Abstract
Background
Unlike cigarettes, there is little information about the association between other tobacco products and the risk of gastroesophageal reflux disease (GERD) and esophageal adenocarcinoma.
Methods
We used the baseline data from the Pars Cohort Study conducted in southern Iran. In 2012, 9264 local residents between 40 and 75 years old were enrolled, with detailed information about lifestyle, including duration and frequency of tobacco use. GERD was defined based on questions assessing heartburn and regurgitation in the past 12 months, frequency and severity. Associations were calculated by logistic regression models adjusted for age, sex, education, cigarettes and body mass index.
Results
In the study, 25.4% of the participants had severe GERD (interfering with participants’ routines), 25.1% had frequent GERD (at least once a week) and 11.2% had both severe and frequent GERD, all more common among women (p < 0.001); 45.6% of women and 28.3% of men smoked waterpipes. Among people not using medications against reflux symptoms, there was an association between waterpipe smoking and severe [odds ratio (OR) = 1.18; 95% confidence interval (CI):1.04–1.35], frequent (OR = 1.16; 95% CI: 1.02–1.32) and severe and frequent reflux (OR = 1.30; 95% CI: 1.08–1.56). The associations increased with the duration of use, intensity and cumulative waterpipe-years, reaching an OR of 1.44 (95% CI: 1.12–1.86) for severe and frequent reflux in those who had smoked more than 48 waterpipe-years. There was effect modification by sex, and all the associations were only seen among women.
Conclusion
The increasing trend in the association between cumulative waterpipe use and reflux disease among women is particularly important given the growing waterpipe tobacco epidemic in many populations.
Keywords: gastroesophageal reflux disease, esophageal adenocarcinoma, tobacco, waterpipe
Key Messages
Waterpipe smoking is becoming more prevalent in many parts of the world, particularly among young people.
There is little information about the association between waterpipe smoking and gastroesophageal reflux disease (GERD).
In this study, GERD symptoms were more common among women and were associated with waterpipe use.
Among people who did not use anti-reflux medications, longer duration, higher intensity, and more cumulative use of waterpipe (waterpipe-years) were associated with increased chances of developing severe and/or frequent reflux symptoms.
There was effect modification by sex, and the associations were only seen among women.
Introduction
Gastroesophageal reflux disease (GERD) is the main risk factor for esophageal adenocarcinoma (EAC), the histological type of esophageal cancer which has been on the rise in the USA and many European countries.1 GERD itself imposes substantial direct and indirect costs on the health care system, and has important effects on quality of life.2,3 GERD is often diagnosed by its symptoms of heartburn and regurgitation. In 2006, in an effort to develop a global definition of GERD, the International Consensus Group described GERD as ‘a condition which develops when the reflux of stomach contents causes troublesome symptoms and/or complications’.4 The prevalence of GERD varies in different parts of the world, and ranges from an average of 19.8% in North America to 15.2% in Europe and 5.2% in East Asia.5 In the Middle East, the average prevalence of GERD is about 14.4%, which is less common than in most Western populations, but more common than in East Asia.5 Although there are not enough data to understand its trends over time in this region, some studies suggest an increase in recent years like many Western countries.6 In most studies, obesity and smoking have been implicated as the major lifestyle risk factors associated with GERD.5,7,8 Both factors are also associated with increased risk of esophageal adenocarcinoma (EAC).9 Although cigarette smoking has been widely studied in association with the risk of GERD and esophageal adenocarcinoma, there is little information about the association of GERD with other types of tobacco-smoking products, such as waterpipes.
A waterpipe, which is also called hookah, hubbly bubbly, shisha, narghile or qualyan, is used to smoke tobacco. The smoke produced by inhaling through a hose attached to the device passes through water before being inhaled. It is estimated that about 100 million people around the world use a form of waterpipe,10 but this number seems to be rising in many parts of the world, particularly among young people.11,12 A few previous studies have shown health hazards associated with waterpipe smoking, including cancer,13 respiratory diseases14 and early mortality.15 In one study, Islami et al. showed a positive association between waterpipe smoking and severe reflux, but this study included only 533 waterpipe smokers, and was underpowered to further analyse this association.16 Thus, we directed our attention towards a population where both GERD and waterpipe smoking were common. A study of migrating Qashqai Nomads living in Fars Province in south-central Iran found the highest prevalence of weekly GERD symptoms (33.1%) reported in a single population-based study anywhere in the world.17 The south-west of this Province is also the site of Pars Cohort Study (PCS),18 which provides a unique opportunity to study the association between waterpipe smoking and GERD, as more than 35% of the individuals in this study have reported using a waterpipe regularly.
Methods
The PCS is being conducted in the Fars Province of southern Iran to assess risk factors for non-communicable diseases in a population undergoing socio-economic and epidemiologic transition. The recruitment was done in 2012 by inviting all residents of Valashahr district, in the south-west of the Province, who were between 40 and 75 years old. The study protocol and the informed consent used for this investigation were approved by the ethical review committees of the Digestive Disease Research Institute of Tehran University of Medical Sciences and Shiraz University of Medical Sciences.
Out of 9721 eligible invitees, 9264 individuals (95.3%) agreed to participate. They completed an informed consent, were interviewed by the trained cohort staff in the Pars Cohort Study Center (PCSC) and provided biological samples. Active follow-up is planned to continue for 10 years, with complete recording of cause of death, hospitalizations and occurrence of major diseases. Details of PCS recruitment and follow-up procedures have been published before.18 In the current analysis, we are only using the data collected at baseline.
Exposure assessment
A trained nurse completed anthropometric measurements (height, weight, hip and waist circumference) according to a standard protocol. Body mass index (BMI) was calculated as weight in kilograms divided by squared height in metres, and classified according to WHO classification. Waist circumference was measured at the level of the umbilicus, and divided into sex-specific quartiles. The general questionnaire included demographic characteristics, education, history of current and past medications, and detailed questions about lifestyle such as tobacco and alcohol use. Almost all the tobacco used in this population was either cigarettes or waterpipes (13 people also chewed tobacco). Data were collected about the lifetime use of tobacco, the ages of starting and stopping, daily consumption amount and frequency of use. If the participant smoked both cigarettes and waterpipes, or used them intermittently, data were recorded separately for each type and period of use.
Ever waterpipe users were those who smoked a waterpipe at least once a week for 6 months. Former users were defined as people who stopped at least 5 years before entering the study; otherwise they were considered current users. Average intensity of waterpipe use (times per day) was calculated over the individuals’ lifetime reported use and categorized into three groups (less than once a day, once or twice a day and more than twice a day). The lifetime duration of waterpipe use was calculated as the number of years between the first reported use and the time individuals quit (or their current age if they still smoke) and categorized into tertiles (10 years or less, 11–25 years and more than 25 years). For cigarette smoking, we used the same definitions for ever, former and current smoking, intensity and duration. Cumulative cigarette use was calculated in the form of pack-years across all periods of cigarette use, e.g. a person who smoked three packs of cigarettes per day for 5 years and one who smoked half a pack for 30 years both smoked 15 pack-years on average. Cumulative waterpipe use (waterpipe-years) was calculated in a similar way by multiplying duration of use by average number of times per day during each period of use and was summed over the periods as described before.15 Pack-years and waterpipe-years were categorized into tertiles among users (10 or fewer, 11–28 and more than 28 for pack-years; 10 or fewer, 11–48 and more than 48 for waterpipe-years). In 14 women (5 with reflux) and 13 men (7 with reflux), waterpipe intensity and waterpipe-years could not be calculated. Finally, never tobacco users were participants who had never regularly used waterpipe, cigarette or any other type of tobacco (e.g. smokeless tobacco) regularly in their lifetime. This group served as the universal comparison group for all the models described in the statistical analysis.
The questionnaire used in this study has been validated in another study from Iran—the Golestan Cohort Study, in a rural area with lower average education compared with the population in the present study.15 In the validation study, 95% of 40 self-reported current waterpipe smokers had high urinary cotinine levels, similar to those seen among cigarette smokers (20 ng/ml and above; unpublished data).
Definition of GERD
Two sets of questions (a shortened version based on a previously validated questionnaire19) were used to define symptoms of GERD (i.e. reflux symptoms). The first set asked about heartburn and the second set asked about regurgitation. The first question in each set asked ‘Have you experienced [the symptom] in the past 12 months?’. If the answer was yes, the second question asked about the frequency of the symptoms (1–10 times in 12 months, almost once a month, 3–4 times a month, at least once a week, almost daily). We combined the last two categories to define a weekly symptom. The third question assessed symptom severity: mild (‘I don’t feel it unless I think about it’), moderate (‘I feel it, but it does not interfere with my routines’), severe (‘it interferes with my routines’) or very severe (‘it seriously affects my life’). We combined the last two categories to define a severe symptom. For the current analysis, participants who did not report either symptom were classified as ‘no reflux’, whereas ‘any reflux’ was defined as having at least one symptom (heartburn or regurgitation) with any severity or frequency during the past 12 months. If at least one symptom was present weekly, the person was categorized as having ‘frequent reflux’ and, if at least one symptom was perceived as severe or very severe, it was defined as ‘severe reflux’. There was some overlap between these two categories, and individuals who reported having both frequent and severe symptoms were categorized as having ‘severe and frequent reflux’. Current use of any medication that can be used to reduce reflux symptoms was extracted from the medication history (‘anti-reflux medication’). These included antacids (e.g. aluminium/magnesium hydroxide), H2 blockers (e.g. ranitidine) and proton-pump inhibitors (e.g. omeprazole).
Statistical analysis
Categorical variables are presented as numbers and percentages. The chi-square test was used to test the crude associations between these variables and reflux symptoms. We used logistic regression models to calculate the adjusted odds ratios (ORs) and 95% confidence intervals (CIs) of GERD with waterpipe use. Since the use of anti-reflux medication could alter the symptoms, we excluded people reporting such use (n = 783) from the models. For comparison, the results for this group is presented in Supplementary Table 1 (available as Supplementary Data at IJE online). In each model, the presence of one category of reflux symptoms (any/severe/frequent/severe and frequent) vs no reflux served as the dependent variable. The main independent variables were waterpipe use (ever/former/current), average intensity of waterpipe smoking (times per day), lifetime duration of waterpipe use (in years) and cumulative waterpipe use (waterpipe-years). In all models, the reference category was never tobacco users. The models were adjusted for age, sex, education, cigarette smoking (never/former/current) and measured BMI. Alcohol use was uncommon in this population and, since adjustment for alcohol use did not affect the estimates, it was not included in the final models. P-values for trend were calculated in the same models by assigning the median values of each tobacco use category to the individuals in that category of use.
We stratified all analyses by sex, and also analysed the association between cigarette smoking and reflux symptoms only in men (since there were only 35 women who smoked cigarettes). The 5-year cut-off to define past tobacco use was adopted to minimize reverse causation; however, as a sensitivity analysis, we changed this cut-off to those who quit more than 1 year before study entry, and the results were similar.
All statistical analyses were conducted using STATA statistical software, version13 (StataCorp Inc., College Station, TX, USA).
Results
Among 9264 participants (4988 women and 4276 men), 5420 (58.5%) reported experiencing some reflux symptom (i.e. ‘any reflux’). Severe reflux was present in 2357 (25.4% of participants) and frequent symptoms (i.e. at least once a week) in 2325 (25.1% of participants). Shared between these two categories were 1041 people (11.2% of participants) whose reflux was both severe and frequent. All levels of reflux symptoms were more common among women compared with men (63.9% vs 52.1% any reflux, 28.7% vs 21.6% severe reflux, 29.7% vs 19.7% frequent reflux and 13.5% vs 8.7% severe and frequent reflux; all p < 0.001). Regurgitation was a more common symptom than heartburn in both sexes, and people with severe or frequent symptoms were more likely to report both symptoms (Table 1).
Table 1.
Baseline characteristics of the PCS population by reflux symptoms
| Women | Men | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| No reflux (n = 1798) | Any reflux (n = 3190) | Severe reflux (n = 1432) | Frequent reflux (n = 1482) | No reflux (n = 2046) | Any reflux (n = 2230) | Severe reflux (n = 925) | Frequent reflux (n = 843) | ||
| Agea | 51.6 (9.3) | 53.1 (9.6) | 53.4 (9.6) | 53.8 (9.9) | 52.8 (9.8) | 52.7 (10.0) | 53.2 (10.3) | 53.5 (10.3) | |
| Ethnicity | Qashqai | 717 (39.9) | 1182 (37.1) | 514 (35.9) | 561 (37.9) | 826 (40.4) | 871 (39.1) | 369 (39.9) | 334 (39.6) |
| Other | 1081 (60.1) | 2008 (62.9) | 918 (64.1) | 921 (62.1) | 1220 (59.6) | 1359 (60.9) | 556 (60.1) | 509 (60.4) | |
| Education | None | 1119 (62.2) | 2085 (65.4) | 951 (66.4) | 998 (67.3) | 657 (32.1) | 682 (30.6) | 291 (31.5) | 308 (36.5) |
| Elementary | 573 (31.9) | 971 (30.4) | 429 (30) | 424 (28.6) | 588 (28.7) | 601 (27) | 264 (28.5) | 229 (27.2) | |
| 6–8 years | 67 (3.7) | 96 (3) | 41 (2.9) | 42 (2.8) | 348 (17.1) | 463 (20.8) | 186 (20.1) | 153 (18.1) | |
| High school | 28 (1.6) | 31 (1) | 10 (0.7) | 15 (1.0) | 346 (16.9) | 328 (14.7) | 124 (13.4) | 112 (13.3) | |
| University | 11 (0.6) | 7 (0.2) | 1 (0.1) | 3 (0.2) | 107 (5.2) | 156 (7) | 60 (6.5) | 41 (4.9) | |
| BMI | <18 | 38 (2.1) | 78 (2.4) | 37 (2.6) | 32 (2.2) | 131 (6.4) | 144 (6.5) | 62 (6.7) | 62 (7.4) |
| 18–25 | 608 (33.8) | 944 (29.6) | 418 (29.2) | 431 (29.1) | 1064 (52.0) | 1109 (49.7) | 467 (50.5) | 436 (51.7) | |
| 25–29 | 692 (38.5) | 1329 (41.7) | 598 (41.8) | 640 (43.2) | 669 (32.7) | 760 (34.1) | 316 (34.2) | 269 (31.9) | |
| ≥30 | 460 (25.6) | 839 (26.3) | 379 (26.5) | 379 (25.6) | 182 (8.9) | 217 (9.7) | 80 (8.6) | 76 (9.0) | |
| Waist circumference | Q1 | 506 (28.2) | 761 (23.9) | 348 (24.3) | 332 (22.4) | 550 (26.9) | 548 (24.6) | 222 (24.0) | 220 (26.1) |
| Q2 | 463 (25.7) | 805 (25.2) | 342 (23.9) | 381 (25.7) | 545 (26.6) | 572 (25.7) | 248 (26.8) | 234 (27.8) | |
| Q3 | 423 (23.5) | 814 (25.5) | 360 (25.1) | 397 (26.8) | 480 (23.4) | 570 (25.6) | 230 (24.9) | 193 (22.9) | |
| Q4 | 406 (22.6) | 810 (25.4) | 382 (26.7) | 372 (25.1) | 471 (23.1) | 540 (24.2) | 225 (24.3) | 196 (23.3) | |
| Alcohol use (ever) | 9 (0.5) | 21 (0.7) | 8 (0.6) | 10 (0.7) | 71 (3.5) | 95 (4.3) | 41 (4.4) | 38 (4.5) | |
| Cigarette pack-yearsb | Never | 1790 (99.5) | 3173 (99.5) | 1423 (99.4) | 1474 (99.5) | 1186 (57.8) | 1238 (55.5) | 528 (57.1) | 453 (53.7) |
| ≤10 | 6 (0.3) | 12 (0.4) | 7 (0.5) | 5 (0.3) | 292 (14.3) | 321 (14.4) | 104 (11.2) | 117 (13.9) | |
| 11–28 | 1 (0.1) | 3 (0.1) | 1 (0.1) | 3 (0.2) | 274 (13.4) | 343 (15.4) | 154 (16.7) | 133 (15.8) | |
| >28 | 1 (0.1) | 2 (0.1) | 1 (0.1) | 0 (0.0) | 294 (14.4) | 328 (14.7) | 139 (15.0) | 140 (16.6) | |
| Waterpipe-yearsc | Never | 1018 (56.9) | 1697 (53.3) | 725 (50.6) | 757 (51.2) | 1476 (72.3) | 1593 (71.6) | 674 (73.0) | 573 (68.0) |
| ≤10 | 305 (17) | 504 (15.8) | 227 (15.9) | 232 (15.7) | 191 (9.4) | 199 (9) | 72 (7.8) | 84 (10.1) | |
| 11–48 | 265 (14.8) | 513 (16.1) | 252 (17.6) | 249 (16.8) | 160 (7.8) | 185 (8.3) | 70 (7.6) | 82 (9.8) | |
| >48 | 202 (11.3) | 471 (14.8) | 228 (15.9) | 241 (16.3) | 214 (10.5) | 246 (11.1) | 107 (11.6) | 101 (12.1) | |
| Anti-reflux drugs (yes) | 56 (3.1) | 438 (13.7) | 291 (20.3) | 290 (19.6) | 52 (2.5) | 237 (10.6) | 163 (17.6) | 167 (19.8) | |
| Type of symptoms | Heartburn | – | 372 (11.7) | 86 (6.0) | 156 (10.5) | – | 325 (14.6) | 69 (7.5) | 108 (12.8) |
| Regurgitation | – | 1376 (43.1) | 491 (34.3) | 446 (30.1) | – | 1085 (48.7) | 374 (40.4) | 300 (35.6) | |
| Both | – | 1442 (45.2) | 855 (59.7) | 880 (59.4) | – | 820 (36.8) | 482 (52.1) | 435 (51.6) | |
BMI, body mass index; SMR, standardized mortality rate; CI, confidence interval.
aNumbers show frequencies (percentage) except for age, which is mean (SD). bIn 9 men (3 with reflux), cigarette intensity and pack-years could not be calculated. cIn 14 women (5 with reflux) and 13 men (7 with reflux), waterpipe intensity and waterpipe-years could not be calculated.
In both sexes, reflux symptoms were associated with a marked increase in the use of anti-reflux medication, particularly when the symptoms were severe or frequent (p < 0.001). In women, there were larger differences in demographic and lifestyle characteristics between participants with and without reflux (Table 1). Women with all levels of reflux symptoms (any/severe/frequent) were older, less educated, were less likely to have normal BMIs or waist circumferences, and smoked more waterpipes. Among men, differences were more subtle, although men with severe or frequent reflux were slightly older and more likely to smoke cigarettes than the men without reflux.
Overall, 3482 people (37.6%) reported that they smoked waterpipes. After excluding participants who used anti-reflux medications, we observed associations between ever waterpipe smoking and severe or frequent reflux symptoms in the adjusted models, although this association weakened when people with milder or less frequent symptoms (i.e. any reflux) were included (Table 2). The strongest association was in the ‘severe and frequent reflux’ group (OR = 1.30; 95% CI: 1.08–1.56). Both former and current waterpipe smoking were associated with severe and/or frequent reflux: former use had a stronger association with severe reflux and current use with frequent reflux. Associations strengthened with longer duration of smoking waterpipes and reached a 58% higher prevalence of severe and frequent reflux in those who had smoked waterpipes for more than 25 years. The associations also increased with more intense waterpipe smoking, but this increase was not linear in the severe and ‘severe and frequent’ groups. There was an increasing trend between more cumulative waterpipe use and reflux symptoms. Whereas there was little evidence for an association between waterpipe use and reflux symptoms among individuals who had smoked fewer than 10 waterpipe-years, the OR increased with more waterpipe-years (p for trend <0.01). Those who had smoked more than 48 waterpipe-years had increased odds of having any (including mild and infrequent) reflux symptom (OR = 1.19; 95% CI: 1.02–1.38). This OR was 1.30 (95% CI: 1.08–1.57) for severe reflux, 1.25 (95% CI: 1.03–1.50) for frequent reflux and 1.44 (95% CI: 1.12–1.86) for severe and frequent reflux (Table 2).
Table 2.
Adjusted association between waterpipe use and different levels of reflux symptoms among all participants who did not report using anti-reflux medication in the PCS
| Numbera | Any reflux OR (95% CI)b | Severe reflux OR (95% CI)b | Frequent reflux OR (95% CI)b | Severe and frequent reflux OR (95% CI)b | |
|---|---|---|---|---|---|
| Never tobacco users | 4053 | 1 | 1 | 1 | 1 |
| Ever waterpipe usersc | 3151 | 1.09 (0.98, 1.21) | 1.18 (1.04, 1.35)* | 1.16 (1.02, 1.32)* | 1.30 (1.08, 1.56)** |
| Former waterpipe usersc | 979 | 1.14 (0.97, 1.33) | 1.29 (1.06, 1.56)* | 1.10 (0.90, 1.34) | 1.37 (1.05, 1.79)* |
| Current waterpipe usersc | 2172 | 1.07 (0.96, 1.20) | 1.14 (0.99, 1.32)* | 1.18 (1.03, 1.36)* | 1.27 (1.04, 1.55)* |
| Waterpipe use duration (years)c | |||||
| ≤10 | 1229 | 0.96 (0.84, 1.10) | 0.94 (0.79, 1.13) | 1.07 (0.89, 1.27) | 1.06 (0.83, 1.37) |
| 11–25 | 920 | 1.13 (0.97, 1.31) | 1.32 (1.09, 1.59)* | 1.13 (0.93, 1.37) | 1.34 (1.03, 1.75)* |
| >25 | 1001 | 1.23 (1.05, 1.42)* | 1.37 (1.14, 1.65)** | 1.31 (1.09, 1.58)* | 1.58 (1.22, 2.03)** |
| P for trend | 0.008 | <0.001 | 0.008 | <0.001 | |
| Waterpipe intensity (average times per day)c,d | |||||
| <1 | 1106 | 1.05 (0.91, 1.20) | 1.10 (0.92, 1.32) | 1.09 (0.91, 1.30) | 1.23 (0.96, 1.57) |
| 1–2 | 928 | 1.09 (0.94, 1.27) | 1.42 (1.18, 1.71)* | 1.09 (0.90, 1.33) | 1.49 (1.15, 1.92)** |
| >2 | 1090 | 1.16 (1.00, 1.34)* | 1.09 (0.90, 1.32) | 1.32 (1.10, 1.59)** | 1.25 (0.96, 1.62) |
| P for trend | 0.08 | 0.04 | 0.007 | 0.02 | |
| Cumulative waterpipe-yearsc,d | |||||
| ≤10 | 1099 | 1.01 (0.87, 1.16) | 1.05 (0.87, 1.25) | 1.05 (0.87, 1.26) | 1.12 (0.86, 1.44) |
| 11–48 | 1006 | 1.11 (0.96, 1.29) | 1.27 (1.05, 1.52)* | 1.21 (1.01, 1.45)* | 1.40 (1.09, 1.80)** |
| >48 | 1019 | 1.19 (1.02, 1.38)* | 1.30 (1.08, 1.57)* | 1.25 (1.03, 1.50)* | 1.44 (1.12, 1.86)** |
| P for trend | 0.06 | 0.01 | 0.03 | 0.008 | |
OR, odds ratio; CI, confidence interval; *p < 0.05; **p < 0.01.
aExcluding people who used any type of medication against reflux. bLogistic regression models adjusted for sex, age, education, BMI and cigarette smoking. cThe reference category for all models are the same 4035 ‘never tobacco users’. dIn 14 women (5 with reflux) and 13 men (7 with reflux), waterpipe intensity and waterpipe-years could not be calculated.
Waterpipe smoking was more common in women (45.6% vs 28.3%). There was an interaction between sex and waterpipe use in all levels of GERD symptoms (p < 0.001). As Tables 3 and 4 show, when stratified by sex, the association between waterpipe smoking and reflux was only seen among women. Ever, former and current waterpipe smoking were associated with severe and/or frequent reflux in women with a pattern similar to the overall results. The increases in reflux prevalence associated with waterpipe use duration and intensity were also seen among women. As with the overall results, there was no association among women smoking fewer than 10 waterpipe-years, but the OR increased with more cumulative use and reached 1.61 (95% CI: 1.18–2.19) for severe and frequent reflux among women who had smoked over 48 waterpipe-years. In men, there was no association between waterpipe use and reflux symptoms. We observed associations between current cigarette smoking and reflux symptoms (OR = 1.20 for any reflux, OR = 1.22 for severe reflux and OR = 1.26 for frequent reflux) among men. These associations increased to some extent with the intensity, duration and pack-years of cigarette smoking, although these increases were not linear (Table 4). In men, we also analysed reflux associations with waterpipe use, stratified by cigarette smoking, and the results were similar in cigarette smokers and non-smokers (Supplementary Table 2, available as Supplementary Data at IJE online).
Table 3.
Adjusted association between smoking waterpipe and different levels of reflux symptoms among women who did not report using anti-reflux medication in the PCS
| Numbera | Any reflux OR (95% CI)b | Severe reflux OR (95% CI)b | Frequent reflux OR (95% CI)b | Severe and frequent reflux OR (95% CI)b | |
|---|---|---|---|---|---|
| Never tobacco users | 2434 | 1 | 1 | 1 | 1 |
| Ever waterpipe usersc | 2039 | 1.09 (0.97, 1.24) | 1.24 (1.06, 1.44)* | 1.17 (1.01, 1.36)* | 1.40 (1.14, 1.72)** |
| Former waterpipe usersc | 574 | 1.15 (0.94, 1.40) | 1.45 (1.15, 1.83)* | 1.03 (0.81, 1.32) | 1.41 (1.02, 1.93)* |
| Current waterpipe usersc | 1465 | 1.07 (0.94, 1.23) | 1.16 (0.98, 1.38) | 1.22 (1.04, 1.44)* | 1.40 (1.12, 1.75)** |
| Waterpipe use duration (years)c | |||||
| ≤10 | 791 | 0.96 (0.81, 1.13) | 0.97 (0.79, 1.20) | 1.03 (0.84, 1.26) | 1.12 (0.84, 1.49) |
| 11–25 | 611 | 1.13 (0.94, 1.36) | 1.42 (1.14, 1.78)* | 1.12 (0.89, 1.41) | 1.48 (1.09, 2.00)* |
| >25 | 636 | 1.29 (1.07, 1.57)* | 1.46 (1.16, 1.85)* | 1.47 (1.17, 1.85)** | 1.77 (1.31, 2.39)** |
| P for trend | 0.004 | <0.001 | 0.001 | <0.001 | |
| Waterpipe intensity (average times per day)c,d | |||||
| ≤1 | 828 | 1.06 (0.90, 1.25) | 1.20 (0.98, 1.47) | 1.10 (0.90, 1.35) | 1.42 (1.09, 1.86)* |
| 1–2 | 624 | 1.10 (0.92, 1.33) | 1.44 (1.16, 1.79)** | 1.13 (0.90, 1.41) | 1.53 (1.14, 2.06)** |
| >2 | 573 | 1.17 (0.97, 1.43) | 1.12 (0.87, 1.43) | 1.36 (1.08, 1.72)** | 1.28 (0.92, 1.77) |
| P for trend | 0.07 | 0.02 | 0.008 | 0.01 | |
| Cumulative waterpipe-yearsc,d | |||||
| ≤10 | 737 | 1.01 (0.85, 1.20) | 1.08 (0.88, 1.34) | 1.05 (0.85, 1.29) | 1.23 (0.92, 1.64) |
| 11–48 | 684 | 1.10 (0.92, 1.31) | 1.34 (1.08, 1.66)* | 1.17 (0.94, 1.45) | 1.49 (1.12, 1.99)** |
| >48 | 604 | 1.25 (1.03, 1.53)* | 1.40 (1.10, 1.77)* | 1.39 (1.11, 1.76)** | 1.61 (1.18, 2.19)** |
| P for trend | 0.01 | 0.004 | 0.004 | 0.003 | |
OR, odds ratio; CI, confidence interval; *p < 0.05; **p < 0.01.
aExcluding women who used any type of medication against reflux. bLogistic regression models adjusted for sex, age, education, BMI and cigarette smoking. cThe reference category for all models are the same 2434 ‘never tobacco users’. dIn 14 women (5 with reflux), waterpipe intensity and waterpipe-years could not be calculated.
Table 4.
Adjusted association between smoking waterpipe and cigarette and different levels of reflux symptoms among men who did not report using anti-reflux medication in the Pars Cohort Study
| Numbera | Any reflux OR (95% CI)b | Severe reflux OR (95% CI)b | Frequent reflux OR (95% CI)b | Severe and frequent reflux OR (95% CI)b | |
|---|---|---|---|---|---|
| Never tobacco users | 1619 | 1 | 1 | 1 | 1 |
| Ever waterpipe usersc | 1112 | 1.05 (0.87, 1.27) | 1.04 (0.81, 1.33) | 1.09 (0.84, 1.42) | 0.95 (0.64, 1.43) |
| Former waterpipe usersc | 405 | 1.08 (0.83, 1.40) | 0.96 (0.67, 1.38) | 1.14 (0.80, 1.63) | 1.13 (0.67, 1.93) |
| Current waterpipe usersc | 707 | 1.05 (0.86, 1.28) | 1.07 (0.82, 1.40) | 1.07 (0.81, 1.42) | 0.88 (0.56, 1.37) |
| Waterpipe use duration (years)c | |||||
| ≤10 | 438 | 0.95 (0.74, 1.24) | 0.87 (0.61, 1.25) | 1.16 (0.81, 1.65) | 0.82 (0.46, 1.45) |
| 11–25 | 309 | 1.11 (0.85, 1.44) | 1.06 (0.74, 1.51) | 1.14 (0.79, 1.63) | 0.90 (0.50, 1.61) |
| >25 | 365 | 1.09 (0.85, 1.40) | 1.17 (0.84, 1.62) | 1.00 (0.71, 1.42) | 1.10 (0.66, 1.82) |
| P for trend | 0.7 | 0.4 | 0.7 | 0.8 | |
| Waterpipe intensity (average times per day)c,d | |||||
| ≤1 | 278 | 0.99 (0.75, 1.31) | 0.85 (0.57, 1.26) | 1.00 (0.68, 1.48) | 0.49 (0.23, 1.04) |
| 1–2 | 304 | 1.02 (0.77, 1.33) | 1.31 (0.93, 1.85) | 0.95 (0.64, 1.40) | 1.20 (0.70, 2.08) |
| >2 | 517 | 1.11 (0.88, 1.39) | 0.99 (0.73, 1.35) | 1.21 (0.89, 1.65) | 1.05 (0.66, 1.68) |
| P for trend | 0.7 | 0.8 | 0.4 | 0.6 | |
| Cumulative waterpipe-yearsc,d | |||||
| ≤10 | 362 | 0.98 (0.75, 1.28) | 0.95 (0.66, 1.37) | 1.05 (0.72, 1.53) | 0.70 (0.38, 1.31) |
| 11–48 | 322 | 1.10 (0.85, 1.44) | 1.04 (0.72, 1.50) | 1.27 (0.88, 1.82) | 1.00 (0.56, 1.78) |
| >48 | 415 | 1.08 (0.85, 1.36) | 1.11 (0.81, 1.52) | 0.99 (0.71, 1.37) | 1.04 (0.64, 1.69) |
| P for trend | 0.9 | 0.7 | 0.6 | 0.8 | |
| Ever cigarette smokersc | 1736 | 1.14 (0.98, 1.32) | 1.14 (0.93, 1.38) | 1.21 (0.98, 1.49) | 1.23 (0.90, 1.69) |
| Former cigarette smokersc | 412 | 0.93 (0.74, 1.19) | 0.80 (0.57, 1.12) | 1.00 (0.72, 1.40) | 1.10 (0.68, 1.79) |
| Current cigarette smokersc | 1324 | 1.20 (1.02, 1.40)* | 1.22 (1.00, 1.50)* | 1.26 (1.01, 1.57)* | 1.27 (0.92, 1.76) |
| Cigarette smoking duration (years)c | |||||
| ≤20 | 645 | 1.00 (0.82, 1.21) | 0.92 (0.70, 1.21) | 1.10 (0.83, 1.45) | 0.91 (0.58, 1.41) |
| 21–30 | 511 | 1.40 (1.13, 1.73)* | 1.48 (1.12, 1.94)* | 1.47 (1.10, 1.96)** | 1.73 (1.15, 2.61)** |
| >30 | 572 | 1.09 (0.89, 1.33) | 1.11 (0.85, 1.45) | 1.13 (0.85, 1.51) | 1.21 (0.80, 1.82) |
| P for trend | 0.02 | 0.02 | 0.04 | 0.02 | |
| Smoking intensity (average cigarettes per day)c,e | |||||
| <10 | 384 | 1.09 (0.87, 1.38) | 0.87 (0.62, 1.21) | 1.05 (0.75, 1.46) | 0.81 (0.47, 1.41) |
| 10–19 | 482 | 1.17 (0.94, 1.44) | 1.22 (0.92, 1.61) | 1.14 (0.85, 1.54) | 1.30 (0.84, 1.99) |
| ≥20 | 861 | 1.15 (0.96, 1.37) | 1.22 (0.96, 1.55) | 1.33 (1.03, 1.70)* | 1.40 (0.98, 2.01) |
| P for trend | 0.08 | 0.02 | 0.02 | 0.02 | |
| Cigarette pack-yearsc,e | |||||
| ≤10 | 567 | 1.09 (0.89, 1.34) | 0.89 (0.67, 1.19) | 1.11 (0.83, 1.49) | 0.84 (0.52, 1.36) |
| 11–28 | 582 | 1.24 (1.02, 1.52)* | 1.44 (1.11, 1.86)* | 1.34 (1.01, 1.77)* | 1.52 (1.02, 2.26)* |
| >28 | 578 | 1.11 (0.90, 1.35) | 1.12 (0.86, 1.47) | 1.22 (0.92, 1.61) | 1.36 (0.91, 2.02) |
| P for trend | 0.1 | 0.03 | 0.05 | 0.01 | |
OR, odds ratio; CI, confidence interval; *p < 0.05; **p < 0.01.
aExcluding men who used any type of medication against reflux. bLogistic regression models adjusted for sex, age, education, BMI and cigarette smoking. Models for cigarette smoking were adjusted for waterpipe smoking. cThe reference category for all models are the same 1619 ‘never tobacco users’. dIn 13 men (7 with reflux), waterpipe intensity and waterpipe-years could not be calculated. eIn 9 men (3 with reflux), cigarette intensity and pack-years could not be calculated.
Discussion
About a quarter of our study population had severe and/or weekly reflux. GERD symptoms were more common among women and were associated with waterpipe use. Among people who did not use anti-reflux medications, longer duration, higher intensity and more cumulative use of waterpipes (waterpipe-years) increased the chances of developing severe and/or frequent reflux symptoms. There was effect modification by sex and the associations were only seen among women.
The prevalence of frequent GERD (at least weekly) among both the Qashqai and non-Qashqai ethnicities in our study was lower than the previous study from this Province, which was conducted only among the Qashqai Nomads.17 Qashqai Nomads’ lifestyle has evolved dramatically over the past decades and changes in socio-political organizations, new land use policies, and the development of formal education programmes among the nomads have led to the shrinking nomadic lifestyle and increased settling of Qashqais.20 GERD prevalence in our study is similar to that reported in previous studies from Iran,19,21,22 which is higher than the rest of Asia and closer to Western populations.5 Some authors have tried to explain this by the increasing trends of obesity and the influence of a Western lifestyle,6 but one must also not forget cultural differences in the interpretation and reporting of symptoms that can affect the GERD definition. Members of the International Consensus Group have agreed that the ‘troublesome symptoms’ defining GERD are present when mild symptoms occur on 2 or more days a week or moderate/severe symptoms occur on more than 1 day a week (the Montreal Definition).4 Whereas we did not specifically ask whether the symptoms occurred twice a week for mild cases, the ‘frequent reflux’ category in our study (25.1% of the participants) probably comprise individuals closest to this definition. On the other hand, the ‘any reflux symptom’ group in our study, which includes more than half of the study population, may include individuals with non-specific symptoms, or people who do not have so much acid regurgitation to actually cause more severe or frequent GERD symptoms.
Cigarette smoking has been studied in several population-based studies of GERD risk factors. Although a few have failed to show an association,23,24 most of these studies have shown an association between cigarette smoking and GERD.5 The HUNT study in Norway, which is one of the few longitudinal studies on GERD, not only showed an association between cigarette smoking and new-onset GERD symptoms (OR = 1.37),25 but also showed a benefit from smoking cessation on GERD.26 In our study, cigarette smoking showed some association with reflux disease, and we could only detect this association in men, as there were not enough cigarette smokers among women. Waterpipe smoking, on the other hand, was more strongly and consistently associated with reflux disease overall and among women.
Two main mechanisms have been proposed for the association between smoking and GERD symptoms. Smoking can reduce the pressure of the lower esophageal sphincter (LES), thus allowing the gastric contents into the oesophagus.27 The second mechanism involves the effects of tobacco on reducing the salivary bicarbonate secretion, which can neutralize gastric acid.28 Tobacco smoking can also induce systemic inflammatory pathways, which can affect both the development of GERD and its progression to Barrett’s oesophagus and EAC.29 In fact, whereas both smoking and frequent GERD can independently increase the risk of esophageal adenocarcinoma, their combined effect has been shown to be synergistic.30 Our results suggest that these mechanisms may also be relevant for waterpipe smoking. Waterpipe smokers are exposed to very high levels of tobacco-specific nitrosamines, which are comparable to their levels cigarette smokers.31 Waterpipe smoke is a biologically active aerosol that delivers nicotine, CO and tar32 and there are at least 69 different carcinogens in waterpipe smoke for which a causal association with cancers has been established.33,34
There was an effect modification by sex and, in stratified analyses, the waterpipe–GERD associations were present among women. This might be partly because both GERD and waterpipe smoking were more common in women, but we do not have a clear hypothesis for the lack of an association in male waterpipe smokers.
This is one of the largest population-based studies on GERD5,35 and the first to have enough power to detect a dose–response association between waterpipe use and reflux disease. Another strength of this study is the availability of a detailed history of waterpipe smoking and other lifestyle habits. The most important limitation of this study is its cross-sectional design. It is difficult to draw a causal relationship from a cross-sectional study. Waterpipe smoking in this region has been part of a cultural practice for generations. Some participants may have stopped using waterpipes if they felt it could worsen their reflux symptoms. This may explain why some of the associations were stronger among former waterpipe users. However, any possible bias caused in this way should be towards null, and the actual association between waterpipe smoking and reflux disease is likely to be stronger than those we measured. Moreover, we observed an increasing trend with more cumulative use, which was more pronounced in severe and/or frequent reflux, and might be interpreted as evidence for a causal association between the exposure and the outcome.
In conclusion, this study provides evidence for an association between waterpipe tobacco use and reflux disease—a known risk factor for esophageal adenocarcinoma—which, in the face of the growing prevalence of waterpipe smoking in many populations, can inform public health and policy making.
Supplementary Data
Supplementary data are available at IJE online.
Funding
The field work was funded by Shiraz University of Medical Sciences (910210). This manuscript was supported in part by the Intramural Research Program of the US National Cancer Institute, NIH.
Supplementary Material
Acknowledgements
We wish to thank the participants in the study for their cooperation and the Behvarzes for their crucial role in recruitment. We also wish to express our thanks to the general practitioners, nurses and nutritionists in the core team. Author contributions: study concept and design (A.E., N.D.F., C.C.A., R.M.), acquisition of data (A.E., A.G., M.M., M.R.F., H.P., R.M.), analysis and interpretation of data (A.E., N.D.F., F.I., P.B., S.M.D., C.C.A.), drafting of the manuscript (A.E.), critical revision of the manuscript (all authors), statistical analysis (A.E.) and obtaining the funding and supervision of the team (R.M.). Guarantor of the article: Reza Malekzadeh.
Conflict of interest: The authors have no conflicts of interest to declare.
References
- 1. Lagergren J, Bergstrom R, Lindgren A. et al. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med 1999;340:825–31. [DOI] [PubMed] [Google Scholar]
- 2. Nilsson M, Johnsen R, Ye W. et al. Obesity and estrogen as risk factors for gastroesophageal reflux symptoms. JAMA 2003;290:66–72. [DOI] [PubMed] [Google Scholar]
- 3. Ruigomez A, Garcia Rodriguez LA, Wallander MA. et al. Natural history of gastro-oesophageal reflux disease diagnosed in general practice. Aliment Pharmacol Ther 2004;20:751–60. [DOI] [PubMed] [Google Scholar]
- 4. Vakil N, van Zanten SV, Kahrilas P. et al. The Montreal definition and classification of gastroesophageal reflux disease: a global evidence-based consensus. Am J Gastroenterol 2006;101:1900–20, quiz 43. [DOI] [PubMed] [Google Scholar]
- 5. El-Serag HB, Sweet S, Winchester CC. et al. Update on the epidemiology of gastro-oesophageal reflux disease: a systematic review. Gut 2014;63:871–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Sepanlou S, Khademi H, Abdollahzadeh N. et al. Time trends of gastro-esophageal reflux disease (GERD) and peptic ulcer disease (PUD) in Iran. Middle East J Dig Dis 2010;2:78–83. [PMC free article] [PubMed] [Google Scholar]
- 7. Zheng Z, Nordenstedt H, Pedersen NL. et al. Lifestyle factors and risk for symptomatic gastroesophageal reflux in monozygotic twins. Gastroenterology 2007;132:87–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Jacobson BC, Somers SC, Fuchs CS. et al. Body-mass index and symptoms of gastroesophageal reflux in women. N Engl J Med 2006;354:2340–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Cook MB, Kamangar F, Whiteman DC. et al. Cigarette smoking and adenocarcinomas of the esophagus and esophagogastric junction: a pooled analysis from the international BEACON consortium. J Natl Cancer Inst 2010;102:1344–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ward KD, Hammal F, VanderWeg MW. et al. Are waterpipe users interested in quitting? Nicotine Tob Res 2005;7:149–56. [DOI] [PubMed] [Google Scholar]
- 11. Maziak W. Rise of waterpipe smoking. BMJ-Brit Med J 2015;350:h1991. [DOI] [PubMed] [Google Scholar]
- 12. Maziak W, Ben Taleb Z, Bahelah R. et al. The global epidemiology of waterpipe smoking. Tob Control 2015;24:3–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Montazeri Z, Nyiraneza C, El-Katerji H. et al. Waterpipe smoking and cancer: systematic review and meta-analysis. Tob Control 2017;26:92–7. [DOI] [PubMed] [Google Scholar]
- 14. Waziry R, Jawad M, Ballout RA. et al. The effects of waterpipe tobacco smoking on health outcomes: an updated systematic review and meta-analysis. Int J Epidemiol 2017;46:32–43. [DOI] [PubMed] [Google Scholar]
- 15. Etemadi A, Khademi H, Kamangar F. et al. Hazards of cigarettes, smokeless tobacco and waterpipe in a Middle Eastern population: a cohort study of 50 000 individuals from Iran. Tob Control 2016. [epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Islami F, Nasseri-Moghaddam S, Pourshams A. et al. Determinants of gastroesophageal reflux disease, including hookah smoking and opium use- a cross-sectional analysis of 50,000 individuals. PLoS One 2014;9:e89256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Mostaghni A, Mehrabani D, Khademolhosseini F. et al. Prevalence and risk factors of gastroesophageal reflux disease in Qashqai migrating nomads, southern Iran. World J Gastroenterol 2009;15:961–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Gandomkar A, Poustchi H, Moini M. et al. Pars cohort study of non-communicable diseases in Iran: protocol and preliminary results. Int J Public Health 2017;62:397–406. [DOI] [PubMed] [Google Scholar]
- 19. Nasseri-Moghaddam S, Razjouyan H, Alimohamadi SM. et al. Prospective Acid Reflux Study of Iran (PARSI): methodology and study design. BMC Gastroenterol 2007;7:42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Gharakhlou M. A study of cultural changes among the Qashqai tribes in Iran. Researches in Geography 2006;56:1–11. [Google Scholar]
- 21. Nasseri-Moghaddam S, Mofid A, Ghotbi MH. et al. Epidemiological study of gastro-oesophageal reflux disease: reflux in spouse as a risk factor. Aliment Pharmacol Ther 2008;28:144–53. [DOI] [PubMed] [Google Scholar]
- 22. Nouraie M, Razjouyan H, Assady M. et al. Epidemiology of gastroesophageal reflux symptoms in Tehran, Iran: a population-based telephone survey. Arch Iran Med 2007;10:289–94. [PubMed] [Google Scholar]
- 23. Locke GR, 3rd, Talley NJ, Fett SL. et al. Prevalence and clinical spectrum of gastroesophageal reflux: a population-based study in Olmsted County, Minnesota. Gastroenterology 1997;112:1448–56. [DOI] [PubMed] [Google Scholar]
- 24. He J, Ma X, Zhao Y. et al. A population-based survey of the epidemiology of symptom-defined gastroesophageal reflux disease: the Systematic Investigation of Gastrointestinal Diseases in China. BMC Gastroenterol 2010;10:94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Hallan A, Bomme M, Hveem K. et al. Risk factors on the development of new-onset gastroesophageal reflux symptoms: a population-based prospective cohort study: the HUNT study. Am J Gastroenterol 2015;110:393–400, quiz 1. [DOI] [PubMed] [Google Scholar]
- 26. Ness-Jensen E, Lindam A, Lagergren J. et al. Tobacco smoking cessation and improved gastroesophageal reflux: a prospective population-based cohort study: the HUNT study. Am J Gastroenterol 2014;109:171–7. [DOI] [PubMed] [Google Scholar]
- 27. Chattopadhyay DK, Greaney MG, Irvin TT. Effect of cigarette smoking on the lower oesophageal sphincter. Gut 1977;18:833–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Kahrilas PJ, Gupta RR. The effect of cigarette smoking on salivation and esophageal acid clearance. J Lab Clin Med 1989;114:431–8. [PubMed] [Google Scholar]
- 29. Levitzky YS, Guo CY, Rong J. et al. Relation of smoking status to a panel of inflammatory markers: the Framingham offspring. Atherosclerosis 2008;201:217–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Pandeya N, Webb PM, Sadeghi S. et al. Gastro-oesophageal reflux symptoms and the risks of oesophageal cancer: are the effects modified by smoking, NSAIDs or acid suppressants? Gut 2010;59:31–8. [DOI] [PubMed] [Google Scholar]
- 31. Al Ali R, Rastam S, Ibrahim I. et al. A comparative study of systemic carcinogen exposure in waterpipe smokers, cigarette smokers and non-smokers. Tob Control 2015;24:125–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Ward KD. The waterpipe: an emerging global epidemic in need of action. Tob Control 2015;24:1–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Hoffmann D, Hoffmann I, El-Bayoumy K. The less harmful cigarette: a controversial issue. a tribute to Ernst L. Wynder. Chem Res Toxicol 2001;14:767–90. [DOI] [PubMed] [Google Scholar]
- 34. Tobacco smoke and involuntary smoking. IARC monographs on the evaluation of carcinogenic risks to humans. World Health Organization, International Agency for Research on Cancer 2004;83:1–1438. [PMC free article] [PubMed] [Google Scholar]
- 35. Dent J, El-Serag HB, Wallander MA. et al. Epidemiology of gastro-oesophageal reflux disease: a systematic review. Gut 2005;54:710–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.