Abdominal visceral and subcutaneous adipose tissue associations with postmenopausal breast cancer incidence

Abstract

Background: Obesity, classified by body mass index (BMI), is associated with higher postmenopausal breast cancer (BCa) risk. Yet, the associations between abdominal visceral (VAT) and subcutaneous adipose tissue (SAT) with BCa are unclear.

Methods: We assessed BCa associations with abdominal VAT and SAT in a prospective cohort of postmenopausal women without a history of cancer and with 27 years follow-up (N = 9950), during which all new cancers were adjudicated. Dual-energy x-ray absorptiometry scans assessed adiposity at baseline, year 3, and year 6. Competing-risks multivariable sub-hazard ratios (SHR), with adjustments for sociodemographic, behavioral, reproductive, and anthropometric characteristics, were estimated for baseline and time-dependent associations between VAT, SAT, and incident BCa.

Results: Participants averaged 63.3 ± 7.4 years of age and a BMI of 28.20 ± 5.72 kg/m2 at baseline. The models included 738 incident BCa case patients (N = 593 invasive; N = 145 in situ). Baseline VAT and SAT area were associated with statistically significantly increased BCa risk, by 36% and 19%, respectively. Increasing VAT/SAT ratio was associated with an 8% increase in incident BCa. Time-dependent models produced similar results. VAT and VAT/SAT associated BCa risk was highest for African American/Black women, although not statistically significantly different from other groups. Quartiles (Q) of VAT/SAT were also explored; the SHR for Q4 compared with Q1 was 1.49 (95% CI = 1.18 to 1.87).

Conclusion: Higher abdominal VAT and SAT are associated with an increased risk of postmenopausal BCa, and VAT/SAT may provide a distinctive risk estimate. Potential racial and ethnic differences require replication in a larger sample (Women's Health Initiative; NCT00000611; https://clinicaltrials.gov/study/NCT00000611).

Figure 1.

CONSORT diagram of postmenopausal women in the Women’s Health Initiative (WHI) included in the analyses of first incidence of breast cancer related to body composition. ^aCancer includes all types of cancer reported in WHI except nonmelanoma skin cancer. Abbreviation: DXA = dual-energy X-ray absorptiometry.

Figure 2.

Baseline abdominal adipose tissue variables (continuous per 100 cm²) and incident breast cancer overall and stratified by age, race/ethnicity, BMI category, and tumor characteristics (multivariable-adjusted models). Multivariable-adjusted models presented were adjusted for height at baseline, age at baseline, region, education, income, race and ethnicity, trial arm, alcohol intake, smoking status, physical activity (MET-h/wk), physical function (RAND 36 score), total energy intake (kcal/day), HEI-2015 score, menopausal hormone therapy use at baseline, aspirin use at baseline, metformin use at baseline, female relative with breast cancer, age at first birth, total number of months of breastfeeding, age at menopause, age at menarche, and surgical menopause. Adjusted models for American Indian/Alaska Native women did not converge. n = 9950 women observed for 177 294.53 person-years total. Competing risks are death without developing any type of cancer and developing a first primary cancer other than breast cancer. Multivariable-adjusted models used MICE to address missing covariates. Abbreviations: AA = African American; BMI = body mass index; ER = estrogen receptor; MET-h/wk = metabolic equivalent task hours per week; MICE = multiple imputation by chained equations; SAT = subcutaneous adipose tissue; SHR = sub-hazard ratio; VAT = visceral adipose tissue; CI = confidence interval.