Tobacco Documents Online

3 (v) As Francis Roe and I have commented in our critique of Glantz and Parmley's claims about the role of ETS and heart disease, recently published in Human and Ecological Risk Assessment (1999, 5 171-218), even if what has been observed is actually an effect of ETS exposure, it seems wildly speculative to assume that the fmdings are relevant to the longterm development of cardiovascular disease. PNLee 4.10.99

2 exposure ended 2 or more years before the visit (n = 17, mean 5.8, SE 1.0) than those whose exposure was more recent (n = 3, mean 1.2, SE 1.0). The authors clearly infer that endothelial function, "a marker of arterial health" is impaired by ETS exposure and that this is reversed by stopping exposure. Some points to note are as follows: (i) The differences observed between controls and passive smokers are similar to those obtained in a previous study by Celermajer et al (their first reference). In that study, the mean level of dilatation in active smokers (4.4%) was similar to that in passive smokers (3.1 %). As in that study, the current study collected no data on various factors that might effect arterial responsiveness, including diet, alcohol, coffee and general indoor and outdoor pollutants. (ii) Data were collected on reported intensity of exposure to ETS and on salivary cotinine. However, no attempt was made to correlate either of these two indices of extent of ETS exposure to endothelial dilatation. Why not? (iii) It is stated that arterial measurements were conducted blind of the smoking status of each participant and also at "the stage of the experiment", whatever that means. However, it is also stated that the former passive smoking group were "consecutivelv studied". Why was this? Could there have been any time trend in the accuracy of measurement? If so, measuring all of one group of 20 could have led to bias. (iv) There were no significant differences between the groups in respect of other factors measured (see Table on p 579). The authors state that "the determinants of endothelial function were assessed by multiple linear regression analysis" but it is not clear whether or not the results shown in the Table for endothelium-dependent dilatation are actually adjusted for any of these other factors or not.

1 REVIEW 1070 CONFIDENTIAL Subject ref 8f "Arterial endothelial dysfunction related to passive smoking is potentially reversible in healthy young adults" O T Raitakari et al Annals of Internal Medicine (1999), 130, 578-581 This paper describes the results of a cross-sectional study, conducted probably in Australia (though possibly in Finland), comparing three groups of201ifelong nonsmoking adults aged 15 to 39. A Controls: Never regularly exposed to tobacco smoke at home or in the workplace. B. Passive smokers: Exposed to ETS at home and/or at work for at least 1 hour per day for 2 or more years. C. Former passive smokers: Exposed for at least 1 hour per day for 2 or more years, exposure having ceased for at least 1 year before the study. Note that these definitions (based on the text in p 578) would allow those satisfying the requirements for group C to go into either group B or group C. However, the authors also refer to group B as"cun•ent passive smokers" so presumably the definition for "passive smokers" included the fact that exposure had occurred within the last year. The main finding of the study was that endothelium-dependent dilatation was highest in the controls (mean 8.9, standard error SE 0.7) lowest in current passive smokers (mean 2.3, SE 0.5) and intermediate in former passive smokers (mean 5.1, SE 0.9), the differences between groups being very highly statistically significant. The authors also note that there was a significantly (p <0.05) higher endothelial dilatation in the former passive smokers whose