To the Editor.
—My estimate that at least 75% of current breast cancer in the United States is due to earlier medical irradiation1 was called a 12-fold overestimate in the article by Mr Skolnick.2 He is quoting Clark Heath of the American Cancer Society, who asserts that I made "two serious errors."First, an alleged twofold overestimate comes from my "assumption" that dose response is supralinear. This is no assumption. The human evidence from the survivors of the atomic bomb for all cancers combined shows supralinearity fitting the observations provably better than linearity.3 Supralinearity specifically for breast cancer is visible to anyone who inspects the figure on page S26 of Thompson et al,4 although the analysis by Land5 raises questions.Second, an alleged sixfold overestimate comes from transport of Japanese evidence to the United States. Why did I not use North American data? Because (1)