AEP Vol. 12, No. 7 ABSTRACTS (ACE) 493 October 2002: 488–534 crease then stabilization in GCA rates, rates of GCA and NGA among white men were similar in 1994–98. CONCLUSIONS: Declines in the prevalence of smoking and drinking and increases in consumption of fresh fruits and vegeta- bles may contribute to the downward trends in SCE. Increases in gastroesophageal reflux disease and obesity, and declines in Helico- bactor pylori ( H. pylori ) infection may contribute to the upward trends in ACE. Reductions in NGA may be related to improved diet and reductions in smoking and H. pylori prevalence. Factors contributing to the rising incidence of GCA of the cardia during the 1970s and 1980s, but not the 1990s, are less clear. PII S1047-2797(02)00296-X #9 BREAST AND CERVICAL CANCER MORTALITY IN THE MISSISSIPPI DELTA REGION HI Hall , P Jamison, SS Coughlin, Division of Cancer Prevention and Control, Centers for Disease Control and Prevention, Atlanta, GA PURPOSE: Historically, residents of the Mississippi Delta Region have suffered from high unemployment and poverty, deficits in ed- ucation, and inadequate access to health care. These factors may lead to deficits in cancer screening for women, higher incidence rates of late stage disease and higher death rates. This study as- sessed breast and cervical cancer mortality in the Mississippi Delta Region for 1979 through 1998. METHODS: Annual death rates for 1979 through 1998 and aver- age annual rates for 1994-1998 were calculated for the Mississippi Delta Region (235 counties in 8 states) from information reported on death certificates. Rates were age-adjusted to the 1970 U.S. standard population. Rates for all of the Delta Region and state specific rates for the Delta region were compared to rates for the remainder of the U.S. or the states, respectively. Trends were ex- amined with joinpoint regression techniques overall and by age and race over 1979-1998. Average annual death rates were also compared by rurality and a socioeconomic indicator of the county of residence. RESULTS: Overall breast cancer mortality was lower among white Mississippi Delta women than among other U.S. white women throughout the study period but, after rates decreased among both groups in the 1990s, the difference has narrowed. Breast cancer death rates among black women initially increased at a higher rate in the Delta region and rates remained similar in the Delta and other U.S. regions in recent years. Breast cancer death rates did not differ by economic status between regions. Cervical cancer death rates declined more rapidly in the remainder of the U.S. than the Delta region. Among white women, suburban and rural women in the Delta region had higher cervical cancer death rates than their counterparts in the other U.S. Among black women, residents of economically nondistressed or urban Delta counties had higher death rates than their counterparts in the other U.S. Death rates varied by geographic areas within states; Delta rates were elevated in 2 states for breast cancer and 6 states for cervical cancer com- pared to rates for the remainder of the states. CONCLUSION: The results provide information to guide pre- vention and control activities for reducing premature mortality from these diseases. PII S1047-2797(02)00297-1 #10 RACE DIFFERENCES IN MAMMOGRAPHY AND COLORECTAL CANCER SCREENING AMONG OLDER U.S. WOMEN, BEHAVIORAL RISK FACTOR SURVEILLANCE SYSTEM (BRFSS), 1999 D Holtzman , K Mack, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA PURPOSE: National guidelines recommend that persons aged 50 years at average risk for colorectal cancer should have regular screening tests. Breast cancer detection guidelines suggest yearly mammography for all women 40 years of age. This research ex- amines current screening practices for these cancers among US women by age group and by race/ethnicity. METHODS: Data come from the 1999 BRFSS, a state-specific, population-based, random telephone survey of health-related be- haviors and preventive practices (n = 58,127 women  40; 38,280  50). Data were aggregated across states and weighted prevalence estimates were calculated for receipt of mammography or sigmoidoscopy/colonoscopy. RESULTS: Overall, over 70% of women 40 years had a mam- mogram within the past two years. There was little variation by race/ethnicity in receipt of recent mammogram, and a curvilinear effect by age group, with the percentage of women having been re- cently screened dipping under 70% in the youngest (40–44) and oldest (80+) age groups. Conversely, only 42.8% of women 50 years reported having ever received a sigmoidoscopy/colonoscopy. Variation by age group and race/ethnicity was observed among those screened for colorectal cancer. For example, white women (42.8%, 95%CI = 41.9–43.7) were significantly more likely than black (37.6%, CI = 34.7–40.5), Hispanic (33.3%, CI = 37.9– 57.6), or Asian (28.2%, CI = 20.0–36.4) women 50 years to have ever had a sigmoidoscopy/colonoscopy. CONCLUSION: Our findings suggest that most women follow the recommended guidelines for mammography screening, how- ever, the majority are not screened for colorectal cancer, which also varies by age and race/ethnicity. Implications of the findings for early detection and cancer prevention will be discussed. PII S1047-2797(02)00298-3 #11 RACE / ETHNIC DIFFERENCES IN BREAST CANCER SURVIVAL BA Jones, SV Kasl, H Soler, P Van Ness, C Howe, M Lachman, A Beeghly, C Dallal, F Duan PURPOSE: In this study, we investigate the prognostic signifi- cance of tumor characteristics, genetic alterations, medical care,