Abstract P6-22-04: A pooled analysis of waist and hip circumference and postmenopausal breast cancer

Abstract

Abstract Background Many, but not all, epidemiologic studies suggest that waist and hip circumference, and waist-to-hip ratio [WHR] are positively associated with postmenopausal breast cancer. However, results are inconsistent for breast cancer subtypes, by strata of body mass index, or after adjustment for body mass index. We evaluated these associations in 10 prospective cohorts in the Pooling Project of Prospective Studies of Diet and Cancer. Methods These analyses included 256,214 postmenopausal women among whom 10,738 (N=1,346 estrogen receptor [ER]- and 6,587 ER+ tumors) were diagnosed with breast cancer during follow-up ranging up to 8-20 years. Study-specific, multivariable-adjusted hazard ratios [HR] were calculated for risk of breast cancer overall and by ER status using Cox proportional hazards regression analyses and controlled for age, race, education, height, reproductive factors, physical activity, smoking status, energy and alcohol intake, and family history of breast cancer. Pooled estimates were computed using random-effects models. Results The medians across studies ranged from 76.0-88.3 cm for waist circumference, 98.0-104.8 cm for hip circumference, and 0.77-0.84 for WHR. As expected, preliminary analyses showed stronger, positive associations for all three exposures for women who had never used hormone replacement therapy (HRT) compared to women who had used HRT (P for interaction <0.05). Subsequent results are presented only for never users of HRT. For waist circumference, the pooled HR of overall breast cancer was 1.58 (95% confidence interval [CI] 1.44-1.74) for women with waist circumference ≥100 cm compared to 70-80 cm. The corresponding pooled HRs were 1.65 (95% CI 1.45-1.86) for ER+ and 1.38 (95% CI 1.03-1.84) for ER- breast cancer. The associations for hip circumference and WHR were weaker than those observed for waist circumference. The pooled HR for women with hip circumference ≥110 cm compared to 95-100 cm was 1.22 (95% CI 1.11-1.35) for breast cancer risk overall. The HRs for ER+ and ER- tumors were 1.27 (95% CI 1.12-1.43) and 1.04 (95% CI 0.78-1.38), respectively. For WHR, the pooled HRs in women with WHR>0.90 compared to 0.75-0.80 were 1.37 (95% CI 1.23-1.52) for overall breast cancer, 1.33 (95% CI 1.15-1.53) for ER+, and 1.29 (95% CI 0.94-1.77) for ER- breast cancer. The findings for ER+ and ER- tumors were significantly different for waist and hip circumference (P-value, test for common effects < 0.05). After additionally adjusting for body mass index, the associations for all three exposures were attenuated and only the associations for waist circumference and WHR for overall and ER+ breast cancer remained statistically significant. Similar positive associations were observed for waist circumference across strata of body mass index: the HRs (95% CI) for a 5 cm increment were 1.06 (1.03-1.09) for women with body mass index <25 kg/m2 and 1.04 (1.02-1.06) for women with body mass index ≥25 kg/m2. Conclusions In postmenopausal women who had not used HRT, higher waist circumference, hip circumference and WHR were associated with a modestly higher risk of breast cancer. Results were stronger for ER+ than ER- tumors for waist and hip circumference and did not vary by body mass index for all three exposures. Citation Format: Song R, Wang M, Smith-Warner SA, For the Pooling Project of Prospective Studies of Diet and Cancer Investigators. A pooled analysis of waist and hip circumference and postmenopausal breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P6-22-04.