BMI and Cervical Cancer Screening among White,African‐American,and Hispanic Women in the United States

Objectives: We examined cervical cancer screening by BMI in white, African‐American, and Hispanic women and explored women's reasons for not undergoing screening. Research Methods and Procedures: We used logistic regression to examine Pap testing in the preceding 3 years across BMI groups for white (n = 6419), African‐American (n = 1715), and Hispanic women (n = 1859) age 18 to 75 years who responded to the 2000 National Health Interview Survey. We used bivariable analyses to describe women's reasons for not undergoing testing and examined whether unscreened women received physician recommendations for screening. Results: Of 12, 170 women, 50% were normal weight, 26% were overweight, and 21% were obese. The proportion who reported Pap testing in the last 3 years was 86% in whites, 88% in African Americans, and 78% in Hispanics. After adjustment for sociodemographics, health care access, and illness burden, severely obese white women (BMI = 40+ kg/m²) were significantly less likely to undergo Pap testing (relative risk, 0.92; 95% CI, 0.83 to 0.99) compared with normal weight women. BMI was not associated with screening in African Americans or Hispanics. A higher proportion of obese white women than normal weight women cited putting off the test or embarrassment or discomfort as the primary reason for not undergoing screening. Among the unscreened, obese women were as likely as normal weight women to receive a physician recommendation to undergo screening. Discussion: Disparities in cervical cancer screening by body weight persist for white women with severe obesity. Disparities were not explained by differences in the rate of physician recommendations for screening, but obese white women may be more likely to delay screening or to find screening painful, uncomfortable, or embarrassing than normal weight white women. Efforts to increase screening among obese women should address women's reservations about screening.