Erschienen in:
01.01.2012 | Original paper
Bodyweight and other correlates of symptom-detected breast cancers in a population offered screening
verfasst von:
Anne Kricker, Tracey DiSipio, Jennifer Stone, Chris Goumas, Jane E. Armes, Dorota M. Gertig, Bruce K. Armstrong
Erschienen in:
Cancer Causes & Control
|
Ausgabe 1/2012
Einloggen, um Zugang zu erhalten
Abstract
Objective
To determine the factors associated with symptom-detected breast cancers in a population offered screening.
Methods
We interviewed 1,459 Australian women aged 40–69, 946 with symptom-detected and 513 with mammogram-detected invasive breast cancers ≥1.1 cm in diameter about their personal, mammogram, and breast histories before diagnosis and reviewed medical records for tumor characteristics and mammogram dates, calculating ORs and 95% confidence intervals (CIs) for symptom- versus mammogram-detected cancers in logistic regression models.
Results
Lack of regular mammograms (<2 mammograms in the 4.5 years before diagnosis) was the strongest correlate of symptom-detected breast cancer (OR = 3.04 for irregular or no mammograms). In women who had regular mammograms (≥2 mammograms in the 4.5 years before diagnosis), the independent correlates of symptom-detected cancers were low BMI (OR < 25 kg/m2 vs. ≥ 30 kg/m2 = 2.18, 95% CI 1.23–3.84; p = 0.008), increased breast density (available in 498 women) (OR highest quarter vs. lowest = 3.50, 95% CI 1.76–6.97; p
trend = 0.004), high-grade cancer, and a larger cancer (each p < 0.01). In women who did not have regular mammograms, the independent correlates were age <50 years, a first cancer, and a ≥2-cm cancer. Smoking appeared to modify the association of symptom-detected cancer with low BMI (higher ORs for low BMI in current smokers) and estrogen receptor (ER) status (higher ORs for low BMI in ER cancers).
Conclusion
Women with low BMI may benefit from a tailored approach to breast cancer detection, particularly if they smoke.