Using the negative exponential distribution to quantitatively review the evidence on how rapidly the excess risk of ischaemic heart disease declines following quitting smoking

► No previous review has formally modelled the decline in IHD risk following quitting smoking. ► We estimated when the excess risk becomes half that of continuing smoking (the half-life H). ► Based on 41 independent data blocks we estimated H as 4.40 (95% CI 3.26 to 5.95) years. ► Individual estimates are extremely variable: 10years for 12 blocks. ► Claims that excess risk approximately halves after quitting for 1year do not reflect the overall data. No previous review has formally modelled the decline in IHD risk following quitting smoking. From PubMed searches and other sources we identified 15 prospective and eight case-control studies that compared IHD risk in current smokers, never smokers, and quitters by time period of quit, some studies providing separate blocks of results by sex, age or amount smoked. For each of 41 independent blocks, we estimated, using the negative exponential model, the time, H, when the excess risk reduced to half that caused by smoking. Goodness-of-fit to the model was adequate for 35 blocks, others showing a non-monotonic pattern of decline following quitting, with a variable pattern of misfit. After omitting one block with a current smoker RR 1.0, the combined H estimate was 4.40 (95% CI 3.26–5.95) years. There was considerable heterogeneity, H being 10years for 12. H increased (p