Reply to the letter to the editor Lissowska J, Brinton LA, Zatonski W, Blair A, Bardin‐Mikolajczak A, Peplonska B, Sherman ME, Szeszenia‐Dabrowska N, Chanock S, García‐Closas M. Tobacco smoking, NAT2 acetylation genotype and breast cancer risk. Lissowska et al. (Int J Cancer 2006; 119:1961–69). More evidence for passive and active smoking and breast cancer risk among younger women

Dear Sir, Lissowska et al. are to be congratulated on collecting lifetime residential and workplace exposure to secondhand smoke for their analysis of passive and active smoking and breast cancer in Poland.1 The report adds to the mounting evidence that active smoking is associated with increased breast cancer risk in younger/premenopausal women,2 (Table 6 in that review) as well as a more general trend toward studies observing increased breast cancer risk associated with smoking including 3 large American cohort studies,3, 4, 5 a large Canadian cohort study,6 two European cohorts7, 8 and studies that have controlled for passive smoking.9 Although the authors conclude that the “data indicate that passive smoking is not associated with breast cancer risk”, they present passive smoking breast cancer risk estimates of 1.28 (95% confidence interval (CI) 0.52–3.11) and 1.27 (95% CI 0.76–2.11) for women aged less than 45 and 45–55, respectively. The estimates are not statistically significant, however they are consistent with results for younger women from three recently published meta-analyses, which found elevated passive smoking summary risk estimates: 2 based on 14 studies of younger, primarily premenopausal women (1.68 (95% CI 1.33–2.12)),9, 10 (Figure 1) and most recently the estimate in the US Surgeon General's 2006 report based on 11 studies of premenopausal women (1.64 (95%CI 1.25–2.14)) (Table 7–10)10 Furthermore, Lissowska et al.'s passive smoking results among older women were also consistent with these meta-analyses—both finding little indication of increased risk for all exposed postmenopausal women.9, 10, 10 Finally, the negative conclusion of the authors is based on very low power in the critical younger age groups. Summary of 14 studies of breast cancer risk associated with passive smoking in younger/premenopausal women 9,10 and Lissowska et al. 1 for age <45 and age 45–55. The passive smoking dose-response analysis presented in the report is inadequate because it only examined risks among all-aged women combined. For women over age 55 the passive risk point estimate was only 1.04, 60% of the passively exposed women were in this age group and these older women likely dominated the highest cumulative exposure category given they have had the longest time period to accumulate exposure. The authors have provided dose-response analyses by age group for active smoking—the same is warranted for passive smoking given the difference in risk estimates by age group. Even though the power would be low, it would be informative for the authors to present tables of breast cancer risk for tertiles of hours/day-years of passive exposure for never active smokers for each of the following categorizations: (i) women under 45; (ii) women 45–55 years; (iii) women under 45 and 45–55 combined given the passive risk estimates were similar for under <45 and 45–55, and to help stabilize risk estimates; (iv) premenopausal women, (the report states that the age groupings were chosen to approximate pre, peri and post menopausal status—why not use menopausal status itself as well, given that it was collected); and (v) for women under 50, as the risk from the passive and active analysis suggests the risks are concentrated in younger women as do meta-analyses. Analyses which collapse the least exposed tertile of passively-exposed women with the never exposed would also help to stabilize risks, but at the expense of some misclassification. Presentation of these analyses would provide a more complete picture of this important work. Yours sincerely, Kenneth C Johnson Ph.D.*, * Evidence and Risk Assessment Division, Centre for Chronic Disease Prevention and Control Public Health Agency of Canada, Ottawa, ON, Canada.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.