- Email: [email protected]
Waterpipe tobacco smoking prevalence and illegal underage use in waterpipe-serving premises: a cross-sectional analysis among schoolchildren in Stoke-on-Trent
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Waterpipe tobacco smoking prevalence and illegal underage use in waterpipe-serving premises: a cross-sectional analysis among schoolchildren in Stoke-on-Trent M. Jawad a,*, C. McIver b a b
Public Health Policy Evaluation Unit, Imperial College London, Hammersmith, W6 8RP, United Kingdom Department of Public Health, Stoke-on-Trent City Council, Stoke-on-Trent, ST4 1HH United Kingdom
article info
abstract
Article history:
Objectives: Waterpipe tobacco smoking has received little epidemiological and policy
Received 16 August 2016
attention in the UK despite reports of increasing prevalence alongside an anecdotally non-
Received in revised form
compliant industry. This study aimed to determine how waterpipe tobacco smoking is
14 December 2016
changing among young people in the UK, both in terms of prevalence and sociodemo-
Accepted 18 December 2016
graphic correlates of use, and to quantify the extent of illegal underage use in waterpipeserving premises in the UK. Study design: Repeat cross-sectional.
Keywords:
Methods: A secondary analysis of two cross-sectional surveys (total N ¼ 3376), conducted in
Waterpipe
2013 and 2015 among secondary school students aged 11e16 years in Stoke-on-Trent,
Shisha
measured lifetime (both surveys) and regular (at least monthly; 2015 survey only) water-
Hookah
pipe tobacco prevalence and location of usual use. Logistic regression models measured
Tobacco
the association between independent variables (age, sex, ethnicity, presence of free school
Smoking
meals, cigarette smoking status) with lifetime and regular waterpipe tobacco use, and with
Legislation
illegal underage use; the latter defined as usually smoking waterpipe tobacco in a
Policy
waterpipe-serving premise. Results: Lifetime waterpipe tobacco prevalence remained similar in 2013 (13.7%, 95% confidence interval [CI] 12.0e15.4%) and 2015 (14.6%, 95% CI 12.8e16.4%), whereas regular use was measured at 2.9% (95% CI 2.1e3.8%) in 2015. Older, non-white, males who concurrently used cigarettes had higher odds of lifetime waterpipe tobacco use. Illegal underage use was reported among 27.1% of all regular users, correlates of which included increasing age and South Asian ethnicity. The presence of free school meals was not associated with lifetime or regular waterpipe tobacco prevalence, nor illegal underage use. Conclusions: Increased monitoring of waterpipe tobacco prevalence and patterns, including the underage policy compliance of waterpipe-serving premises, is needed to help inform policy decisions to control waterpipe tobacco use. © 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. http://dx.doi.org/10.1016/j.puhe.2016.12.028 0033-3506/© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Introduction Waterpipe tobacco smoking is a centuries-old practice where charcoal-heated tobacco smoke passes through a multistemmed, water-containing instrument before inhalation. It has been dubbed a ‘global epidemic’ given high prevalence estimates worldwide:1e3 in areas of Eastern Europe, over 20% of secondary school students have tried waterpipe tobacco in the past 30 days; a figure reaching over 30% for secondary school students in Lebanon and the West Bank.4 High prevalence of use is secondary to widespread social acceptability, accessibility of waterpipe-serving premises and affordability. Central to a reduced harm perception is the false belief that its smoke is safely filtered by the water at the base of the instrument;5 a fallacy that has been since disproven after the identification of significant quantities of tar, nicotine, carbon monoxide and other carcinogens in waterpipe tobacco smoke.6e9 As a combustible tobacco product, waterpipe tobacco has, expectedly, been shown to be associated with an array of tobacco-related diseases, including cancers of the lung and oropharynx, cardiovascular diseases and other respiratory conditions.10,11 Despite this, waterpipe tobacco has been given relatively little attention in the United Kingdom (UK), both epidemiologically and legislatively. The first prevalence study in the UK was published in 2008, reporting that 8.0% of students in one university were regular waterpipe tobacco users, and nearly 40% were lifetime users.12 Since then studies have been largely confined to non-nationally representative populations, such as local surveys among school students13,14 and university students,15,16 with lifetime waterpipe tobacco prevalence estimates ranging from 24% to 66%. No longitudinal or repeat cross-sectional studies on waterpipe tobacco prevalence have been conducted on young people in the UK. One large repeat cross-sectional online survey of adults in Great Britain (England, Wales and Scotland only) reported that 11.6% of adults had ever smoked waterpipe tobacco and that 1.0% smoked waterpipe tobacco at least monthly, with nonsignificant differences in prevalence between 2012 and 2013 .17 In this study, higher odds of use were seen among younger males of non-white ethnicities and higher socioeconomic groups. Despite reports of high prevalence in selected population groups, the UK has been slow to respond legislatively. While waterpipe tobacco is not exempt from any of the UK's tobacco laws, which are among the strongest in Europe,18 there have been numerous reports of difficulty in applying these laws to waterpipe-serving premises.19,20 A classic example is enforcing indoor public smoking bans on a product which is smoked for approximately an hour in duration, in a country with a generally cold climate. Another is enforcing health warning labels on the tobacco-packed apparatuses, often cumbersome and large, which are presented to customers at such premises.20 Illegal underage consumption is also a concern; whereas the legal age for tobacco purchases in England and Wales was increased from 16 to 18 years old in 2007, qualitative reports from law enforcers suggest illegal underage use is commonplace in waterpipe-serving premises,20 but this has not been quantified.
33
In light of the lack of waterpipe prevalence data among young people in the UK, and the lack of research attention given to illegal underage use in waterpipe-serving premises, this study has two main aims. First, we aimed to determine how waterpipe tobacco prevalence is changing among young people, both in terms of prevalence and sociodemographic correlates of use. Second, we aimed to quantify the extent of illegal underage use in waterpipe-serving premises.
Methods Design, sample, setting This study was a secondary analysis of two cross-sectional surveys conducted in 2013 and 2015 among secondary school students aged 11e16 years in Stoke-on-Trent, one of the most deprived cities in the United Kingdom. Both surveys used similar methodological designs, in which all schools in Stoke-on-Trent were invited to participate. In the 2015 survey, fourteen secondary schools, two Pupil Referral Units and four Special Schools were invited to participate, of which six secondary schools agreed to take part. The main reasons given by schools declining to take part in the survey were lack of time and competing priorities (such as upcoming government inspections and regulatory visits). Two schools gave no response to the invitation. More details on the 2013 survey methodology can be found elsewhere.21
Questionnaire and measures The Stoke-on-Trent Young People's Lifestyle Survey was designed through a multi-partner consultation which included local schools, the Schools Health Education Unit, the National Health Service and the local government. Some questions were developed locally while others were taken from previously validated surveys. The 2013 survey contained 62 questions and the 2015 survey contained 64 questions; these were organised into six themes: tobacco use, food and drink, alcohol and drugs, emotional wellbeing, sexual health and sociodemographic questions. In both years, waterpipe tobacco prevalence was gathered using the following question: ‘have you ever smoked ‘shisha’ (also known as a ‘waterpipe’)?’ In 2013, the response options were ‘no/don't know/yes', whereas in 2015 the response options were ‘I have never smoked shisha/I have tried shisha once or twice/I use them sometimes (more than once a month)/I use them often (more than once a week)’. The 2015 survey also included an image of a waterpipe apparatus next to this question. Both surveys also asked ‘Where do you smoke it [waterpipe]?’ as a follow on question, and response options were ‘at home/at school/in a shisha bar or lounge/at a friend's house/other’. The primary outcome measures were lifetime waterpipe use (ascertained by answering ‘yes’ to the 2013 waterpipe tobacco prevalence question, and anything except ‘I have never smoked shisha’ for the 2015 waterpipe tobacco prevalence question), and regular waterpipe use, ascertained by answering positively to smoking either more than once a month or more than once a week in the 2015 survey. Independent variables included age (reported as a continuous
34
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
variable from 11 to 16 years), sex, ethnicity (22 categories taken from the UK census, including ‘I don't know’ and ‘I don't want to say’), the presence of free school meals (yes/no/don't know/don't want to say) and lifetime (ever) and current (occasional or regular) cigarette use. The presence of free school meals was used as a proxy measure for low socioeconomic status.
Statistical analysis Ethnicity was recoded into three categories: white (white British, Irish, Traveller of Irish heritage, Romany or Gypsy, Polish and other white background), South Asian (British Asian, Indian Asian, Pakistani Asian, Bangladeshi Asian and any other Asian background) and other (mixed backgrounds, black backgrounds, other backgrounds not listed and those reporting ‘I don't know’ and ‘I don't want to say’). Free school meals was categorised into three categories: no, I don't know/I don't want to say, and yes. Categorical data were described using frequencies and percentages, and continuous data were described using means and standard deviations. Sample characteristics were descriptively reported, with independent sample ttests and Chi-squared tests to test the differences between both surveys' sample characteristics. Sample characteristics were also descriptively reported by waterpipe status (lifetime or regular). Two multi-variable logistic regression models assessed the association of the independent variables (age, sex, ethnicity, free school meals, current cigarette use) with each outcome measure (lifetime or regular waterpipe tobacco use). In the model reporting lifetime use, both the 2013 and 2015 samples were combined. In the model reporting regular use, only the 2015 sample was used. Due to the anonymised nature of the data repeat responders could not be identified. These models reported adjusted odds ratios (AORs) and 95% confidence intervals (95% CIs) and used age as a continuous variable rather than categorical. To assess illegal underage use, we measured the prevalence of regular waterpipe tobacco use by smoking location, then a final multi-variable logistic regression model assessed the association of the independent variables with usual waterpipe smoking location being at a shisha bar/ lounge, which for the purposes of this study is our measure of ‘illegal underage use’ given all respondents were under 18 years old.
Results Sample characteristics Both surveys had over 1500 respondents each, from which the pooled mean age was 13.5 ± 1.5 years, 48.0% were male, 83.2% were white, 18.2% reported having free school meals and the prevalence of lifetime and current cigarette use was 20.8% and 6.9%, respectively. Table 1 shows the sample characteristics for each survey. These were broadly similar however after statistical testing, the 2015 sample was found to be significantly older than the 2013 sample (13.3 vs 13.7 years, P < 0.001), have a lower proportion of white respondents
(85.2% vs 81.0%, P < 0.01) and a lower proportion of those reporting free school meals (21.0% vs 15.4%, P < 0.001).
Waterpipe tobacco prevalence Lifetime waterpipe prevalence remained similar in 2013 (13.7%, 95% CI 12.0e15.4%) and 2015 (14.6%, 95% CI 12.8e16.4%). In 2015, regular waterpipe use was measured at 2.9% (95% CI 2.1e3.8%). Table 2 shows sample characteristics by waterpipe tobacco status. The main difference in lifetime waterpipe prevalence between the two surveys was found for sex: while males reported lifetime waterpipe use more than females in 2013 (16.8% vs 10.9%), this appeared more balanced in 2015 (14.9% vs 14.0%). Otherwise, for both lifetime and regular use, waterpipe prevalence increased with age, nonwhite ethnicities and cigarette use. For example, in 2015 regular waterpipe prevalence was 5.5% among those aged 16 years compared with 1.6% among those aged 11 years, and 12.0% among lifetime cigarette users compared with 0.7% among never cigarette users. In adjusted logistic regression models (Table 3), the odds of lifetime waterpipe tobacco use increased by 45% for every year of age (AOR 1.45, 95% CI 1.34e1.59). Odds of lifetime use were also higher for males (AOR 1.48, 95% CI 1.17e1.86) non-White ethnicities (e.g. South Asian vs White: AOR 3.25, 95% CI 2.36e4.47) and current cigarette use (AOR 7.22, 95% CI 5.23e9.96). The odds of regular waterpipe tobacco smoking increased by 40% for every year of age (AOR 1.40, 95% CI 1.09e1.80). Odds of regular use were also higher for current cigarette users (AOR 8.87, 95% CI 4.16e18.91). For both measures of waterpipe use, no association was seen for free school meals.
Table 1 e Sample characteristics. Characteristic
Total Age 11 years 12 years 13 years 14 years 15 years 16 years Sex Female Male Ethnicity White South Asian Other Free school meals No Don't know/don't want to say Yes Lifetime cigarette use No Yes Current cigarette use No Yes
2013
2015
N
%
N
%
1772
100.0
1604
100.0
176 368 433 394 230 146
10.1 21.1 24.8 22.6 13.2 8.4
135 289 263 400 257 233
8.6 18.3 16.7 25.4 16.3 14.8
897 875
50.6 49.4
830 723
53.4 46.6
1490 147 112
85.2 8.4 6.4
1275 168 132
81.0 10.7 8.4
1320 63 368
75.4 3.6 21.0
1268 65 242
80.5 4.1 15.4
1324 375
77.9 22.1
1233 298
30.5 19.5
1576 123
92.8 7.2
1469 104
93.4 6.6
35
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Table 2 e Waterpipe tobacco smoking by sample characteristics. Characteristic
2013
2015
Lifetime waterpipe users
Total Age 11 years 12 years 13 years 14 years 15 years 16 years Sex Female Male Ethnicity White South Asian Other Free school meals No Don't know/don't want to say Yes Lifetime cigarette use No Yes Current cigarette use No Yes
Age Sex Female Male Ethnicity White South Asian Other Free school meals No Don't know/don't want to say Yes Current cigarette use No Yes
Regular waterpipe users
N
%
N
%
N
%
221
13.7
214
14.6
43
2.9
1 13 51 51 66 35
0.7 3.9 13.1 14.3 29.7 25.9
8 15 23 68 50 49
6.5 5.6 9.7 18.5 21.4 22.6
2 2 2 14 10 12
1.6 0.8 0.8 3.8 4.3 5.5
90 131
10.9 16.8
97 109
14.9 14.0
20 20
2.6 3.1
166 41 11
12.1 31.8 11.6
161 29 21
13.7 18.5 18.6
31 5 5
2.6 3.2 4.4
165 12 40
13.6 20.0 12.4
167 8 35
14.3 13.3 16.1
32 3 7
2.7 5.0 3.2
90 115
7.4 33.3
94 109
8.2 38.5
8 34
0.7 12.0
146 59
10.1 52.2
164 45
12.1 45.5
27 16
2.0 16.2
Table 3 e Correlates of regular waterpipe tobacco use (2015 sample only) and lifetime waterpipe tobacco use (2013 and 2015 pooled samples). Characteristic
Lifetime waterpipe users
Regular waterpipe users
Table 4 e Regular waterpipe tobacco use by location of use (2015 sample only). Location
Lifetime waterpipe users
AOR
95% CI
AOR
95% CI
1.40
1.09, 1.80**
1.45
1.34, 1.59***
1.00 1.22
0.63, 2.39
1.00 1.48
1.17, 1.86**
1.00 2.20 2.48
0.79, 6.15 0.88, 6.98
1.00 3.25 1.70
2.36, 4.47*** 1.10, 2.62*
1.00 1.15
0.24, 5.38
1.00 1.10
0.62, 1.94
1.20
0.50, 2.87
0.94
0.69, 1.27
1.00 8.87
4.16, 18.91***
1.00 7.22
5.23, 9.96***
Abbreviations: AOR, adjusted odds ratio; CI, confidence interval. *P < 0.05, **P < 0.01, ***P < 0.001.
Friend's home No Yes Home No Yes Shisha bar/lounge No Yes School No Yes Other No Yes
Regular waterpipe users N
%
18 22
13.3 40.7
31 7
18.8 35.0
26 13
18.2 27.1
16 22
14.0 29.3
28 10
21.1 18.5
However, the prevalence of regular waterpipe tobacco use was the highest in friend's homes (40.7%) and own homes (35.0%) than in shisha bars/lounges (27.1%).
Location of waterpipe tobacco smoking
Underage use
Table 4 shows regular waterpipe tobacco use prevalence, as a proportion of all waterpipe tobacco users, by location of usual use. The prevalence of regular waterpipe tobacco use was higher in all reported locations except the ‘other’ location.
Table 5 presents the correlates of illegal underage shisha bar/ lounge use. For every 1 year increase in age, the odds of illegal underage use increased by 36% (AOR 1.36, 95% CI 1.14e1.62). The odds of illegal underage use were also higher among
36
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Table 5 e Correlates of illegal underage shisha bar/lounge use pooled across both samples. Characteristic
Age Sex Female Male Ethnicity White South Asian Other Free school meals No Don't know/don't want to say Yes Current cigarette use No Yes
Illegal underage shisha bar/lounge use AOR
95% CI
1.36
1.14, 1.62**
1.00 1.20
0.78, 1.84
1.00 2.99 0.83
1.65, 5.42*** 0.35, 1.94
1.00 0.83 0.76
0.29, 2.39 0.43, 1.32
1.00 1.80
1.09, 2.99*
Abbreviations: AOR, adjusted odds ratio; CI, confidence interval. *P < 0.05, **P < 0.01, ***P < 0.001.
South Asians (vs White, AOR 2.99, 95% CI 1.65e5.42) and current cigarette users (AOR 1.80, 95% CI 1.09e2.99).
Discussion Main findings This study among 11- to 16-year-old secondary school students in Stoke-on-Trent showed that nearly one in seven had tried waterpipe tobacco smoking, with no significant change in use between 2013 and 2015, and that 3% were regular users. Waterpipe tobacco use was associated with increasing age, males, non-white ethnicities and current cigarette use, but not the presence of free school meals, our proxy measure for low socioeconomic status. Our correlates of use are largely in line with the reports from the US22 and Europe.4 Regular use was highest in those reporting usual smoking locations being friends' or own homes, although over a quarter of lifetime users reported illegal underage use, which was more apparent among older and South Asian students.
Previous literature To the best of knowledge, there are no published national surveys of waterpipe tobacco prevalence among young people in the UK, however similar school-based surveys in London report a higher lifetime prevalence (24.0e39.6%)13,14 than our sample (13.7e14.6%). This difference may be due to the higher proportion of non-White students in London, who are most likely to use waterpipe tobacco, or the extremely high number of waterpipe-serving premises (approximately 400) in known operation there.20 Online cross-sectional surveys among adults in Great Britain in 2012 and 2013 showed 11.6% had tried waterpipe tobacco, and there was little variation between the two years.17 In the absence of longitudinal data, the existing evidence may therefore suggest waterpipe tobacco
smoking prevalence is plateauing in the UK, but longitudinal studies are needed to confirm or refute this. The 2015 wave of a yearly multiple cross-sectional survey of school students in the US also suggests a plateau in waterpipe tobacco use (at around 7%);23 however, continual rising rates have been documented in several Arab countries24,25 and Sweden.26 Our findings on illegal under age use in waterpipe-serving premises are informative to health policy makers. Studies from the US and UK have shown that educational establishment proximity to waterpipe-serving premises are an independent predictor of waterpipe use.14,27,28 While this suggests that such premises should be a focus of policy intervention, our findings showed that regular use is more highly reported in domestic settings. This corroborates with qualitative research from university students who report smoking in commercial waterpipe premises can be expensive and less accessible than domestic use28,29 and lends itself to the need for a multi-faceted policy approach that goes beyond simply stronger enforcement of waterpipe-serving premises. Illegal underage use in waterpipe-serving premises has been previously identified in London, both from surveys among young people14 and law enforcers.20 In addition to accessibility, affordability and acceptability, reasons for this are likely to include the attraction towards flavoured tobacco,30 the general lack of tobacco legislation compliance from the waterpipe tobacco industry,20 and ritualistic aspects of waterpipe tobacco use.31 It is important to note that waterpipe-serving premises are not homogenousdsome resemble loud club-like environments, whereas others resemble quiet coffee shops.31 Further research should identify which types, if any, young people are particularly attracted to, and interventions to kerb such use.
Implications Given a quarter of lifetime waterpipe users report illegal underage use, this calls into question the degree of tobacco legislation self-enforcement in waterpipe-serving premises. Such premises are required to display a statutory notice explaining that it is illegal to sell tobacco products to under 18s; however, qualitative reports with law enforcers in London suggest waterpipe-serving premises are deliberately and recurrently non-compliant with such regulations.20 Turkey and several Gulf countries have zoning laws in place to prohibit the sale of waterpipe tobacco within a certain distance of educational establishments in a bid to kerb youth waterpipe tobacco smoking.19 Research evaluating tobacco control policy on waterpipe tobacco smoking behaviour are lacking32 and should be considered a research priority in this field. This study has other research implications. Research should explore how the location and frequency of waterpipe tobacco smoking changes over time. Whether students are introduced to waterpipe tobacco at commercial premises then transition to regular waterpipe tobacco smoking at home remains unknown. This has implications for tobacco control policy: an emphasis on enforcing policy on commercial premises may simply be offset by increased home use.29 Better quality epidemiological data concerning prevalence of use and surveillance of waterpipe-serving premises are also needed.
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Limitations The sampling method may be affected by selection bias; schools whose staff believe tobacco smoking is an issue for them may be more likely to agree to take part in this survey. However, our sample characteristics are broadly representative of the local population.21 Due to the low number of regular waterpipe tobacco users, statistical analysis may be underpowered to detect important associations; this may have been the case looking at the association between ethnicity and regular waterpipe tobacco use, which displayed a high odds ratio (greater than 2) but non-significant associations due to wide confidence intervals. Tobacco use prevalence is selfreported and may be prone to measurement bias (underreporting); however, self-reported tobacco use is known to correlate well with biochemical markers such as serum cotinine.33 Finally, given we could not identify data repeat responders from the anonymised data sets, students may have been double counted for the lifetime pooled analyses shown in Tables 3 and 5; however, we believe this to be minimal.
Conclusions Nearly one in seven students in this sample have tried waterpipe tobacco smoking, of which a quarter report illegal underage use in waterpipe-serving premises. Increased monitoring of waterpipe tobacco prevalence and of waterpipe-serving premises is required with a view to generate more policy evaluation in this public health area.
Author statements
5.
6.
7.
8.
9. 10.
11.
12.
13.
14.
Ethical approval Data from this study were taken from routine health survey data and as such ethical approval was not required.
Funding
15.
16.
None.
Competing interests
17.
None declared.
references
18. 19.
1. Ward KD. The waterpipe: an emerging global epidemic in need of action. Tob Control 2015;24(Suppl. 1):i1e2. 2. Maziak W. The global epidemic of waterpipe smoking. Addict Behav 2011;36:1e5. 3. Maziak W. Commentary: the waterpipeea global epidemic or a passing fad. Int J Epidemiol 2010;39:857e9. 4. Jawad M, Lee JT, Millett C. Waterpipe tobacco smoking prevalence and correlates in 25 Eastern Mediterranean and Eastern European countries: cross-sectional analysis of the
20.
21.
37
Global Youth Tobacco Survey. Nicotine Tob Res 2016;18(4):395e402. Akl EA, Jawad M, Lam WY, Co CN, Obeid R, Irani J. Motives, beliefs and attitudes towards waterpipe tobacco smoking: a systematic review. Harm Reduct J 2013;10:12. Daher N, Saleh R, Jaroudi E, Sheheitli H, Badr T, Sepetdjian E, et al. Comparison of carcinogen, carbon monoxide, and ultrafine particle emissions from narghile waterpipe and cigarette smoking: sidestream smoke measurements and assessment of second-hand smoke emission factors. Atmos Environ 1994;2010(44):8e14. Monzer B, Sepetdjian E, Saliba N, Shihadeh A. Charcoal emissions as a source of CO and carcinogenic PAH in mainstream narghile waterpipe smoke. Food Chem Toxicol 2008;46:2991e5. Sepetdjian E, Shihadeh A, Saliba NA. Measurement of 16 polycyclic aromatic hydrocarbons in narghile waterpipe tobacco smoke. Food Chem Toxicol 2008;46:1582e90. Shihadeh A. Investigation of mainstream smoke aerosol of the argileh water pipe. Food Chem Toxicol 2003;41:143e52. Waziry K, Jawad M, Ballout RA, Al Akel M, Akl EA. The effects of waterpipe tobacco smoking on health outcomes: an updated systematic review and meta-analysis. Int J Epidemiol 2016. http://dx.doi.org/10.1016/10.1093/ije/dyw021 [Epub ahead of print]. Raad D, Gaddam S, Schunemann HJ, Irani J, Jaoude PA, Honeine R, et al. Effects of water-pipe smoking on lung function: a systematic review and meta-analysis. Chest 2011;139:764e74. Jackson D, Aveyard P. Waterpipe smoking in students: prevalence, risk factors, symptoms of addiction, and smoke intake. Evidence from one British University. BMC Public Health 2008;8:174. Jawad M, Power G. Prevalence, correlates and patterns of waterpipe smoking among secondary school students in southeast London: a cross-sectional study. BMC Public Health 2016;16:1. Jawad M, Wilson A, Lee JT, Jawad S, Hamilton FL, Millett C. Prevalence and predictors of water pipe and cigarette smoking among secondary school students in London. Nicotine Tob Res 2013;15:2069e75. Jawad M, Choaie E, Brose L, Dogar O, Grant A, Jenkinson E, et al. Waterpipe tobacco use in the United Kingdom: a crosssectional study among university students and stop smoking practitioners. PloS One 2016;11:e0146799. Jawad M, Abass J, Hariri A, Rajasooriar KG, Salmasi H, Millett C, et al. Waterpipe smoking: prevalence and attitudes among medical students in London. Int J Tuberc Lung Dis 2013;17:137e40. Grant A, Morrison R, Dockrell MJ. Prevalence of waterpipe (shisha, narghille, hookah) use among adults in Great Britain and factors associated with waterpipe use: data from crosssectional Online Surveys in 2012 and 2013. Nicotine Tob Res 2014;16:931e8. Joossens L, Raw M. The Tobacco Control Scale: a new scale to measure country activity. Tob Control 2006;15:247e53. Jawad M, El Kadi L, Mugharbil S, Nakkash R. Waterpipe tobacco smoking legislation and policy enactment: a global analysis. Tob Control 2015;24(Suppl. 1):i60e5. Jawad M. Legislation enforcement of the waterpipe tobacco industry: a qualitative analysis of the London experience. Nicotine Tob Res 2014;16:1000e8. Jawad M, McIver C, Iqbal Z. Prevalence and correlates of lifetime waterpipe, cigarette, alcohol and drug use among secondary school students in Stoke-on-Trent, UK: a post hoc cross-sectional analysis. J Public Health (Oxf) 2014;36:615e21.
38
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
22. Maziak W, Taleb ZB, Bahelah R, Islam F, Jaber R, Auf R, et al. The global epidemiology of waterpipe smoking. Tob Control 2015;24:i3e12. 23. Singh T, Arrazola RA, Corey CG, Husten CG, Neff LJ, Homa DM, et al. Tobacco use among middle and high school students e United States, 2011e2015. MMWR Morb Mortal Wkly Rep 2016;65:361e7. 24. McKelvey KL, Wilcox ML, Madhivanan P, Mzayek F, Khader YS, Maziak W. Time trends of cigarette and waterpipe smoking among a cohort of school children in Irbid, Jordan, 2008e11. Eur J Public Health 2013;23:862e7. 25. Saade G, Warren CW, Jones NR, Asma S, Mokdad A. Linking Global Youth Tobacco Survey (GYTS) data to the WHO Framework Convention on Tobacco Control (FCTC): the case for Lebanon. Prev Med 2008;47(Suppl. 1):S15e9. 26. Galanti MR, Al-Adhami M. Use of a water pipe is not an alternative to other tobacco or substance use among adolescents: results from a national survey in Sweden. Nicotine Tob Res 2015;17:74e80. 27. Sutfin EL, McCoy TP, Reboussin BA, Wagoner KG, Spangler J, Wolfson M. Prevalence and correlates of waterpipe tobacco smoking by college students in North Carolina. Drug Alcohol Depend 2011;115:131e6.
28. Kassem NO, Jackson SR, Boman-Davis M, Kassem NO, Liles S, Daffa RM, et al. Hookah smoking and facilitators/barriers to lounge use among students at a US University. Am J Health Behav 2015;39:832e48. 29. Jawad M, Jawad S, Mehdi A, Sardar A, Jawad AM, Hamilton FL. A qualitative analysis among regular waterpipe tobacco smokers in London universities. Int J Tuberc Lung Dis 2013;17:1364e9. 30. Salloum RG, Maziak W, Hammond D, Nakkash R, Islam F, Cheng X, et al. Eliciting preferences for waterpipe tobacco smoking using a discrete choice experiment: implications for product regulation. BMJ open 2015;5:e009497. 31. Carroll MV, Chang J, Sidani JE, Barnett TE, Soule E, Balbach E, et al. Reigniting tobacco ritual: waterpipe tobacco smoking establishment culture in the United States. Nicotine Tob Res 2014;16:1549e58. 32. Jawad M, Jawad S, Waziry RK, Ballout RA, Akl EA. Interventions for waterpipe tobacco smoking prevention and cessation: a systematic review. Sci Rep 2016;6:25872. € la € T, Lillsunde P, Puska P. Validation of self 33. Vartiainen E, Seppa reported smoking by serum cotinine measurement in a community-based study. J Epidemiol Community Health 2002;56:167e70.
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Waterpipe tobacco smoking prevalence and illegal underage use in waterpipe-serving premises: a cross-sectional analysis among schoolchildren in Stoke-on-Trent M. Jawad a,*, C. McIver b a b
Public Health Policy Evaluation Unit, Imperial College London, Hammersmith, W6 8RP, United Kingdom Department of Public Health, Stoke-on-Trent City Council, Stoke-on-Trent, ST4 1HH United Kingdom
article info
abstract
Article history:
Objectives: Waterpipe tobacco smoking has received little epidemiological and policy
Received 16 August 2016
attention in the UK despite reports of increasing prevalence alongside an anecdotally non-
Received in revised form
compliant industry. This study aimed to determine how waterpipe tobacco smoking is
14 December 2016
changing among young people in the UK, both in terms of prevalence and sociodemo-
Accepted 18 December 2016
graphic correlates of use, and to quantify the extent of illegal underage use in waterpipeserving premises in the UK. Study design: Repeat cross-sectional.
Keywords:
Methods: A secondary analysis of two cross-sectional surveys (total N ¼ 3376), conducted in
Waterpipe
2013 and 2015 among secondary school students aged 11e16 years in Stoke-on-Trent,
Shisha
measured lifetime (both surveys) and regular (at least monthly; 2015 survey only) water-
Hookah
pipe tobacco prevalence and location of usual use. Logistic regression models measured
Tobacco
the association between independent variables (age, sex, ethnicity, presence of free school
Smoking
meals, cigarette smoking status) with lifetime and regular waterpipe tobacco use, and with
Legislation
illegal underage use; the latter defined as usually smoking waterpipe tobacco in a
Policy
waterpipe-serving premise. Results: Lifetime waterpipe tobacco prevalence remained similar in 2013 (13.7%, 95% confidence interval [CI] 12.0e15.4%) and 2015 (14.6%, 95% CI 12.8e16.4%), whereas regular use was measured at 2.9% (95% CI 2.1e3.8%) in 2015. Older, non-white, males who concurrently used cigarettes had higher odds of lifetime waterpipe tobacco use. Illegal underage use was reported among 27.1% of all regular users, correlates of which included increasing age and South Asian ethnicity. The presence of free school meals was not associated with lifetime or regular waterpipe tobacco prevalence, nor illegal underage use. Conclusions: Increased monitoring of waterpipe tobacco prevalence and patterns, including the underage policy compliance of waterpipe-serving premises, is needed to help inform policy decisions to control waterpipe tobacco use. © 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. http://dx.doi.org/10.1016/j.puhe.2016.12.028 0033-3506/© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Introduction Waterpipe tobacco smoking is a centuries-old practice where charcoal-heated tobacco smoke passes through a multistemmed, water-containing instrument before inhalation. It has been dubbed a ‘global epidemic’ given high prevalence estimates worldwide:1e3 in areas of Eastern Europe, over 20% of secondary school students have tried waterpipe tobacco in the past 30 days; a figure reaching over 30% for secondary school students in Lebanon and the West Bank.4 High prevalence of use is secondary to widespread social acceptability, accessibility of waterpipe-serving premises and affordability. Central to a reduced harm perception is the false belief that its smoke is safely filtered by the water at the base of the instrument;5 a fallacy that has been since disproven after the identification of significant quantities of tar, nicotine, carbon monoxide and other carcinogens in waterpipe tobacco smoke.6e9 As a combustible tobacco product, waterpipe tobacco has, expectedly, been shown to be associated with an array of tobacco-related diseases, including cancers of the lung and oropharynx, cardiovascular diseases and other respiratory conditions.10,11 Despite this, waterpipe tobacco has been given relatively little attention in the United Kingdom (UK), both epidemiologically and legislatively. The first prevalence study in the UK was published in 2008, reporting that 8.0% of students in one university were regular waterpipe tobacco users, and nearly 40% were lifetime users.12 Since then studies have been largely confined to non-nationally representative populations, such as local surveys among school students13,14 and university students,15,16 with lifetime waterpipe tobacco prevalence estimates ranging from 24% to 66%. No longitudinal or repeat cross-sectional studies on waterpipe tobacco prevalence have been conducted on young people in the UK. One large repeat cross-sectional online survey of adults in Great Britain (England, Wales and Scotland only) reported that 11.6% of adults had ever smoked waterpipe tobacco and that 1.0% smoked waterpipe tobacco at least monthly, with nonsignificant differences in prevalence between 2012 and 2013 .17 In this study, higher odds of use were seen among younger males of non-white ethnicities and higher socioeconomic groups. Despite reports of high prevalence in selected population groups, the UK has been slow to respond legislatively. While waterpipe tobacco is not exempt from any of the UK's tobacco laws, which are among the strongest in Europe,18 there have been numerous reports of difficulty in applying these laws to waterpipe-serving premises.19,20 A classic example is enforcing indoor public smoking bans on a product which is smoked for approximately an hour in duration, in a country with a generally cold climate. Another is enforcing health warning labels on the tobacco-packed apparatuses, often cumbersome and large, which are presented to customers at such premises.20 Illegal underage consumption is also a concern; whereas the legal age for tobacco purchases in England and Wales was increased from 16 to 18 years old in 2007, qualitative reports from law enforcers suggest illegal underage use is commonplace in waterpipe-serving premises,20 but this has not been quantified.
33
In light of the lack of waterpipe prevalence data among young people in the UK, and the lack of research attention given to illegal underage use in waterpipe-serving premises, this study has two main aims. First, we aimed to determine how waterpipe tobacco prevalence is changing among young people, both in terms of prevalence and sociodemographic correlates of use. Second, we aimed to quantify the extent of illegal underage use in waterpipe-serving premises.
Methods Design, sample, setting This study was a secondary analysis of two cross-sectional surveys conducted in 2013 and 2015 among secondary school students aged 11e16 years in Stoke-on-Trent, one of the most deprived cities in the United Kingdom. Both surveys used similar methodological designs, in which all schools in Stoke-on-Trent were invited to participate. In the 2015 survey, fourteen secondary schools, two Pupil Referral Units and four Special Schools were invited to participate, of which six secondary schools agreed to take part. The main reasons given by schools declining to take part in the survey were lack of time and competing priorities (such as upcoming government inspections and regulatory visits). Two schools gave no response to the invitation. More details on the 2013 survey methodology can be found elsewhere.21
Questionnaire and measures The Stoke-on-Trent Young People's Lifestyle Survey was designed through a multi-partner consultation which included local schools, the Schools Health Education Unit, the National Health Service and the local government. Some questions were developed locally while others were taken from previously validated surveys. The 2013 survey contained 62 questions and the 2015 survey contained 64 questions; these were organised into six themes: tobacco use, food and drink, alcohol and drugs, emotional wellbeing, sexual health and sociodemographic questions. In both years, waterpipe tobacco prevalence was gathered using the following question: ‘have you ever smoked ‘shisha’ (also known as a ‘waterpipe’)?’ In 2013, the response options were ‘no/don't know/yes', whereas in 2015 the response options were ‘I have never smoked shisha/I have tried shisha once or twice/I use them sometimes (more than once a month)/I use them often (more than once a week)’. The 2015 survey also included an image of a waterpipe apparatus next to this question. Both surveys also asked ‘Where do you smoke it [waterpipe]?’ as a follow on question, and response options were ‘at home/at school/in a shisha bar or lounge/at a friend's house/other’. The primary outcome measures were lifetime waterpipe use (ascertained by answering ‘yes’ to the 2013 waterpipe tobacco prevalence question, and anything except ‘I have never smoked shisha’ for the 2015 waterpipe tobacco prevalence question), and regular waterpipe use, ascertained by answering positively to smoking either more than once a month or more than once a week in the 2015 survey. Independent variables included age (reported as a continuous
34
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
variable from 11 to 16 years), sex, ethnicity (22 categories taken from the UK census, including ‘I don't know’ and ‘I don't want to say’), the presence of free school meals (yes/no/don't know/don't want to say) and lifetime (ever) and current (occasional or regular) cigarette use. The presence of free school meals was used as a proxy measure for low socioeconomic status.
Statistical analysis Ethnicity was recoded into three categories: white (white British, Irish, Traveller of Irish heritage, Romany or Gypsy, Polish and other white background), South Asian (British Asian, Indian Asian, Pakistani Asian, Bangladeshi Asian and any other Asian background) and other (mixed backgrounds, black backgrounds, other backgrounds not listed and those reporting ‘I don't know’ and ‘I don't want to say’). Free school meals was categorised into three categories: no, I don't know/I don't want to say, and yes. Categorical data were described using frequencies and percentages, and continuous data were described using means and standard deviations. Sample characteristics were descriptively reported, with independent sample ttests and Chi-squared tests to test the differences between both surveys' sample characteristics. Sample characteristics were also descriptively reported by waterpipe status (lifetime or regular). Two multi-variable logistic regression models assessed the association of the independent variables (age, sex, ethnicity, free school meals, current cigarette use) with each outcome measure (lifetime or regular waterpipe tobacco use). In the model reporting lifetime use, both the 2013 and 2015 samples were combined. In the model reporting regular use, only the 2015 sample was used. Due to the anonymised nature of the data repeat responders could not be identified. These models reported adjusted odds ratios (AORs) and 95% confidence intervals (95% CIs) and used age as a continuous variable rather than categorical. To assess illegal underage use, we measured the prevalence of regular waterpipe tobacco use by smoking location, then a final multi-variable logistic regression model assessed the association of the independent variables with usual waterpipe smoking location being at a shisha bar/ lounge, which for the purposes of this study is our measure of ‘illegal underage use’ given all respondents were under 18 years old.
Results Sample characteristics Both surveys had over 1500 respondents each, from which the pooled mean age was 13.5 ± 1.5 years, 48.0% were male, 83.2% were white, 18.2% reported having free school meals and the prevalence of lifetime and current cigarette use was 20.8% and 6.9%, respectively. Table 1 shows the sample characteristics for each survey. These were broadly similar however after statistical testing, the 2015 sample was found to be significantly older than the 2013 sample (13.3 vs 13.7 years, P < 0.001), have a lower proportion of white respondents
(85.2% vs 81.0%, P < 0.01) and a lower proportion of those reporting free school meals (21.0% vs 15.4%, P < 0.001).
Waterpipe tobacco prevalence Lifetime waterpipe prevalence remained similar in 2013 (13.7%, 95% CI 12.0e15.4%) and 2015 (14.6%, 95% CI 12.8e16.4%). In 2015, regular waterpipe use was measured at 2.9% (95% CI 2.1e3.8%). Table 2 shows sample characteristics by waterpipe tobacco status. The main difference in lifetime waterpipe prevalence between the two surveys was found for sex: while males reported lifetime waterpipe use more than females in 2013 (16.8% vs 10.9%), this appeared more balanced in 2015 (14.9% vs 14.0%). Otherwise, for both lifetime and regular use, waterpipe prevalence increased with age, nonwhite ethnicities and cigarette use. For example, in 2015 regular waterpipe prevalence was 5.5% among those aged 16 years compared with 1.6% among those aged 11 years, and 12.0% among lifetime cigarette users compared with 0.7% among never cigarette users. In adjusted logistic regression models (Table 3), the odds of lifetime waterpipe tobacco use increased by 45% for every year of age (AOR 1.45, 95% CI 1.34e1.59). Odds of lifetime use were also higher for males (AOR 1.48, 95% CI 1.17e1.86) non-White ethnicities (e.g. South Asian vs White: AOR 3.25, 95% CI 2.36e4.47) and current cigarette use (AOR 7.22, 95% CI 5.23e9.96). The odds of regular waterpipe tobacco smoking increased by 40% for every year of age (AOR 1.40, 95% CI 1.09e1.80). Odds of regular use were also higher for current cigarette users (AOR 8.87, 95% CI 4.16e18.91). For both measures of waterpipe use, no association was seen for free school meals.
Table 1 e Sample characteristics. Characteristic
Total Age 11 years 12 years 13 years 14 years 15 years 16 years Sex Female Male Ethnicity White South Asian Other Free school meals No Don't know/don't want to say Yes Lifetime cigarette use No Yes Current cigarette use No Yes
2013
2015
N
%
N
%
1772
100.0
1604
100.0
176 368 433 394 230 146
10.1 21.1 24.8 22.6 13.2 8.4
135 289 263 400 257 233
8.6 18.3 16.7 25.4 16.3 14.8
897 875
50.6 49.4
830 723
53.4 46.6
1490 147 112
85.2 8.4 6.4
1275 168 132
81.0 10.7 8.4
1320 63 368
75.4 3.6 21.0
1268 65 242
80.5 4.1 15.4
1324 375
77.9 22.1
1233 298
30.5 19.5
1576 123
92.8 7.2
1469 104
93.4 6.6
35
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Table 2 e Waterpipe tobacco smoking by sample characteristics. Characteristic
2013
2015
Lifetime waterpipe users
Total Age 11 years 12 years 13 years 14 years 15 years 16 years Sex Female Male Ethnicity White South Asian Other Free school meals No Don't know/don't want to say Yes Lifetime cigarette use No Yes Current cigarette use No Yes
Age Sex Female Male Ethnicity White South Asian Other Free school meals No Don't know/don't want to say Yes Current cigarette use No Yes
Regular waterpipe users
N
%
N
%
N
%
221
13.7
214
14.6
43
2.9
1 13 51 51 66 35
0.7 3.9 13.1 14.3 29.7 25.9
8 15 23 68 50 49
6.5 5.6 9.7 18.5 21.4 22.6
2 2 2 14 10 12
1.6 0.8 0.8 3.8 4.3 5.5
90 131
10.9 16.8
97 109
14.9 14.0
20 20
2.6 3.1
166 41 11
12.1 31.8 11.6
161 29 21
13.7 18.5 18.6
31 5 5
2.6 3.2 4.4
165 12 40
13.6 20.0 12.4
167 8 35
14.3 13.3 16.1
32 3 7
2.7 5.0 3.2
90 115
7.4 33.3
94 109
8.2 38.5
8 34
0.7 12.0
146 59
10.1 52.2
164 45
12.1 45.5
27 16
2.0 16.2
Table 3 e Correlates of regular waterpipe tobacco use (2015 sample only) and lifetime waterpipe tobacco use (2013 and 2015 pooled samples). Characteristic
Lifetime waterpipe users
Regular waterpipe users
Table 4 e Regular waterpipe tobacco use by location of use (2015 sample only). Location
Lifetime waterpipe users
AOR
95% CI
AOR
95% CI
1.40
1.09, 1.80**
1.45
1.34, 1.59***
1.00 1.22
0.63, 2.39
1.00 1.48
1.17, 1.86**
1.00 2.20 2.48
0.79, 6.15 0.88, 6.98
1.00 3.25 1.70
2.36, 4.47*** 1.10, 2.62*
1.00 1.15
0.24, 5.38
1.00 1.10
0.62, 1.94
1.20
0.50, 2.87
0.94
0.69, 1.27
1.00 8.87
4.16, 18.91***
1.00 7.22
5.23, 9.96***
Abbreviations: AOR, adjusted odds ratio; CI, confidence interval. *P < 0.05, **P < 0.01, ***P < 0.001.
Friend's home No Yes Home No Yes Shisha bar/lounge No Yes School No Yes Other No Yes
Regular waterpipe users N
%
18 22
13.3 40.7
31 7
18.8 35.0
26 13
18.2 27.1
16 22
14.0 29.3
28 10
21.1 18.5
However, the prevalence of regular waterpipe tobacco use was the highest in friend's homes (40.7%) and own homes (35.0%) than in shisha bars/lounges (27.1%).
Location of waterpipe tobacco smoking
Underage use
Table 4 shows regular waterpipe tobacco use prevalence, as a proportion of all waterpipe tobacco users, by location of usual use. The prevalence of regular waterpipe tobacco use was higher in all reported locations except the ‘other’ location.
Table 5 presents the correlates of illegal underage shisha bar/ lounge use. For every 1 year increase in age, the odds of illegal underage use increased by 36% (AOR 1.36, 95% CI 1.14e1.62). The odds of illegal underage use were also higher among
36
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Table 5 e Correlates of illegal underage shisha bar/lounge use pooled across both samples. Characteristic
Age Sex Female Male Ethnicity White South Asian Other Free school meals No Don't know/don't want to say Yes Current cigarette use No Yes
Illegal underage shisha bar/lounge use AOR
95% CI
1.36
1.14, 1.62**
1.00 1.20
0.78, 1.84
1.00 2.99 0.83
1.65, 5.42*** 0.35, 1.94
1.00 0.83 0.76
0.29, 2.39 0.43, 1.32
1.00 1.80
1.09, 2.99*
Abbreviations: AOR, adjusted odds ratio; CI, confidence interval. *P < 0.05, **P < 0.01, ***P < 0.001.
South Asians (vs White, AOR 2.99, 95% CI 1.65e5.42) and current cigarette users (AOR 1.80, 95% CI 1.09e2.99).
Discussion Main findings This study among 11- to 16-year-old secondary school students in Stoke-on-Trent showed that nearly one in seven had tried waterpipe tobacco smoking, with no significant change in use between 2013 and 2015, and that 3% were regular users. Waterpipe tobacco use was associated with increasing age, males, non-white ethnicities and current cigarette use, but not the presence of free school meals, our proxy measure for low socioeconomic status. Our correlates of use are largely in line with the reports from the US22 and Europe.4 Regular use was highest in those reporting usual smoking locations being friends' or own homes, although over a quarter of lifetime users reported illegal underage use, which was more apparent among older and South Asian students.
Previous literature To the best of knowledge, there are no published national surveys of waterpipe tobacco prevalence among young people in the UK, however similar school-based surveys in London report a higher lifetime prevalence (24.0e39.6%)13,14 than our sample (13.7e14.6%). This difference may be due to the higher proportion of non-White students in London, who are most likely to use waterpipe tobacco, or the extremely high number of waterpipe-serving premises (approximately 400) in known operation there.20 Online cross-sectional surveys among adults in Great Britain in 2012 and 2013 showed 11.6% had tried waterpipe tobacco, and there was little variation between the two years.17 In the absence of longitudinal data, the existing evidence may therefore suggest waterpipe tobacco
smoking prevalence is plateauing in the UK, but longitudinal studies are needed to confirm or refute this. The 2015 wave of a yearly multiple cross-sectional survey of school students in the US also suggests a plateau in waterpipe tobacco use (at around 7%);23 however, continual rising rates have been documented in several Arab countries24,25 and Sweden.26 Our findings on illegal under age use in waterpipe-serving premises are informative to health policy makers. Studies from the US and UK have shown that educational establishment proximity to waterpipe-serving premises are an independent predictor of waterpipe use.14,27,28 While this suggests that such premises should be a focus of policy intervention, our findings showed that regular use is more highly reported in domestic settings. This corroborates with qualitative research from university students who report smoking in commercial waterpipe premises can be expensive and less accessible than domestic use28,29 and lends itself to the need for a multi-faceted policy approach that goes beyond simply stronger enforcement of waterpipe-serving premises. Illegal underage use in waterpipe-serving premises has been previously identified in London, both from surveys among young people14 and law enforcers.20 In addition to accessibility, affordability and acceptability, reasons for this are likely to include the attraction towards flavoured tobacco,30 the general lack of tobacco legislation compliance from the waterpipe tobacco industry,20 and ritualistic aspects of waterpipe tobacco use.31 It is important to note that waterpipe-serving premises are not homogenousdsome resemble loud club-like environments, whereas others resemble quiet coffee shops.31 Further research should identify which types, if any, young people are particularly attracted to, and interventions to kerb such use.
Implications Given a quarter of lifetime waterpipe users report illegal underage use, this calls into question the degree of tobacco legislation self-enforcement in waterpipe-serving premises. Such premises are required to display a statutory notice explaining that it is illegal to sell tobacco products to under 18s; however, qualitative reports with law enforcers in London suggest waterpipe-serving premises are deliberately and recurrently non-compliant with such regulations.20 Turkey and several Gulf countries have zoning laws in place to prohibit the sale of waterpipe tobacco within a certain distance of educational establishments in a bid to kerb youth waterpipe tobacco smoking.19 Research evaluating tobacco control policy on waterpipe tobacco smoking behaviour are lacking32 and should be considered a research priority in this field. This study has other research implications. Research should explore how the location and frequency of waterpipe tobacco smoking changes over time. Whether students are introduced to waterpipe tobacco at commercial premises then transition to regular waterpipe tobacco smoking at home remains unknown. This has implications for tobacco control policy: an emphasis on enforcing policy on commercial premises may simply be offset by increased home use.29 Better quality epidemiological data concerning prevalence of use and surveillance of waterpipe-serving premises are also needed.
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
Limitations The sampling method may be affected by selection bias; schools whose staff believe tobacco smoking is an issue for them may be more likely to agree to take part in this survey. However, our sample characteristics are broadly representative of the local population.21 Due to the low number of regular waterpipe tobacco users, statistical analysis may be underpowered to detect important associations; this may have been the case looking at the association between ethnicity and regular waterpipe tobacco use, which displayed a high odds ratio (greater than 2) but non-significant associations due to wide confidence intervals. Tobacco use prevalence is selfreported and may be prone to measurement bias (underreporting); however, self-reported tobacco use is known to correlate well with biochemical markers such as serum cotinine.33 Finally, given we could not identify data repeat responders from the anonymised data sets, students may have been double counted for the lifetime pooled analyses shown in Tables 3 and 5; however, we believe this to be minimal.
Conclusions Nearly one in seven students in this sample have tried waterpipe tobacco smoking, of which a quarter report illegal underage use in waterpipe-serving premises. Increased monitoring of waterpipe tobacco prevalence and of waterpipe-serving premises is required with a view to generate more policy evaluation in this public health area.
Author statements
5.
6.
7.
8.
9. 10.
11.
12.
13.
14.
Ethical approval Data from this study were taken from routine health survey data and as such ethical approval was not required.
Funding
15.
16.
None.
Competing interests
17.
None declared.
references
18. 19.
1. Ward KD. The waterpipe: an emerging global epidemic in need of action. Tob Control 2015;24(Suppl. 1):i1e2. 2. Maziak W. The global epidemic of waterpipe smoking. Addict Behav 2011;36:1e5. 3. Maziak W. Commentary: the waterpipeea global epidemic or a passing fad. Int J Epidemiol 2010;39:857e9. 4. Jawad M, Lee JT, Millett C. Waterpipe tobacco smoking prevalence and correlates in 25 Eastern Mediterranean and Eastern European countries: cross-sectional analysis of the
20.
21.
37
Global Youth Tobacco Survey. Nicotine Tob Res 2016;18(4):395e402. Akl EA, Jawad M, Lam WY, Co CN, Obeid R, Irani J. Motives, beliefs and attitudes towards waterpipe tobacco smoking: a systematic review. Harm Reduct J 2013;10:12. Daher N, Saleh R, Jaroudi E, Sheheitli H, Badr T, Sepetdjian E, et al. Comparison of carcinogen, carbon monoxide, and ultrafine particle emissions from narghile waterpipe and cigarette smoking: sidestream smoke measurements and assessment of second-hand smoke emission factors. Atmos Environ 1994;2010(44):8e14. Monzer B, Sepetdjian E, Saliba N, Shihadeh A. Charcoal emissions as a source of CO and carcinogenic PAH in mainstream narghile waterpipe smoke. Food Chem Toxicol 2008;46:2991e5. Sepetdjian E, Shihadeh A, Saliba NA. Measurement of 16 polycyclic aromatic hydrocarbons in narghile waterpipe tobacco smoke. Food Chem Toxicol 2008;46:1582e90. Shihadeh A. Investigation of mainstream smoke aerosol of the argileh water pipe. Food Chem Toxicol 2003;41:143e52. Waziry K, Jawad M, Ballout RA, Al Akel M, Akl EA. The effects of waterpipe tobacco smoking on health outcomes: an updated systematic review and meta-analysis. Int J Epidemiol 2016. http://dx.doi.org/10.1016/10.1093/ije/dyw021 [Epub ahead of print]. Raad D, Gaddam S, Schunemann HJ, Irani J, Jaoude PA, Honeine R, et al. Effects of water-pipe smoking on lung function: a systematic review and meta-analysis. Chest 2011;139:764e74. Jackson D, Aveyard P. Waterpipe smoking in students: prevalence, risk factors, symptoms of addiction, and smoke intake. Evidence from one British University. BMC Public Health 2008;8:174. Jawad M, Power G. Prevalence, correlates and patterns of waterpipe smoking among secondary school students in southeast London: a cross-sectional study. BMC Public Health 2016;16:1. Jawad M, Wilson A, Lee JT, Jawad S, Hamilton FL, Millett C. Prevalence and predictors of water pipe and cigarette smoking among secondary school students in London. Nicotine Tob Res 2013;15:2069e75. Jawad M, Choaie E, Brose L, Dogar O, Grant A, Jenkinson E, et al. Waterpipe tobacco use in the United Kingdom: a crosssectional study among university students and stop smoking practitioners. PloS One 2016;11:e0146799. Jawad M, Abass J, Hariri A, Rajasooriar KG, Salmasi H, Millett C, et al. Waterpipe smoking: prevalence and attitudes among medical students in London. Int J Tuberc Lung Dis 2013;17:137e40. Grant A, Morrison R, Dockrell MJ. Prevalence of waterpipe (shisha, narghille, hookah) use among adults in Great Britain and factors associated with waterpipe use: data from crosssectional Online Surveys in 2012 and 2013. Nicotine Tob Res 2014;16:931e8. Joossens L, Raw M. The Tobacco Control Scale: a new scale to measure country activity. Tob Control 2006;15:247e53. Jawad M, El Kadi L, Mugharbil S, Nakkash R. Waterpipe tobacco smoking legislation and policy enactment: a global analysis. Tob Control 2015;24(Suppl. 1):i60e5. Jawad M. Legislation enforcement of the waterpipe tobacco industry: a qualitative analysis of the London experience. Nicotine Tob Res 2014;16:1000e8. Jawad M, McIver C, Iqbal Z. Prevalence and correlates of lifetime waterpipe, cigarette, alcohol and drug use among secondary school students in Stoke-on-Trent, UK: a post hoc cross-sectional analysis. J Public Health (Oxf) 2014;36:615e21.
38
p u b l i c h e a l t h 1 4 6 ( 2 0 1 7 ) 3 2 e3 8
22. Maziak W, Taleb ZB, Bahelah R, Islam F, Jaber R, Auf R, et al. The global epidemiology of waterpipe smoking. Tob Control 2015;24:i3e12. 23. Singh T, Arrazola RA, Corey CG, Husten CG, Neff LJ, Homa DM, et al. Tobacco use among middle and high school students e United States, 2011e2015. MMWR Morb Mortal Wkly Rep 2016;65:361e7. 24. McKelvey KL, Wilcox ML, Madhivanan P, Mzayek F, Khader YS, Maziak W. Time trends of cigarette and waterpipe smoking among a cohort of school children in Irbid, Jordan, 2008e11. Eur J Public Health 2013;23:862e7. 25. Saade G, Warren CW, Jones NR, Asma S, Mokdad A. Linking Global Youth Tobacco Survey (GYTS) data to the WHO Framework Convention on Tobacco Control (FCTC): the case for Lebanon. Prev Med 2008;47(Suppl. 1):S15e9. 26. Galanti MR, Al-Adhami M. Use of a water pipe is not an alternative to other tobacco or substance use among adolescents: results from a national survey in Sweden. Nicotine Tob Res 2015;17:74e80. 27. Sutfin EL, McCoy TP, Reboussin BA, Wagoner KG, Spangler J, Wolfson M. Prevalence and correlates of waterpipe tobacco smoking by college students in North Carolina. Drug Alcohol Depend 2011;115:131e6.
28. Kassem NO, Jackson SR, Boman-Davis M, Kassem NO, Liles S, Daffa RM, et al. Hookah smoking and facilitators/barriers to lounge use among students at a US University. Am J Health Behav 2015;39:832e48. 29. Jawad M, Jawad S, Mehdi A, Sardar A, Jawad AM, Hamilton FL. A qualitative analysis among regular waterpipe tobacco smokers in London universities. Int J Tuberc Lung Dis 2013;17:1364e9. 30. Salloum RG, Maziak W, Hammond D, Nakkash R, Islam F, Cheng X, et al. Eliciting preferences for waterpipe tobacco smoking using a discrete choice experiment: implications for product regulation. BMJ open 2015;5:e009497. 31. Carroll MV, Chang J, Sidani JE, Barnett TE, Soule E, Balbach E, et al. Reigniting tobacco ritual: waterpipe tobacco smoking establishment culture in the United States. Nicotine Tob Res 2014;16:1549e58. 32. Jawad M, Jawad S, Waziry RK, Ballout RA, Akl EA. Interventions for waterpipe tobacco smoking prevention and cessation: a systematic review. Sci Rep 2016;6:25872. € la € T, Lillsunde P, Puska P. Validation of self 33. Vartiainen E, Seppa reported smoking by serum cotinine measurement in a community-based study. J Epidemiol Community Health 2002;56:167e70.