Abstract IA42: Multiethnic differences in BMI, body composition, and survival in colorectal and breast cancer

Abstract

Abstract Rationale: The disproportionate burden of obesity among minority populations is hypothesized to contribute to racial/ethnic disparities in cancer mortality rates (1). However, several studies note differences by race/ethnicity in the relationship of higher BMI to cancer survival, with the strongest adverse associations often seen among non-Hispanic white patients and the weakest among black patients (2-6). A potential explanation is that BMI is an imperfect proxy for total adiposity that does not distinguish muscle from fat mass. Further, data from noncancer patients suggests that body composition at a given BMI may vary by race/ethnicity, underscoring the need for more accurate measures of fat and lean mass to understand the contribution of the body habitus to disparities in cancer survival. Methods: We will summarize the existing literature and present new data on: (1) differences in the BMI-mortality relationship by race/ethnicity, (2) differences in body composition by race/ethnicity, and (3) associations of body composition with mortality among survivors of breast and colorectal cancer by race/ethnicity. For brevity, we will focus on black-white differences in overall survival from a cohort of ~5,000 nonmetastatic breast or colorectal cancer patients at Kaiser Permanente Northern California. Body composition was quantified using clinically-acquired computed tomography (CT) scans, a gold standard to assess adiposity and muscle mass. Results: Body composition at a given BMI differs by race/ethnicity in the general population and among cancer survivors. Among 234 black and 2,290 white breast cancer patients, BMI was not associated with survival. By contrast, higher total adiposity and lower skeletal muscle were both associated with increased risk of death after breast cancer: the HR (95% CI) comparing the highest (versus lowest) tertile of total adiposity index was 2.43 (1.20, 4.92) among black and 1.48 (1.13, 1.93) among white women, respectively, and the HR comparing the lowest (versus highest) tertile of skeletal muscle index was 1.67 (0.82, 3.41) among black and 1.37 (1.05, 1.80) among white women, respectively. Among 243 black and 2,192 white colorectal cancer patients, BMI was associated with survival only among white patients (overweight was associated with a 22% reduction in mortality). By contrast, higher visceral adiposity and lower skeletal muscle were associated with increased risk of death after colorectal cancer diagnosis in both groups: the HR (95% CI) comparing the highest (versus lowest) tertile of visceral adiposity index was 2.64 (1.39, 4.99) among black patients and 1.35 (1.07, 1.70) among white patients, respectively, and the HR comparing the lowest (versus highest) tertile of skeletal muscle index was 2.13 (1.17, 3.88) among black and 1.71 (1.39, 2.12) among white patients, respectively. Conclusion: Body composition assessed via CT scans produces measures of adiposity and muscle that associate with survival across racial/ethnic groups where BMI does not. Accurate assessment of body composition is critical for equitable prognostic assessments and for making appropriate lifestyle recommendations to diverse groups of survivors (e.g., recommending weight loss through caloric restriction only when beneficial and offering complementary interventions, such as building muscle mass through resistance training or physical activity, when appropriate).