant for the
comparison between controls and group II cases for the highest
category of exposure to WP-years (
41 WP-years) (OR ¼ 2.94; 95%
CI 1.04e8.33); with findings showing a doseeresponse relation
increasing with increasing exposure to lifetime WP use.
Furthermore, the risk of CAD increased significantly with age and
was higher among males than females. Whereas diabetes and
hyperlipidemia were associated positively with CAD in both sets of
caseecontrol comparisons; alcohol consumption and physical ac-
tivity were negatively associated with both outcomes. Similarly,
lifetime exposure to waterpipe smoking exceeding 40 WP-years
was associated significantly with CAD Index (b ¼ 7.835, p-
value ¼ 0.027). Other significant co-variates included age, gender,
cigarette smoking, alcohol consumption, diabetes, and family his-
tory of CAD.
4. Discussion
To date, this study is the first to examine the relationship be-
tween WP smoking and CAD as manifested by the extent of an-
giographically determined stenosis. The study showed that lifetime
cumulative WP smoking (frequency and duration of risk exposure)
appears to be a more critical risk factor to CAD than the status of
WP smoking (never, ex- and current), with individuals accumu-
lating greater than 40 WP-years being almost three times more at
risk of having severe stenosis (
70%) than nonsmokers, net of the
effect of the traditional CAD risk factors including cigarette smok-
ing. Using the CAD Prognostic Index, the study also established a
significant increase in the extent of CAD with increasing exposure
to WP use. The Index was originally derived as a more detailed
reflection of important prognostic aspects of coronary anatomy
than the traditional classification of one-, two- and three-vessel

Table 2
Distribution of co-variates by categories of exposure to waterpipe years.

Co-variate Waterpipe years

Never 1e20 21e40 41þ P-value

n % n % n % n %

Total 975 80.6 140 11.6 49 4.0 46 3.8

Demographics
Gender (% male) 605 62.1 96 68.6 27 55.1 28 60.9 0.323
Age (years)
40e49 212 21.7 43 30.7 13 26.5 8 17.4 0.023
50e59 279 28.6 46 32.9 9 18.4 10 21.7
60e69 280 28.7 33 23.6 14 28.6 12 26.1

70 204 20.9 18 12.9 13 26.5 16 34.8
Mean  SE 60.3  0.4 56.6  0.9 61.6  1.7 64.1  1.7 <0.001
Education
Illiterate 119 12.1 9 6.4 7 14.3 12 26.1 <0.001
Primary 262 26.9 22 15.7 18 36.7 11 23.9
Secondary 337 34.6 63 45.0 17 34.7 17 37.0
University 257 26.4 46 32.9 7 14.3 6 13.0
Work status
Never 286 29.3 33 23.6 22 44.9 18 39.1 0.003
Current 488 51.1 92 65.7 14 28.6 18 39.1
Retired 201 20.6 15 10.7 13 26.5 10 21.7
Marital status
Single 62 6.4 7 5.1 2 4.1 3 6.7 0.542
Married 860 88.8 129 93.5 44 89.8 41 91.1
Divorced/widowed 47 4.9 2 1.4 3 6.1 1 2.2
Health related characteristics
Cigarette smoking
Never 316 32.4 67 47.9 25 51.0 37 80.4 <0.001
Past 183 18.8 24 17.1 11 22.4 3 6.5
Current 476 48.8 49 35.0 13 26.5 6 13.0
Alcohol
Never/rare 846 86.9 124 88.6 44 89.8 42 91.3 0.831
Occasional 59 6.1 9 6.4 3 6.1 3 6.5
Frequent 68 7.0 7 5.0 2 4.1 1 2.2
Physical activity (%yes) 679 69.8 112 80.0 21 42.9 28 60.9 <0.001
Diabetes (%yes) 266 27.3 48 34.3 20 40.8 21 45.7 0.006
Hypertension (%yes) 484 49.6 76 54.3 34 69.4 22 47.8 0.043
Hyperlipidemia (%yes) 401 41.4 76 54.7 27 55.1 24 53.3 0.004
Family history (%yes) 514 53.4 81 59.1 29 60.4 28 63.6 0.282
458 A.M. Sibai et al. / Atherosclerosis 234 (2014) 454e460

disease, and was used to estimate the mortality hazard for treat-
ment modalities against levels of coronary disease severity [11].
Importantly, the clinical implications of these findings in terms of
how much does WP smoking contribute to increasing CAD-related
deaths or myocardial infarctions, and thereafter how to account for
WP smoking in CAD risk assessment models are compelling ques-
tions to address.
The present study reinforces previous findings of the hazards of
WP smoking [5,15]. The association observed with CAD is expected
given that WP and cigarettes share common toxicants that are
associated with lung and cardiovascular diseases. The elevated risk
of CAD among WP smokers is likely to be related to various in-
flammatory and arrhythmic changes within the heart [6]. WP
smoking has been shown to be associated with reduction in heart
rate variability, which increases the risk of coronary heart disease
and mortality [16], and causes an increase in systemic inflammation
[17] and susceptibility to arrhythmia [18]. Moreover, recent studies
have demonstrated a significant elevation in heart rate and systolic
blood pressure after a single session of WP smoking [7]. In fact, some
studies have linked WP smoking to in vivo oxidation injury and
atherosclerosis in peripheral and coronary vascular tissue [19,20].
While functionally significant obstructive CAD is more present in
patients with increasing severity of stenosis [8], it is important to
note that the majority of acute coronary syndrome (ACS) cases are
due to ruptured plaques with <70% stenosis. The role of WP use in
ACS remains a question that needs to be answered.
WP smoking has been increasingly gaining popularity world-
wide. This popularity is pronounced mainly in the youth lured by a
powerful tobacco industry and the appealing social circumstances
accompanying the behavior [3]. In addition and unlike cigarettes,
the production of WP is not regulated, its content is not monitored,
and in some countries, its use is exempted from banning policies at
restaurants or closed spaces [21]. Adolescents gather in WP cafes
and most often share the same apparatus, a practice similar to
recreational drug use. The attractiveness of WP smoking can be
attributed to several reasons: it is inexpensive, comes in multi
flavors, and it is erroneously perceived to be less harmful and less
addictive than cigarette smoking [22]. In several studies, mis-
conceptions were reported regarding the safety of WP smoking and
a belief that water filters out the toxins from the smoke before
inhalation, with cigarette smokers often switching to WP smoking
when they decide to quit smoking [23,24].
While it is difficult to document the equivalence of each WP
smoker in this study, our finding of potentially more harmful effect
of WP compared to cigarette smoking merits attention. Studies
comparing the harmful constituents of WP versus cigarettes
smoking reveal a wide equivalence range which depends on several
factors including duration of the WP smoking session, the amount
and type of tobacco mounted in one WP and the number of puffs.
Whereas a cigarette can be consumed in 8e12 puffs, a WP may last
50e200 puffs, delivering up to 1 L of smoke in each WP session,
compared to 0.6 L of cigarette smoke [25]. The World Health Or-
ganization released an advisory note in 2005 explaining that one
session of WP smoking may deliver a smoke volume equivalent to
100 cigarettes and Masters and colleagues have shown that a 20-
min, 45-min, or 80-min WP smoking session is equivalent to 25,
60, or 100 cigarettes, respectively [26]. More recently, Cobb et al.
(2011) demonstrated that a single session of WP smoking delivers
“1.7 times the nicotine, 6.5 times the CO, and 46.4 times the tar”
compared to cigarette smoking [27].
The findings of the study need to be discussed in light of its
potential biases and offsetting strengths. The clinical profile of the

Table 3
Logistic regression analysis of both sets of case control studies.

Cases
50% stenosis Cases
70% stenosis

Variable (Referent category)
Model 1a Model 2a Model 3a Model 1a Model 2a Model 3a

OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI

Waterpipe-year (Never)
1e20 0.77 0.52e1.13 0.86 0.57e1.29 0.79 0.51e1.24 0.90 0.59e1.38 1.05 0.66e1.68 0.97 0.58e1.63
21e40 1.66 0.79e3.50 1.76 0.79e3.40 1.62 0.71e3.73 1.79 0.79e4.06 1.83 0.75e4.44 1.68 0.65e4.39
41þ 2.24 1.00e4.99 1.89 0.79e4.36 2.00 0.80e5.04 1.79 0.72e4.45 2.00 0.75e5.32 2.94 1.04e8.33
p-value for trendb 0.031 0.093 0.076 0.137 0.098 0.027
Age (40e49)
50e59 1.73 1.18e2.52 1.45 0.96e2.19 1.88 1.20e2.94 1.51 0.92e2.46
60e69 2.88 1.90e4.36 2.50 1.59e3.94 3.25 2.00e5.28 2.95 1.73e5.05

70 4.33 2.65e7.07 4.47 2.56e7.73 5.31 3.00e9.40 5.23 2.78e9.85
Gender (Females) 3.88 2.51e6.01 4.48 2.79e7.19 6.28 3.49e11.28 8.73 4.55e16.75
Education (Illiterate)
Primary 0.69 0.41e1.17 0.76 0.43e1.33 0.68 0.37e1.25 0.88 0.45e1.74
Complementary 0.60 0.36e0.99 0.71 0.41e1.22 0.68 0.38e1.21 0.99 0.51e1.90
University 0.44 0.26e0.77 0.61 0.33e1.11 0.47 0.25e0.89 0.78 0.38e1.61
Occupation (Never worked)
Current 0.83 0.49e1.37 1.05 0.59e1.86 0.60 0.31e1.13 0.78 0.38e1.60
Retired 0.85 0.50e1.43 1.01 0.58e1.77 0.44 0.22e0.89 0.51 0.24e1.08
Pack years (Never)
1e20 1.17 0.74e1.84 0.97 0.56e1.70
21e40 1.98 1.30e3.03 2.17 1.33e3.54
41e60 1.87 1.09e3.21 1.99 1.06e3.73
61þ 1.58 0.99e2.52 1.85 1.08e3.18
Alcohol (Never)
Occasionally 0.84 0.45e1.55 0.87 0.44e1.74
Often 0.54 0.30e0.96 0.54 0.28e1.06
Physical activity (Not active) 0.74 0.48e1.15 0.57 0.33e0.97
Diabetes 2.54 1.77e3.65 2.59 1.70e3.96
Hypertension 1.31 0.95e1.80 1.28 0.88e1.87
Hyperlipidemia 1.57 1.14e2.15 1.74 1.19e2.53
Family history 1.35 0.99e1.83 1.46 1.01e2.10
a
Model 1: unadjusted; Model 2: controlling for socio-demographic characteristics; Model 3: controlling for socio-demographic and traditional risk factors of CAD.
b
Linear trend for WP-years as a continuous variable.
 A.M. Sibai et al. / Atherosclerosis 234 (2014) 454e460 459

Table 4
Linear regression analysis with CAD index as the outcome variable.

Variable (Referent category) Model 1a Model 2a Model 3a

b se p-value b se p-value b se p-value

Waterpipe-year (Never)
1e20 3.581 2.122 0.092 2.517 2.064 0.223 3.452 2.091 0.099
21e40 4.080 3.437 0.235 4.206 3.331 0.207 1.992 3.319 0.549
41þ 10.684 3.542 0.003 8.924 3.432 0.009 7.835 3.534 0.027
Age (40e49)
50e59 4.249 1.868 0.023 2.473 1.886 0.19
60e69 7.984 1.959 <0.001 6.648 2.029 0.001

70 14.191 2.261 <0.001 12.932 2.357 <0.001
Gender (Females) 9.767 2.050 <0.001 11.199 2.084 <0.001
Education (Illiterate)
Primary 0.143 2.328 0.951 0.100 2.374 0.966
Complementary 2.140 2.282 0.349 1.178 2.337 0.614
University 5.155 2.525 0.041 3.038 2.605 0.244
Occupation (Never worked)
Current 2.236 2.438 0.359 3.646 2.566 0.156
Retired 1.341 2.416 0.579 2.571 2.404 0.285
Pack Years (Never)
1e20 0.379 0.379 2.084 0.856
21e40 3.771 3.771 1.882 0.045
41e60 3.970 3.970 2.212 0.073
61þ 2.773 2.773 2.051 0.177
Alcohol (Never)
Occasionally 0.434 0.434 2.770 0.876
Often 10.330 10.330 2.708 <0.001
Physical activity (Not active) 3.488 3.488 1.859 0.061
Diabetes 8.545 8.545 1.528 <0.001
Hypertension 0.840 0.840 1.414 0.552
Hyperlipidemia 1.811 1.811 1.393 0.194
Family history 3.812 3.812 1.342 0.005
a
Model 1: unadjusted; Model 2: controlling for socio-demographic characteristics; Model 3: controlling for socio-demographic and traditional risk factors of CAD.

patients and the presence of multiple risk factors influence the rate
of admission to the catheterization laboratory and, hence, the rate
of enrollment into the study. Nevertheless, there is no reason to
believe that the practice of WP smoking increases clinical aware-
ness of CAD or potentiates the search for atherosclerosis as WP
smoking is not yet perceived as an established risk factor for heart
disease by either the public or the medical community [24]. Second,
as in any caseecontrol study, differential reporting bias of exposure
between cases and controls is problematic; however, this is un-
likely in this study as subjects were not informed of the study’s
specific objectives and the majority were interviewed before
knowledge of the results of the cardiac catheterization. Despite the
large sample size of the present study, it was difficult to measure
the pure effect of WP smoking because a considerable proportion of
current WP smokers were current (29.3%) or former (12.2%) ciga-
rette smokers. This may have resulted in an inflation of the effect of
WP smoking acting synergistically with cigarette smoking. Finally,
a major challenge in any study of the health effects of WP remains
the quantification of the exposure. WPs are variable in the amount,
type and composition of the tobacco and the length of the tube
through which the tobacco is smoked, and thus the amount of
inhaled smoke [28]. Because WP smoking sessions are generally
tagged as social events with smokers being in the company of
others sharing in the same smoking ritual and sometimes the same
device, the puff frequency, the interval between puffs, the amount
of smoke inhaled and, hence, the levels of toxicants in the blood can
vary widely [29].
In spite of the above, our study provides the first evidence of a
link between cumulative exposure to WP tobacco and increased
risk of severe CAD stenosis. Furthermore, our results on the asso-
ciation of the classical risk factors with the outcome were direc-
tionally similar to those established in the literature:
atherosclerosis increased with age and was more severe in men
than women; diabetes, hyperlipidemia and family history were
found to increase the risk of CAD; and physical activity, alcohol and
education associated negatively with the outcome, all of which
substantiate the internal validity of the study. Clinically, our study
argues for the need to inquire about the history of WP smoking
from patients with suspected CAD and document WP-years in their
medical charts, in the same manner cigarette pack-years are
documented. This would place WP use at the same level of risk as
cigarette smoking and is likely t6o mitigate misconceptions around
WP smoking and also increase awareness of the hazards of WP
smoking among both the physicians and the patients. With the
increasing global burden of WP smoking, more attention is needed
to document the harmful effects of this form of tobacco smoking,
and to start implementing cessation services and preventive con-
trol measures to address this growing public health epidemic.
Funding
This work was supported by the International Development
Research Centre (IDRC-EDR) [grant number 103436-001] to [AMS].
The funding source had no role in the study design, conduct or
analysis; in the writing of the paper or the decision to submit the
paper for publication.
Contributors
AMS proposed and designed the study, planned the analyses
and contributed significantly to the interpretation of the results and
write-up of the manuscript. RAT coordinated and supervised the
study conduct and analysis. RAT, TI, SY, EN and MMA contributed to
the analysis and helped draft the manuscript. HAI provided critical
input on the plan of analysis and contributed significantly to the
write-up of the manuscript.
460 A.M. Sibai et al. / Atherosclerosis 234 (2014) 454e460
Competing interests
None declared for all authors.
Acknowledgments
We thank all the cardiologists at the American University of
Beirut, Makassed General Hospital, Rizk Hospital and Beirut
Governmental Hospital for contributing patients for this study, the
cardiologists who facilitated conduction of this study (Dr. Samir
Alam, American University of Beirut, Dr. George Ghanem, Rizk
Hospital, Dr. Shawki Abdallah, Makassed General hospital, Dr. Nabil
Shasha, Beirut Governmental hospital, Dr. George Saadeh), and the
interviewers at these hospitals (Mireille Azar, Haitham Awdeh, Bilal
Dimassi, Ghada Hamdar, Rasha Issa, and Fatima Al- Sayah). We also
thank the two anonymous reviewers’ for their comments and
feedback on an earlier version of the manuscript.
References
[1] WHO. WHO report on the global tobacco epidemic; warning about the dan-
gers of tobacco. Geneva: World Health Organization; 2011.
[2] Maziak W. The global epidemic of waterpipe smoking. Addict Behav 2011;36:
1e5.
[3] Warren CW, Lea V, Lee J, et al. Change in tobacco use among 13-15 year olds
between 1999 and 2008: findings from the Global Youth Tobacco Survey. Glob
Health Promotion 2009;16:38e90.
[4] American Lung Association. An emerging deadly trend; waterpipe tobacco
use. Washington, DC: American Lung Association; 2007.
[5] Akl EA, Gaddam S, Gunukula SK, et al. The effects of waterpipe tobacco
smoking on health outcomes: a systematic review. Int J Epidemiol 2010;39:
834e57.
[6] Cobb CO, Sahmarani K, Eissenberg T, et al. Acute toxicant exposure and cardiac
autonomic dysfunction from smoking a single narghile waterpipe with to-
bacco and with a “healthy” tobacco-free alternative. Toxicol Lett 2012;215:
70e5.
[7] Hakim F, Hellou E, Goldbart A, et al. The acute effects of water-pipe smoking
on the cardiorespiratory system. Chest 2011;139:775e81.
[8] Pijls NH, De Bruyne B, Peels K, et al. Measurement of fractional flow reserve to
assess the functional severity of coronary-artery stenoses. N Engl J Med
1996;334:1703e8.
[9] Fried LP, Pearson TA. The association of risk factors with arteriographically
defined coronary artery disease: what is the appropriate control group? Am J
Epidemiol 1987;125:844e53.
[10] Scanlon PJ, Faxon DP, Audet AM, et al. ACC/AHA guidelines for coronary
angiography: executive summary and recommendations. A report of the
American College of Cardiology/American Heart Association Task Force on
Practice Guidelines (Committee on Coronary Angiography) developed in
collaboration with the Society for Cardiac Angiography and Interventions.
Circulation 1999;99:2345e57.
[11] Mark DB, Nelson CL, Califf RM, et al. Continuing evolution of therapy for
coronary artery disease. Initial results from the era of coronary angioplasty.
Circulation 1994;89:2015e25.
[12] Maziak W, Ward KD, Afifi Soweid RA, et al. Standardizing questionnaire items
for the assessment of waterpipe tobacco use in epidemiological studies. Public
Health 2005;119:400e4.
[13] Taylor-Piliae RE, Norton LC, Haskell WL, et al. Validation of a new brief
physical activity survey among men and women aged 60-69 years. Am J
Epidemiol 2006;164:598e606.
[14] McElduff P, Dobson AJ. How much alcohol and how often? Population based
case-control study of alcohol consumption and risk of a major coronary event.
Brit Med J 1997;314:1159e64 [Clinical Research ed.].
[15] Maziak W. The waterpipe: an emerging global risk for cancer. Cancer Epi-
demiol 2013;37:1e4.
[16] Dekker JM, Crow RS, Folsom AR, et al. Low heart rate variability in a 2-minute
rhythm strip predicts risk of coronary heart disease and mortality from
several causes: the ARIC Study. Atherosclerosis risk in communities. Circula-
tion 2000;102:1239e44.
[17] von Kanel R, Carney RM, Zhao S, et al. Heart rate variability and biomarkers of
systemic inflammation in patients with stable coronary heart disease: find-
ings from the Heart and Soul Study. Clin Res Cardiol Off J Ger Cardiac Soc
2011;100:241e7.
[18] Chen CY, Chow D, Chiamvimonvat N, et al. Short-term secondhand smoke
exposure decreases heart rate variability and increases arrhythmia suscepti-
bility in mice. Am J Physiol Heart Circ Physiol 2008;295:H632e9.
[19] Wolfram RM, Chehne F, Oguogho A, et al. Narghile (water pipe) smoking in-
fluences platelet function and (iso-)eicosanoids. Life Sci 2003;74:47e53.
[20] Mehrabi MR, Ekmekcioglu C, Tatzber F, et al. The isoprostane, 8-epi-PGF2
alpha, is accumulated in coronary arteries isolated from patients with coro-
nary heart disease. Cardiovasc Res 1999;43:492e9.
[21] Noonan D. Exemptions for hookah bars in clean indoor air legislation: a public
health concern. Public Health Nurs (Boston, Mass) 2010;27:49e53.
[22] Asfar T, Ward KD, Eissenberg T, et al. Comparison of patterns of use, beliefs,
and attitudes related to waterpipe between beginning and established
smokers. BMC Public Health 2005;5:19.
[23] Aljarrah K, Ababneh ZQ, Al-Delaimy WK. Perceptions of hookah smoking
harmfulness: predictors and characteristics among current hookah users. Tob
Induced Dis 2009;5:16.
[24] Smith-Simone S, Maziak W, Ward KD, et al. Waterpipe tobacco smoking:
knowledge, attitudes, beliefs, and behavior in two U.S. samples. Nicotine Tob
Res Off J Soc Res Nicotine Tob 2008;10:393e8.
[25] WHO. World Health Organization Study Group on tobacco product regulation
(TobReg) advisory note: waterpipe tobacco smoking: health effects, research
needs and recommended actions by regulators. World Health Organization;
2005. p. 3.
[26] Masters N, Tutt C, Yaseen N. Waterpipe tobacco smoking and cigarette
equivalence. Br J General Pract J Royal Coll General Pract 2012;62:127.
[27] Cobb C, Ward KD, Maziak W, et al. Waterpipe tobacco smoking: an emerging
health crisis in the United States. Am J Health Behav 2010;34:275e85.
[28] Knishkowy B, Amitai Y. Water-pipe (narghile) smoking: an emerging health
risk behavior. Pediatrics 2005;116:e113e119.
[29] Shihadeh A, Azar S, Antonios C, et al. Towards a topographical model of
narghile water-pipe cafe smoking: a pilot study in a high socioeconomic
status neighborhood of Beirut, Lebanon. Pharmacol Biochem Behav 2004;79:
75e82.