Association Between Being Underweight and Outcomes in Breast Cancer: Alternative Explanations

Abstract

erable interest. The authors are to be commended for an interesting analysis from a national multicenter data set, concluding that being underweight should be considered as a risk factor for death and recurrence after breast cancer surgery. A possible alternative explanation for the observed association between being underweight and breast cancer outcomes may be the contribution of mammographic screening in older women. Given that the authors present data showing weight increases from younger to older age groups, it seems possible that a higher proportion of mammographic screening among older women will result in a survival bias and stage shifting for older individuals. Mammographic screening is commonly recommended for women in Asia older than age 40 years. This threshold is based on the observation that the proportion of patients diagnosed at a younger age is higher than in the West. Although the authors have controlled for age and histologic grade, it is established that the length and lead time bias contributed by mammographic screening are obvious even after controlling for these effects. 2 In addition, the apparent increase in breast cancer–specific mortality associated with being underweight may be confounded by a cohort effect. These cohort effects in breast cancer incidence have been observed in other Asian countries such as Singapore, where rapid industrialization has resulted in dramatic changes in lifestyle. 3 In Korea, where the study took place over a period of 15 years, increases in average body mass index and obesity have been observed over recent years for each cohort. 4 Further, with the introduction of the Korean National Cancer Screening program for breast cancer from 1999 onward, 3 there is a clear potential for a cohort effect. To control for possible cohort effects, the authors could examine 5-year relative survival against birth cohort using a 3-year central moving average 5