Prevalence of hookah smoking among Iranian pupils and university students: An updated systematic review and meta‐analysis

Abstract Objective Today, smoking is considered a pressing global health issue. The present study aimed to estimate the total prevalence of hookah smoking among pupils and university students in Iran. Materials and Methods This systematic review and meta‐analysis were conducted via searching in databases such as Scientific Information Database (SID), MagIran, Scopus, PubMed and Web of Sciences from inception to October 2021. We targeted observational studies evaluating the prevalence or frequency of hookah smoking among Iranian pupils and university students. Data analysis was performed using a random‐effects model, and the heterogeneity of the articles was assessed using Cochran's Q test and the I 2 statistic. Results In total, 124 studies conducted on 155 115 subjects were reviewed. The lifetime prevalence of hookah smoking among high school students and university students was estimated at 34.4% and 32.3%, respectively. In addition, the frequency of hookah smoking within the past month/week (point prevalence) was estimated at 21.5% and 16.6% in university students and pupils, respectively. The frequency of hookah smoking within the past year (period prevalence) was also reported to be 22.5% and 20.8% in these groups, respectively. No significant correlation was observed between the prevalence of hookah smoking, sample size, year of publication and the mean age of the participants. Region 5 had the highest lifetime prevalence (41.7%) and period prevalence (27.1%). However, Region 1 had the highest point prevalence of hookah smoking (27.2%). Conclusions According to the results, hookah smoking is highly prevalent among Iranian pupils and university students. Therefore, proper educational interventions are required in the form of workshops and curricula to raise awareness regarding the hazardous effects of this unhealthy habit on the young generation.

K E Y W O R D S hookah, Iran, prevalence, pupil, student

| INTRODUCTION
Smoking is a leading cause of premature and preventable deaths in the world. Each year, more than eight million die due to the diseases caused by tobacco consumption across the globe, which costs the global economy 1.4 billion dollars. 1 Hookah is a traditional way of smoking, which has gained popularity among pupils and students, currently becoming an epidemic. 2 The risks of hookah smoking are often undermined due to various misconceptions, such as the use of flavouring agents in tobacco products and the false public opinion that hookah is healthy because the hazardous agents in the tobacco are filtered through the water inside the tank. 3 Nonetheless, hookah is similar to cigarettes in terms of causing diseases such as cancer, respiratory and gastrointestinal disorders, chronic pulmonary diseases and cardiovascular diseases. 4,5 Furthermore, the persistent use of hookah could lead to nicotine dependency. 6,7 Statistics suggest that the prevalence of hookah smoking is 4% to34% worldwide. 8 In a comprehensive study conducted on 105 000 university students in the United States, the prevalence of hookah smoking was reported to be 8.4% within the past month. 9 In another research in the United States, 19 000 students were evaluated, and the recent prevalence rate (past 30 days) of hookah smoking was estimated at 3.4% in high school students. 10 Different results have been reported in the studies conducted on Iranian pupils and university students. Because smoking habits often initiate in adolescence, 11 the first step to the prevention of this issue is to identify the status of hookah smoking in the Iranian community so that proper interventions and managerial strategies could be developed in this regard. The present study aimed to estimate the prevalence of hookah smoking among Iranian pupils and university students.

| MATERIALS AND METHODS
This systematic review and meta-analysis aimed to determine the lifetime prevalence rate of hookah smoking in Iranian pupils and university students based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The protocol of this study has been sent to PROSPERO and is under review.
Because the national databases were not sensitive to Boolean operators, the database search was performed in a single-word manner. To retrieve more eligible papers and have access to 'grey literature', the references lists of the identified articles were also manually searched. The literature search was conducted during 20 September-20 October 2021. Selected articles were analysed in the EndNote software version 8. Following that, the papers were screened based on title, abstract and full text. The screening stages were carried out by two authors independently. In case of disagreement, a third author would resolve the matter. Finally, all the authors attended a meeting in which the final articles were put to discussion to be selected based on the inclusion and exclusion criteria.

| Inclusion and exclusion criteria
The inclusion criteria were observational studies published in English and Persian regarding the prevalence of hookah smoking among Iranian pupils and university students. Interventional studies, reviews, letters to the editor, qualitative/analytical studies, care reports and cohorts without essential data were excluded from the review.

| Data extraction
Two authors extracted data on the authors' names, year of publication, sample size, mean age of the participants, study area (city), target population (pupils/ university students) and the prevalence of hookah smoking.

| Qualitative assessment of the selected articles
Two authors qualitatively assessed the retrieved articles based on the Newcastle-Ottawa quality assessment scale (NOS). The NOS checklist is used to qualitatively assess observational studies, especially cross-sectional studies. The NOS evaluates articles based on six items in three dimensions of sample collection method, comparison and analysis of study groups, and the outcome measurement and analysis. If these criteria are met, each item is scored one, and the maximum achievable score for each article is nine. In the current review, discrepancies in the allocated score to the selected articles were resolved by the third author through discussion to reach a consensus. 12

| Statistical analysis
To perform the meta-analysis, the prevalence rates extracted from the selected studies were calculated based on standard error. The random-effects model by DerSimonian-Liard was also employed to estimate the cumulative prevalence of hookah smoking (95% confidence interval [CI]) using the Metaprop and Metan commands in the Stata software version 16. To assess the homogeneity and variance of the selected studies for the meta-analysis, we used Cochran's Q test and the I 2 statistic. Moreover, publication bias was evaluated using the funnel plot and Egger's test. Meta-regression analysis was also utilized to determine the correlations of age, sample size and year of publication with the prevalence of hookah smoking. In all the statistical tests, we considered α = 0.05.

| RESULTS
An initial search yielded 588 articles. After removing duplicates, the titles and abstracts of the remaining 218 articles were evaluated. Based on the inclusion criteria, irrelevant studies were excluded and the full text of the remaining articles was studied ( Figure 1).
In total, 124 eligible studies were selected for the current review, which had been conducted on 155 115 subjects (mean per study: 1251). The studies were performed on high school and university students. In addition, 44.5% of the studies had a larger sample size than 1000. Most of the studies (28.2%) were conducted in Region 1 of the country. The oldest and most recent F I G U R E 1 Article selection process studies were published in 2001 (n = 1) and 2021 (n = 5), respectively ( Table 1). The reviewed studies classified hookah smoking into three main categories, including lifetime prevalence (n = 92), point prevalence (n = 48) and period prevalence (n = 21). Correspondingly, the analysis of the findings indicated that the lifetime, period and point prevalence of hookah were 34.4% (95% CI: 31.5-37.2), 22% (95% CI: 19-25) and 18% (95% CI: 16-21), respectively ( Figure 2). Based on the country's districts and type of the samples (high school and university students), the findings indicated that the lifetime prevalence of hookah smoking was higher in the high school students (34.4%) compared with the university students (32.3%). On the other hand, the point (21.5% vs. 16 Table 2 shows the findings in more detail. According to the results of meta-regression analysis, the lifetime prevalence and period prevalence of hookah smoking were not correlated with the mean age of the participants, sample size and publication year. However, a significant correlation was observed between the point prevalence of hookah smoking and sample size (P = 0.015) (Figure 3).
One of the most important variables in metaregression analysis is the year of publication of the articles, because it shows how the prevalence of hookah use has changed over time. The lack of correlation between the year of publication and the prevalence of hookah indicates that all health advice and workshops held for students at universities and schools do not lead to a F I G U R E 3 The results of meta-regression. The relationship between year of publication with lifetime prevalence (A1), period prevalence (A2) and point prevalence (A3) and relationship between mean age of participants with lifetime prevalence (A4), period prevalence (A5) and point prevalence (A6). The size of the circles indicates the sample size of the articles F I G U R E 4 Publication bias for lifetime (B1), period (B2) and point (B3) prevalence of hookah smoking reduction in hookah use. One of the reasons for not reducing the prevalence of hookah is the misconception that the water in a hookah bottle is able to absorb nicotine and harmful substances, so hookah is not harmful. It seems that in order to reduce hookah use, it is necessary to correct these misconceptions and impose heavy penalties for hookah use in parks, promenades and even dormitories. Publication bias was not considered significant for the period prevalence of hookah smoking (P = 0.195), while it was significant for lifetime prevalence (P = 0.033) and point prevalence (P = 0.01) of hookah smoking (Figure 4).

| DISCUSSION
The current meta-analysis aimed to evaluate the prevalence of hookah smoking (lifetime, period and point prevalence) among Iranian pupils and university students. According to the findings, the lifetime prevalence of hookah smoking among Iranian pupils and university students was 34.4% and 32.3%, respectively. In this meta-analysis, we used a precise and inclusive strategy to retrieve and analyse the largest number of studies possible without a time limit. Meanwhile, previous meta-analyses in this regard evaluated fewer studies due to the lack of division based on lifetime, period and point prevalence. 137,138 In the previous two meta-analyses, 58 and 37 studies were analysed, respectively, while a much larger number of studies were conducted in this field. However, due to the incomplete search, many studies were missed in the previous two meta-analyses. Given that the decisions of health professionals are influenced by the findings of these studies, and poor findings can lead to wrong decisions, the implementation of this study (by eliminating the weakness of previous studies) seemed necessary. The number of studies that have examined the prevalence of hookah use in Iran is very large, which has a cultural aspect. In Iran, hookah is considered a recreational device that is used by families even in parks and recreation areas and even in some student dormitories, and despite health warnings, its use is still increasing. Consistent with our findings, the results of the metaanalysis conducted by Aydogan et al indicated that the prevalence of hookah smoking was 31% among Turkish students. 139 A similar study was performed on more than 2000 students in London, and the obtained results showed the lifetime prevalence of hookah smoking to be 40.1%, which is slightly higher than the estimated rate in the present study. 10 In a cross-sectional study conducted by Al-Delaimy and Al-Ani on 847 male high school students in Iraq, the prevalence of hookah smoking within the past 30 days was reported to be 46.1%, which is high than the current research. This discrepancy could be due to the fact that the mentioned study was performed on male students aged 15-18 years. 140 According to a meta-analysis, the prevalence of hookah smoking in university students in Saudi Arabia was 17%, which is a lower rate compared with the results of the present study. 141 The high prevalence of hookah smoking in Iranian pupils and students could be attributed to cultural factors and general attitudes. Hookah is commonly served in Iranian ceremonies, and hookah is frequently carried around by Iranian families' outdoors as a recreational activity. In the current research, no significant correlation was observed between the prevalence of hookah smoking and the mean age of the participants. It seems that hookah smoking has become inherent to the youth culture. One of the main reasons is the misconception that hookah is less harmful than cigarettes because people believe that its toxins are dissolved by the water in the tank. In the present study, the results of meta-regression analysis indicated no significant correlation between the prevalence of hookah smoking and the publication year of the studies. In other words, no ascending or descending trends were observed in this regard. Therefore, it could be concluded that the current measures taken to reduce the prevalence of hookah smoking are practically ineffective. The lifetime and point prevalence had a significant publication bias. The reason for this finding can be attributed to the tendency of journals to publish articles with specific results or that articles with specific results may not have been published.
One of the limitations of the present study was that data were not reported completely in the reviewed studies, which restricted our analysis. If the prevalence of hookah smoking had been reported based on gender and the lifetime prevalence, more practical findings would have been available. Also, in many articles, only one prevalence of hookah smoking was investigated and reported. Therefore, it is recommended that these limitations be addressed in further investigations in this regard. Despite these limitations, we analysed more data and a larger number of studies purposively, and the obtained results regarding the prevalence of hookah smoking in pupils and university students are more reliable.

| CONCLUSION
According to the results, hookah smoking is highly prevalent among Iranian pupils and university students, who are the future of our country, and this unhealthy habit affects almost one third of these populations. Therefore, proper training should be developed in the form of educational workshops or through academic and school curricula in order to raise the awareness of youngsters regarding the adverse health effects of hookah smoking.

ACKNOWLEDGMENTS
We would like to thank all the authors whose articles have been used in this systematic review.

CONFLICT OF INTEREST
The authors declare that they have no competing interests.
ETHICS STATEMENT Not applicable.
AUTHOR CONTRIBUTIONS HZ and RGG: concept development (provided idea for the research). YR and HS: search strategy. HZ, NK and RGG: data collection. RGG: supervision. YM: analysis/ interpretation. All authors: writing (responsible for writing a substantive part of the manuscript).

DATA AVAILABILITY STATEMENT
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.