The negative health risks of exposure to cigarette smoke are well known; including personal, second-hand and foetal exposure during pregnancy. In spite of this, rates of smoking (even during pregnancy) remain high, typically above 10% in most countries. This rate is even higher among certain subgroups of the population, and appears to be increasing in low and middle-income countries.
Smoking status is therefore of great importance in most epidemiological studies. Although questionnaires are the simplest and least invasive way to interrogate smoking status, the result is often unreliable and may be biased, particularly if based on self-report during pregnancy. Furthermore, questions about the number of cigarettes smoked may not capture exposure differences resulting from personal variation related to cigarette brands, smoking patterns such as depth of inhalation, metabolism rates and exposure to second-hand smoke.
Serum cotinine, a metabolite of cigarette smoke, can be measured but is limited by its short biological half-life of 17 hours for the average person. This half-life is even shorter (8 hours) for pregnant women, due to markedly accelerated metabolic clearance of cotinine during pregnancy. Cotinine measurements taken after even a single day of smoking cessation could result in false negatives. It certainly could not be used to estimate smoking intensity or pack-years.
DNA methylation, by contrast, appears to provide reliable and long-term biomarkers of tobacco smoke exposure. In fact, it is possible to use linear combinations of methylation levels at smoking-associated CpG sites to not only differentiate among current, former and never smokers but to also detect smoking intensity and number of smoking years.
We will use ARIES DNA methylation data to develop predictors of smoking behaviour which we will then test by comparing against self-reported smoking behaviour from questionnaires and ALSPAC cotinine data. This should allow us to assess the degree of smoking behaviour misreporting across defined groups of individuals (see aims and objectives).