Background: Body mass index (BMI) has been consistently associated with development of postmenopausal breast cancer (BC). There is accumulating evidence that waist circumference (WC) is a measure of visceral fat and thus may more accurately reflect the biological sequelae of adiposity on BC carcinogenesis. WC has been associated with development of BC, but its added value to BMI as a risk factor over time is less clear. This study leverages repeated in-person measurements to examine the association between WC, BMI and BC incidence in a diverse cohort of women.

Methods: We conducted a prospective study of 7626 (BC=564) postmenopausal women (29% Black, 71% white) in the ARIC Study who were 44-65 years old and cancer-free at baseline (1987-89). Women were categorized by time-varying WC (low <88cm, high ≥88cm) and BMI (normal <25, overweight 25-29.9, obese ≥30 kg/m2) measured at 4 clinical visits every 3 years, and followed until diagnosis of BC or another cancer, loss to follow up or 12/31/12, whichever occurred first. Cox proportional hazards models were used to estimate adjusted hazards ratios (HR) and 95% confidence intervals (CI) for BC using age as the time metric, overall and stratified by BMI (WC only). In a subset that had ≥3 visits and attended Visit 1 and 4 (N=4760, BC=244), we also 1) calculated the percent difference in Visit 1 and 4 measures, and 2) used linear regression to estimate the slope of change across all measures. The percent difference and slope coefficients were modeled as average change per year. Models were adjusted for age, race and study center, education, age at menarche, parity, frequency of routine exams, family history of BC, height, and hormone replacement therapy use. Change in WC/BMI analyses were also adjusted for WC/BMI at Visit 1.

Results: Compared to low WC, women with high WC had an increased risk for BC (HR 1.39, 95% CI 1.12, 1.73) that was slightly attenuated after adjustment for BMI (HR 1.27, 95% CI 0.97, 1.68). Stratification by BMI among normal and overweight women showed WC was associated with BC in normal weight (HR 1.47, 95% CI 1.06, 2.03) but not overweight women (HR 0.95, 95% CI 0.60, 1.50) (Pinteraction=0.12). In the subset that we analyzed change over time, the HRs for a 1% annual increase in WC and BMI were 1.12 (95% CI 0.99, 1.28) and 1.07 (95% CI 0.95, 1.21), respectively. For the slope change analyses, the HRs for a 1 cm annual increase in WC and 1 kg/m2 annual increase in BMI were 1.11 (95% CI 0.97, 1.27) and 1.22 (95% CI 0.81, 1.85). Results were similar among Black and white women.

Conclusion: In-person measured WC was associated with postmenopausal BC risk even after adjustment for BMI. Even in normal weight women, having a high WC increased risk of BC by 47%. Independent of baseline WC, increasing WC over time was associated with BC. This study highlights the importance of monitoring WC for risk of postmenopausal BC associated with adiposity.

Citation Format: Maneet Kaur, Avonne E. Connor, Elizabeth A. Platz, Kala Visvanathan, Corinne E. Joshu, for ARIC Cancer Investigators. A prospective evaluation of waist circumference over time and incidence of postmenopausal breast cancer in the Atherosclerosis Risk in Communities (ARIC) Study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 618.