ing frequency of eatingtchewing aji, we did see an increased risk in analyses versus stone controls of using aji 3-4 timestweek versus fewer than 3 timest week: the unadjusted OR(C1) was 3.6 (0.90-14.51, increasing to 5.9 (1.2-28.7) with adjustment for race. However, higher levels of frequency of aji use had odds ratios of less than 2. In Mexico, the variable we analyzed was the ingestion of hot sauce. Using the stone controls, relative to a reference group of ingesting it less than twice each week, ingesting 2 or 3 timeslweek, or 4 or more timestweek, were associated with OR(C1) of 0 (0-4.33) and 1.15 (0.19-12.41, respectively.
As to our reasons for not exploring gallstone disease, this was by design. It is very clear that gallstone disease is an extremely strong risk factor for gallbladder cancer. Whether gallstone disease is involved in the etiology of the cancer or, rather, the gallbladder cancer is a nidus for the formation of the stones is not clear, nor has it been clear why only a very small subset of patients with stones go on to develop cancer. As such, our study was designed with two control groups, one with stones and one without stones, to try to examine these questions. However, by guaranteeing that our control subjects either did or did not have stones, by design, we lost the ability to study gallstones as a risk factor. We did, in fact, analyze the two control groups separately. In general, our results with the gallstone control group and the nonstone control group were essentially the same, except as indicated in the original article.