Background: Reported family medical history may be the most widely accessible genomics screen currently available. The worth of any population-screening tool may be measured by parameters such as sensitivity and specificity. There is some evidence and theory suggesting that family history of breast cancer may be more frequently communicated about maternal (versus paternal) lines. If true, a discrepancy in reporting family history of breast cancer could mean suboptimal accuracy for this genomics-screening tool.
Methods: A cross-sectional analysis of reported family history of breast cancer was conducted beginning in early fall 2005 of data collected from April 2003 to March 2005. Study participants were Women's Health Clinic patients without breast cancer aged at least 40 years. The number of women reporting extended paternal relatives (e.g., aunts, grandmothers) with breast cancer was compared to the number of women reporting maternal relatives.
Results: More women reported a maternal relative with breast cancer (16%) compared to those reporting paternal relatives (10%) (McNemar odds ratio = 1.71, 95% confidence interval = 1.26 to 2.34). A discrepancy remained in analyses adjusting for potential covariates (race/ethnicity, awareness of breast cancer, family communication about breast cancer, and perceived risk).
Conclusions: Self-reported family history of breast cancer may not optimally capture inherited risk from the father's side of the family. Further research is needed to look for modifiable contributors to this discrepancy to optimize screening for familial breast cancer.