Chen · Nutrition and cancer 2023 · umbrella review of observational meta-analyses · n=18 meta-analyses

Body Mass Index and Cancer Risk: An Umbrella Review of Meta-Analyses of Observational Studies.

Cited 32 times in the scientific literature.

Level 3 - non-randomized controlled study

Umbrella review of systematic reviews and meta-analyses of observational studies

PubMed 37139871 · doi:10.1080/01635581.2023.2180824 · record verified 2026-08-26

What was done

Authors conducted an umbrella review of systematic reviews and meta-analyses searched via PubMed, Embase, and Web of Science to evaluate the associations between body mass index (BMI) categories (underweight, overweight, obesity) and cancer incidence. Eighteen meta-analyses of observational studies met inclusion criteria, with 10 providing dose-response data.

What was found

Underweight was inversely associated with brain tumor incidence and positively associated with risks of esophageal and lung cancer. Overweight was linked to increased incidence of brain, kidney, endometrial, ovarian, bladder, and liver cancers, and multiple myeloma. Obesity was linked to higher incidence of 14 cancer types: brain, cervical, kidney, endometrial, esophageal, gastric, ovarian, multiple myeloma, gallbladder, bladder, colorectal, liver, thyroid, and Hodgkin's lymphoma. In dose-response analyses (10 studies), each 5 kg/m² increase in BMI was associated with a 1.01- to 1.13-fold increase in risk for brain tumors, multiple myeloma, bladder, pancreatic, breast cancer, and non-Hodgkin's lymphoma. Each 1 kg/m² increase in BMI was associated with a 6% and 4% risk increase for kidney and gallbladder cancers, respectively. Point estimates and confidence intervals for the categorical associations were not reported in the abstract.

Why it matters

This synthesis provides comprehensive broad-spectrum mapping of BMI-related oncologic risk across organ systems, reinforcing the relevance of adiposity management in population-level cancer prevention strategies.

Limits

All included evidence comes from observational studies, preventing causal inference and leaving susceptibility to residual confounding (e.g., smoking, diet). The abstract lacks total individual participant numbers, effect sizes and confidence intervals for categorical outcomes, and details on primary study overlap between included meta-analyses.

Cited by