Age at cancer diagnosis for blacks compared with whites in the United States

Abstract

Background: Younger ages at diagnosis for blacks compared with whites have been reported for several cancer types. However, the US black population is younger than the white population, which may bias age comparisons that do not account for the populations at risk.

Methods: We analyzed Surveillance, Epidemiology, and End Results data for non-Hispanic blacks and non-Hispanic whites from 18 regions for the year 2010. We calculated crude mean ages at diagnosis among cases of 29 cancer types for whites and blacks. Separately, we calculated adjusted means that corrected for differences in population structure, which we obtained by fitting linear regression models to the ages at diagnosis with statistical weights specific to age and sex. Negative differences indicate younger ages in blacks, while positive differences indicate older ages in blacks. All statistical tests were two-sided.

Results: Based on crude means, blacks were diagnosed at younger ages than whites for nearly every cancer type. However, adjustment for population structure shifted the comparisons toward older ages among blacks, and only six statistically significant differences of three or more years remained. Blacks were younger than whites at diagnosis for Kaposi sarcoma (-10.2 years), male soft tissue cancer (-5.6), male anal cancer (-5.5), and non-Hodgkin's lymphoma (-3.7), but older for cervical cancer (+4.7 years) and female thyroid cancer (+3.3). Smaller differences (<3 years) were present for female breast, female colon, lung, pancreas, prostate, and uterine corpus cancers (all P ≤ .001).

Conclusions: Most differences between blacks and whites in the age at cancer diagnosis are small. Large differences for a few cancer types may be driven by etiologic and subtype heterogeneity as well as disparities in access to care.

Figure 1.

Population age distributions for blacks and whites age 84 years or younger in 18 regions of the United States, 2010.

Figure 2.

Crude (A) and adjusted (B) differences in the mean age at cancer diagnosis between blacks and whites in 18 US regions, 2010. Crude mean differences were calculated among observed cancer cases, without adjustment for underlying population structure. Adjusted means use statistical weights to account for differences between blacks and whites in the population distribution of age and sex. Analyses are stratified by sex for cancers where the adjusted difference in mean age at diagnosis between blacks and whites was statistically significantly different between males and females. Dark bars indicate statistical significance of adjusted differences, based on a Bonferroni correction applied to P values derived from weighted linear regression models. We did not test for statistical significance of crude differences. Rectum includes rectosigmoid junction, anus includes anal canal and anorectum, and kidney includes renal pelvis.

Figure 3.

Age-specific cancer incidence rates for cancers with statistically significant adjusted differences of at least three years between blacks and whites in the mean age at cancer diagnosis in 18 US regions, 2010. Error bars represent 95% exact confidence intervals. A) Male anal cancer; B) cervical cancer; C) Kaposi sarcoma; D) non-Hodgkin’s lymphoma; E) male soft tissue cancer; F) female thyroid cancer.