LAUSR

Concern Regarding Age Distribution of Breast Cancer To the Editor The research letter by Stapleton et al1 presented the age distribution of individuals with breast cancer by race and ethnicity for women between the ages of 40 and 75 years using data from the Surveillance, Epidemiology, and End Results (SEER) Program database. The letter concluded that the higher proportion of individuals with breast cancer at younger ages in minority populations suggests a need for earlier screening for minority women. However, the results presented did not account for the different age distributions of the populations considered. As a result, the authors used the wrong denominator to assess risk. Instead of considering the proportion of women in a particular age group and race/ethnicity diagnosed with breast cancer, they considered the proportion of all breast cancers within that race/ ethnicity that occurred in a particular age group. To see the flaw in the authors’ argument, consider this hypothetical example. Imagine that the first members of a new race were born today. For the next 40 years, 100% of all cancers diagnosed in this race would occur in those younger than 40 years. This alone would not tell us that this new race was particularly prone to early cancers. To conclude this, we would need to compare the cancer rates of individuals of this race with the cancer rates for individuals of other races who were of comparable age. Consider non-Hispanic white women and Hispanic women (who have the highest portion of breast cancers diagnosed at younger ages) to address the impact of varying population distributions in the United States.2 Relative to case counts, the percentage of individuals diagnosed at an earlier age is higher for Hispanic women (Figure, A), with 6% of diagnoses of breast cancer in non-Hispanic white women and 12% in Hispanic women occurring between ages 40 to 44 years. However, Hispanic women are also more likely to be younger given the age distribution of these 2 populations (Figure, B). When combining case counts with the populations to estimate the age-specific incidence rates (Figure, C), the rates are lower for Hispanic women compared with non-Hispanic white women for all ages shown, reflecting a lower risk of developing breast cancer. Screening recommendations are based on balancing the benefits and harms of screening for an individual woman. These depend on the risk of cancer for women her age within her racial/ethnic group. An individual young woman’s cancer risk is not any greater just because a higher proportion of cancer in that group is diagnosed at younger ages.