Dear Editor,

The first error in Dr Imed Harrabi and colleagues’ article is geographical as our country, Tunisia, is not located in the "in the northeast of Africa" as stated by the authors [1]. Probably this error is a consequence of another one: the fact that ««waterpipe»» use would be « growing in popularity worldwide, essentially in the Eastern Mediterranean region »[1]. Indeed, the « Western Mediterranean Region » (Libya, Tunisia, etc.) is also affected by the shisha "epidemic" that the authors describe as having reached “worldwide” proportions. Obviously there is a serious contradiction. The geographical distribution of shisha use is something and the official ««waterpipe »» antismoking research -mainly led by the the US-American University of Beirut and the US-Syrian Centre for Tobacco Studies/US-SCTS (Wasim Maziak, co-author of the paper, is also director of the latter)-, undoubtedly based in the “Eastern Mediterranean”, is something else.
Then, we have two other errors, of a methodological nature this time. The first one is that Harrabi et al [1] have forgotten to discuss their “self administered questionnaire to measure tobacco consumption by frequency of smoking, age of commencement and age of regular use”. In fact, questionnaires used for such epidemiological surveys have not been established with the necessary scientific rigour [2, 3].
Declaring that a “”waterpipe”” questionnaire is “standardised” (corresponding US-SCTS reference in the authors’ article) is not enough, particularly when such a tool has not been independently evaluated but rather imposed to the tobacco research community with no minimal scientific discussion [4].
The other error, of a methodological nature, this time, is to group under one universal entity (««waterpipe»», particularly in one word) different types of pipes which are actually used with different smoking products in different contexts. This error is in not only a scientific reductionism but also a nominalism that has fuelled a world confusion [2, 3, 4].
For instance, in our country, researchers have early understood the difference between pipes and products, the necessary quantification that the need to take into account anthropological details of utmost importance (type of charcoal, of smoking mixture (with or without glycerol; quantity), involved temperatures, the way the pipe is prepared, etc.)[2, 3, 4], Harrabi et al have also used ““waterpipe”” to list a series of related (health) problems [1]. However, they have not realised that in our country, as in many others, there is not one sort of water pipe only but several ones. If we leave aside the old « arguila » that is not anymore in use in our country for decades now [5], we have the shisha (spelled « chicha » in French) under two configurations: traditional on one side and modern on the other. Both imply different smoking products and a different heating/burning system. In the former one, the most popular product has been, for a long time now, « Cheikh el beled » (a national brand) which is an unflavoured plain tobacco-molasses mixture (“moassel”, i.e. “honeyed” in Arabic] with which the charcoal is in direct contact. This product is strong and it is one of the reasons for its exclusive use by males.
For two decades or so, it is also common to add jurak (pronounced « jirak » in our country) to the previous product. Jurak is a blackish paste (minced fruits and sugar are some of its main ingredients) which makes the above moassel even stronger. Its users commonly refer to a « buzz » perhaps due to a synergistic effect of nicotine and carbon monoxide although official advertised research is wading in other directions so that no study so far has clarified this phenomenon of utmost importance.
The other product in use in our country, not only by males but also, little by little, by women (although only for a few years - since our society remains conservative in many aspects) is also called moassel. However, it is a much more lighter flavoured tobacco-molasses mixture (in nicotine in particular) than «Cheikh el beled» or jurak. An important aspect is that unlike the two previous forms, the charcoal is not in direct contact but is separated by a pierced aluminium foil which serves as thermal screen whereby the smoking mixture is only heated (the temperature of the mixture does not go in excess of ca. 200°C)[2,3,4]. These are not secondary details in the light of the systematic comparison with cigarettes. Indeed, the tip of the tobacco rod in a cigarette can easily reach ca. 900°C. The consequences of these differences cannot be downplayed so easily. Shisha smoke, in its modern fashionable form, and unlike cigarettes, is mainly made up of water and glycerol (known to have no biological activity) and is far less concentrated in chemicals (hundreds vs. thousands) than cigarette smoke. It is surprising that the authors did not pay attention to these facts even more that the latter have been highlighted by a Tunisian team in a respectable peer-reviewed journal [2, 3]. This means that such chemical differences are going to induce health effects that will be different, not only from those induced by cigarettes but also different in each case : plain unflavoured moassel with or no jurak with no thermal screen, flavoured moassel with thermal screen, etc. In these conditions, stating that "waterpipe smoke contains many of the same toxicants as cigarette smoke » [1] is not exact from a scientific standpoint. Most of Harrabi et al’s references in this respect are mainly based on a smoking machine designed at the US-American University of Beirut which has been criticised for its biased underlying methodology [2, 3, 4].
In the light of the available sound science, only a few toxicants, among the thousands to be found in cigarette smoke, are common and these are in varying proportions: sometimes higher, sometimes lesser. Smoking patterns are also completely different in each case so that direct comparisons such as those based on consultations with the antismoking experts (e.g. “1 shisha equals 200 cigarettes”) [6], are also unscientific.
Harrabi et al cite a German study (by Fromme et al.) on the harm caused to «non-smokers exposed to waterpipe smoke » [1]. However, they omit to add that this study was not only biased but unrealistic and unethical (Helsinki Declaration on protection of subjects in medical research). Indeed, it was based on four successive smoking sessions involving four smokers in an unventilated room; windows and doors closed; quicklighting charcoal (not natural) still emitting hazardous particles at the beginning of each session; water not changed after each session; etc. In spite of such a surprising design, the German researchers came up with results that -after dividing the yields of toxicants by 4, the number of sessions- do not support Harrabi et al's claims [7].
One of Harrabi et al's objective seems to demonstrate that the « gateway » hypothesis (shisha smoking or the shisha experience leads to cigarette smoking) is true. They cite recent US articles that would show that ««waterpipe»» is a « strong[est] » predictor of current cigarette smoking ». In fact, the design of the corresponding surveys is very poor and the « tobacco questionnaire » used by Weglicki et al and Virginia Rice at Wayne University University (cited by Harrabi et al) was never disclosed and is expected to contain linguistic (including translation) bias (Arab-speaking communities there).
Furthermore, it is amazing that Harrabi et al [1] do not cite an Australian study whose much more solid methodology (and large sample) contradicts the gateway’ hypothesis [8].
From our follow-up of sound independent research on this issue, we believe that, as in the case of smokesless tobacco, direct evidence that shisha use triggers cigarette smoking, is totally lacking and that sound independent studies are needed.
Indeed, even for common sense, the above gateway hypothesis (not to mention Harrabi et al's huge confidence intervals) is simply not consistent with what we know about tobacco smoking in our country, Tunisia.
smoking in our country, Tunisia.
The nicotine-equivalent of 10 cigarettes per day for a daily ««waterpipe»» smoker is not only in contradiction with the opinion of the US-SCTS and US-AUB experts [6], but is also based on a biased “metaanalysis” (by Neergaard et al) that pooled and compared studies in which the pipes and the smoking products were completely different [4].
Concerning “”waterpipe”” “addiction”, at the core of Harrabi et al’s paper, is not it amazing that the tobacco industry –unlike antismoking research centres- has not produced one single study on narghile ? Obviously it knows that it is much less addictive (if not at all in some instances) than cigarettes and therefore not lucrative.
Finally, we would like to stress other weak points:
1. In their "study design”, Harrabi et al [1] have actually selected a sample of schoolchildren of Sousse in 2003. We have many reservations because by that time, flavoured moassel was not so common.
2. The great majority of the bibliographical references in Harrabi et al's paper are from the United States of America as if our researchers in Tunisia (from Maalej to Hsairi and from Ourari to Ben Saad), had not produced any relevant literature on this issue. This is a striking form of publication (bibliographical) bias [4, 7].
3. It is also amazing to see the authors cite the WHO report (whose main co-author is Wasim Maziak) without adding that it has been criticized for its numerous serious errors [4]. The first two sentences of the WHO report contain an error and a misquotation. In particular, this document states that in North
Africa (therefore in Tunisia), “it is not uncommon” (sic) to see children smoking the narghile with their parents. As Tunisian researchers (one of them being the author of the critique of the WHO erroneous report) [4], we must say that we are very concerned with these facts and this situation.
We hope that from now on, the Editors of La Tunisie Medicale will keep a watchful eye on any manuscript on shisha smoking and that they will demand more scientific rigour. For some reasons, this issue has amazingly become highly sensitive.

1. Harrabi I, Maaloul Jm, Gaha R, Kebaili R, Maziak W, Ghannem H. Comparison of cigarette and waterpipe smoking among pupils in the urban area of Sousse, Tunisia. Tunis Med 2010;88:470-3.
2. Ben Saad H. Le narguilé et ses effets sur la santé. Partie I: le narguilé, description générale et propriétés. Rev Pneumol Clin 2009;65:369-75.
3. Ben Saad H. Le narguilé et ses effets sur la santé. Partie II : les effets du narguilé sur la santé [The narghile and its effects on health. Rev Pneumol Clin 2010;66:132-44.
4. Chaouachi K, Sajid KM. A critique of recent hypotheses on oral (and lung) cancer induced by water pipe (hookah, shisha, narghile) tobacco smoking. Med Hypotheses 2010; 74: 843-6.
5. Gobert E. M?urs des fumeurs de chanvre. In: Toxicomanies orientales, Vichy médical. Tunis, Archives de l’Institut Pasteur, Dec 1937-Jan 1938 (pages 422- 433).
6. ASH (Action on Smoking and Health). ““Shisha 200 times worse than a cigarette” say Middle East experts””. 27 March 2007 (prepared by Martin Dockrell)(accessed 13 June, 2008). [based, among others, on an interview with Wasim Maziak and Alan Shihadeh] http://www.newash.org.uk/ash_4q8eg0ft.htm
7. Chaouachi K. Hookah (shisha, narghile, "water pipe") indoor air contamination in German unrealistic experiment. Serious methodological biases and ethical concern. Food Chem Toxicol 2010;48:992-5.
8. Carroll T, Poder N, Perusco A. Is concern about waterpipe tobacco smoking warranted? Aust N Z J Public Health 2008;32:181-2.