Extreme obesity and risk of endometrial cancer in the Cohort Consortium

The proposed study requires no funding as no new variables are needed. Our analysis would include those covariates already available in the BMI and physical activity meta-analyses conducted by our group. Any harmonization of variables will be done by in-house analysts and funding obtained within the Intramural research program.

Obesity is an established risk factor for developing endometrial cancer, but the nature of the relationship at extreme levels of adiposity is not well understood, nor is it clear whether differences exist by endometrial cancer subtypes. Recent data suggest that the dose-response relationship between body mass index (BMI) and endometrial cancer may be non-linear, thus previous estimations of the population-attributable risk (PAR) may underestimate the true public health impact of the obesity epidemic for this cancer.

Endometrial cancer is the most common gynecological cancer and the fourth leading cancer among women. In the United States, endometrial cancer incidence rates have remained steady over the past decade overall, but increased for some age groups. Mortality rates have continued to increase(3). A systematic review and meta-analysis of 26 prospective studies conducted as part of the American Institute for Cancer Research/World Cancer Research Fund (AICR/WCRF) Continuous Update Project (CUP) showed a 50% higher risk of endometrial cancer per 5-unit increase in BMI (RR=1.50; 95% CI: 1.42-1.59); updated meta-analyses have supported this finding. The association has been observed among both premenopausal and postmenopausal women, among users and nonusers of MHT with evidence for heterogeneity across subgroups, and across different racial groups. This relationship persists when assessing adiposity using different anthropometric measures. Controlled bariatric surgery interventions have also demonstrated that bariatric surgery with weight loss reduces risk of endometrial cancer. A 10% increase in BMI has also been associated with a similar magnitude of increase in risk for all-cause mortality among women with endometrial cancer. Taken together, the evidence for an association between increasing adiposity and endometrial cancer is convincing.

A non-linear dose-response relationship between BMI and endometrial cancer has been observed, with a steeper increase in risk at higher levels of BMI. For women in the class III obese range (i.e. BMI ≥40kg/m2), these estimates translate to approximately a 10-fold or higher risk of developing endometrial cancer compared with being in the healthy weight range. However, of studies conducted to date, the majority have categorized the highest level of obesity as ≥28kg/m2 to ≥40kg/m2 with no exploration of the shape of the relationship for BMI>40kg/m2. This could be influenced by the lower prevalence of morbid obesity contributing to low study case numbers and power to explore associations within this unique population. Population Attributable Risks (PARs) for incident cancer attributable to excess BMI increase substantially when modeling the shape of the BMI-cancer relationship as curvilinear versus linear. This implies that previous estimates of the burden of disease that could be prevented through obesity prevention have likely been underestimated, especially at the upper end of body weight distribution in the population.

Few studies have explored the relationship between BMI and endometrial cancer grade, or by subtype classifications (e.g. World Health Organization (WHO) histological features, Bockman classification, or more recent genetic classification according to common genetic alterations). The Cancer Prevention Study II Nutrition Cohort found a statistically significant association between higher BMI and increased risk of both type I and type II tumors, although the magnitudes of association differed. The NIH-AARP Diet and Health Study found differences in risk factors associated with Type I and Type II endometrial carcinomas, including MHT use, BMI, race, and family history of breast cancer. Few other studies have explored differences by cancer subtypes, particularly within the BMI class III range. Given the rare class III BMI subgroup as well as the number of cancer subtype classifications by histology and grade, a large sample size is required to explore the association between morbid obesity and endometrial cancer, and to accurately estimate PARs.

The major objective of this study is to provide an assessment of the association between morbid obesity and endometrial cancer.

1. To examine morbid obesity in relation to endometrial cancer.
This will help improve estimates of the dose-response relation between extreme levels of excess adiposity and cancer of the endometrium, and allow us to more accurately estimate the proportion of endometrial cases that would not occur if morbid obesity was prevented.

2. To examine morbid obesity in relation to endometrial cancer, stratified by grade/histological subtype and MHT use.

Proportional hazards regression will be used to examine obesity in 5-unit BMI increments compared with the healthy weight range (BMI 18.5-24.99kg/m2) in relation to endometrial cancer risk, including categories within the obese class III range (BMI ≥40Kg/m2). We will further assess the shape of the dose-response relationship using cubic spline regression. We will stratify analysis by endometrial cancer subtype and MHT use. Women will be excluded that reported at baseline that their uterus had been removed or with unknown uterus status.

A large sample size is needed to adequately determine the dose-response relationship between the rare class III BMI and endometrial cancer incidence. Using pooled cohort data would allow us to model the shape of the dose-response relationship at BMI>40kg/m2, and further determine the relationships by cancer subtype and MHT use with greater power. A pooled analysis would also provide better precision for estimating PARs.

100 total cancer cases

Incident endometrial cancer.

BMI or height and weight.

cohort, age, race, history of diabetes, age at menarche, parity, oral contraceptive use, age at menopause, menopausal status, menopausal hormone therapy (MHT) use at baseline, hysterectomy at baseline, number of births, age at first birth, first degree family history of endometrial cancer, previous history of breast cancer, tamoxifen use, smoking, physical activity, education, fruit and vegetable intake, fiber intake, energy intake, percentage energy intake from fat/carbohydrate, red/processed meat intake, alcohol consumption, hypertension; endometrial cancer subtype by histology (endometrioid, serous, clear cell, adenocarcinoma, mucinous, carcinosarcoma, sarcoma, other) and grade (1-4).

No