Social class, race/ethnicity, and incidence of breast, cervix, colon, lung, and prostate cancer among Asian, black, Hispanic, and white residents of the San Francisco Bay Area, 1988-92 (United States)
N. Krieger et al., Social class, race/ethnicity, and incidence of breast, cervix, colon, lung, and prostate cancer among Asian, black, Hispanic, and white residents of the San Francisco Bay Area, 1988-92 (United States), CANC CAUSE, 10(6), 1999, pp. 525-537
Background: To date only eight US studies have simultaneously examined canc
er incidence in relation to social class and race/ethnicity; all but one in
cluded only black and white Americans. To address gaps in knowledge we thus
investigated socioeconomic gradients in cancer incidence among four mutual
ly exclusive US racial/ethnic groups - Asian and Pacific Islander, black, H
ispanic, and white - for five major cancer sites: breast, cervix, colon, lu
ng, and prostate cancer.
Methods: We generated age-adjusted cancer incidence rates stratified by soc
ioeconomic position using: (a) geocoded cancer registry records, (b) census
population counts, and (c) 1990 census block-group socioeconomic measures.
Cases (n = 70,899) were diagnosed between 1988 and 1992 and lived in seven
counties located in California's San Francisco Bay Area.
Results: Incidence rates varied as much if not more by socioeconomic positi
on than by race/ethnicity, and for each site the magnitude - and in some ca
ses direction - of the socioeconomic gradient differed by race/ethnicity an
d, where applicable, by gender. Breast cancer incidence increased with affl
uence only among Hispanic women. Incidence of cervical cancer increased wit
h socioeconomic deprivation among all four racial/ethnic groups, with trend
s strongest among white women. Lung cancer incidence increased with socioec
onomic deprivation among all but Hispanics, for whom incidence increased wi
th affluence. Colon and prostate cancer incidence were inconsistently assoc
iated with socioeconomic position.
Conclusions: These complex patterns defy easy generalization and illustrate
why US cancer data should be stratified by socioeconomic position, along w
ith race/ethnicity and gender, so as to improve cancer surveillance, resear
ch, and control.