Abstract B25: Information on perceived breast cancer risk may help tailor risk assessment among high risk women

B25

Information on perceived breast cancer risk may help tailor risk assessment among high risk women. />Background: Women with a family history of breast cancer (FHBC) are believed to overestimate their breast cancer risk more than women from the general population. However there is little research examining which factors predict variations in perceived and actual risk among women with FHBC. We examined whether specific lifestyle and demographic factors are associated with over- or under-estimating BC risk. Methods: Cross-sectional data on 767 women with FHBC who attended the Breast and Ovarian Surveillance Service clinic at Johns Hopkins between February 1996 and March 2008 and were cancer-free were identified from a prospective database. Prior to their appointment women completed a questionnaire measuring breast cancer risk perception (5-year and life-time) and demographic and lifestyle behaviors. Detailed information on family history was also provided. Actual risk was determined using the Tyrer-Cuzick model. Odds ratios (OR) and 95% confidence intervals (CI) of being an over-estimator or under-estimator (in comparison with accurate estimators) were estimated from age-adjusted logistic regression. Results: Risk estimates were provided by 69% (531) of women. Those who did not provide risk estimates had significantly more breast cancer-affected family members (p = 0.040) and were more likely to be African American (p=.016). The median actual risk level was 23.79% (range: 0.70%-96.33%). In comparison with women who were accurate in their BC risk estimates (n=92), women who underestimated their lifetime BC risks by more than 10 percentage points (n=37) were younger (OR=0.94, 95%CI 0.90-0.98). In comparison to accurate women, women who overestimated their lifetime BC risk by more than 30 percentage points (n = 295) had a higher body mass index (BMI) (OR=1.08, 95%CI 1.03-1.13), had fewer years of education (OR=0.86, 95%CI 0.75-0.99), had more family members affected with BC (OR=1.28, 95%CI 1.02-1.60), and were more likely to have ever smoked (OR=1.79, 95%CI 1.08-2.97). Conclusion: Distinct profiles were observed for women who underestimated, overestimated or were accurate in their risk perception suggesting they may benefit from more tailored approaches to risk assessment.

Citation Information: Cancer Prev Res 2008;1(7 Suppl):B25.