Passive smoking's role in diabetes

Smoking is the main cause of lung cancer, chronic obstructive pulmonary disease, and peripheral atherosclerosis and one of the most important risk factors for cardiovascular disease. In particular, the risk attributed to current smoking varies from 40% for coronary heart disease to more than 60% for cancers of the pharynx and oesophagus and more than 80% for lung cancer.¹ In addition, the exposure of non-smokers to environmental tobacco smoke has been associated with a substantial increase in their risk of coronary heart disease and cancer.²,³ Several investigators have suggested that both active and passive smoking affects the cardiovascular system through endothelial dysfunction, increases in oxidised low density lipoprotein cholesterol, platelet adherence, inflammation, and mitochondrial and oxidative damage, as well as an acute deterioration in the elastic properties of the aorta.^4-7 Indeed, some of the effects of passive smoke on the cardiovascular systems of non-smokers are comparable to the effects of smoking on smokers.³ Now comes another effect of active and passive smoking: an increased risk of glucose intolerance.

In this issue (p 1064) Houston et al show an increased risk from tobacco smoke of glucose intolerance,⁸ which is a precursor of diabetes and atherosclerotic disease.⁹ In particular, they found a strong association between exposure to tobacco smoke and the incidence of glucose intolerance during the 15 year follow-up of young adults. The incidence of glucose intolerance was 22% among smokers, 17% among those who never smoked but had been exposed to smoke, 14% among former smokers, and 12% among those who had never smoked and had no passive smoke exposure. In addition, the authors reported that current and never smokers with passive smoke exposure experienced, respectively, 65% and 35% higher risks of developing glucose intolerance than never smokers without passive exposure, even after adjustment for various baseline sociodemographic, biological, and behavioural factors. Moreover, in the analyses stratified by race and sex the risks of glucose intolerance associated with tobacco exposure were greater in men than in women and in white people than in African-Americans.

The number of people with diabetes or glucose intolerance is rising owing to population growth, ageing, urbanisation, and the increasing prevalence of obesity and physical inactivity.⁹ Though some studies have shown a dose-response association between smoking and the incidence of diabetes, others have failed to do so.¹⁰,¹¹ The study of Houston et al clearly showed that both active and passive smoking were associated with the development of glucose intolerance among young adults.⁸ Their study has several strengths, but also some limitations. The strengths include the large sample size (about 4600 participants), stratification by sex and race, validation of passive smoking by serum cotinine concentrations, long term follow-up (15 years), an adequate participation rate (> 74%), and controlling for potential sociodemographic, biological, and behavioural confounders. However, the effect of smoking on the incidence of glucose intolerance occurred irrespective of waist:hip ratio, baseline insulin, and C reactive protein levels, markers that have been associated with the development of diabetes and presence of smoking habits in previous studies.⁹,¹⁰ Potentially, the measurement of these markers at various time points during the follow-up might clarify whether or not they constitute a causal pathway between smoking and glucose intolerance.

In addition, the authors reported that smokers and never smokers with passive smoke exposure were more often African-American and less often women than never smokers with no passive smoke exposure, while current smokers also had less education, drank more alcohol, and had higher fat intakes than never smokers with no passive smoke exposure. Although these sociodemographic and lifestyle characteristics were taken into account in the analyses, residual confounding might still exist, and the variability and misclassification usually observed in measuring healthcare, socioeconomic, and lifestyle variables might mask the true findings.

The tobacco industry has vigorously contested allegations that passive smoking is dangerous,¹² but the evidence for the harmful effects of passive smoking keeps growing, and this study by Houston et al provides evidence for a new risk from exposure to tobacco smoke. The finding needs confirming, but in the meantime most non-smokers wish not to be exposed to tobacco smoke against their will. The momentum for bans on smoking in public places continues.