Genetic evidence showed that fat stored in specific regions of the body could affect cancer risk in different ways.
Abdominal subcutaneous adipose tissue increased the risk of endometrial cancer, hepatic cancer, and esophageal adenocarcinoma, whereas gluteofemoral adipose tissue was associated with lower risks of breast cancer and meningioma. Visceral adipose tissue and liver fat raised the risk of hepatic cancer. Pancreas fat was linked to higher risk of endometrioid ovarian cancer. In contrast, abdominal subcutaneous fat appeared protective against certain breast cancer subtypes, including luminal B/HER2-negative–like and triple-negative disease.
Lead study author Emma Hazelwood, PhD, of the MRC Integrative Epidemiology Unit at the University of Bristol, and colleagues explained the rationale for these findings: “Each deposit originates from a distinct fat progenitor and has a unique microenvironment and cytological makeup.”
The researchers reported that hormone and metabolic pathways partly explained these associations. For example, insulin and sex hormone–binding proteins accounted for part of the link between abdominal fat and endometrial cancer. The protein IGFBP-1 appeared to mediate the association between abdominal fat and esophageal adenocarcinoma. The protein adiponectin mediated part of the protective association between gluteofemoral fat and endometrial cancer.
The researchers used a two-sample Mendelian randomization approach, which applied genetic variants as proxies to assess causality between traits and disease risk. They included five adiposity distribution traits: abdominal subcutaneous fat, visceral fat, gluteofemoral fat, liver fat, and pancreas fat. Risks were evaluated for 12 obesity-related cancer types, including endometrial, ovarian, breast, colorectal, pancreatic, hepatic, renal, thyroid, gallbladder cancers; multiple myeloma; esophageal adenocarcinoma; and meningioma. Genetic data were drawn from large genome-wide association studies of European ancestry, with sample sizes ranging from more than 8,000 to more than 700,000 patients.
The researchers acknowledged several limitations. For instance, all of the data were from patients of European ancestry, which restricted the generalizability of the findings. Some analyses were limited by the absence of genetic instruments, such as IGF-1 for ovarian cancer outcomes. Although Mendelian randomization reduced confounding, it didn’t fully eliminate it. The researchers also noted that overlapping effects between fat depots could influence results, as seen when the associations of abdominal and liver fat with hepatic cancer were reduced in combined models.
Full disclosures can be found in the published study.