Genetic Ancestry and Colorectal Cancer in the All of Us Dataset

Area of Science:

• Genetics and Epidemiology
• Cancer Research
• Health Disparities

Background:

• Understanding colorectal cancer (CRC) disparities requires integrating genetic ancestry with other risk factors.
• Current CRC research has limited analyses incorporating genetic ancestry.
• Genetic ancestry may play a role in CRC incidence and age at diagnosis.

Purpose of the Study:

• To analyze the association between genetic ancestry and CRC burden.
• To investigate the relationship between genetic ancestry and age at CRC diagnosis.
• To develop and evaluate a multiethnic CRC risk-prediction model.

Main Methods:

• Retrospective cohort study using All of Us research program data (July 1986-October 2023).
• Inclusion of participants with linked electronic health record (EHR) and whole-genome sequencing (srWGS) data.
• Statistical analyses included Fisher exact tests, cumulative incidence functions, hazard models, and multivariable logistic regression; prediction models utilized penalized regression and extreme gradient boosting (XGBoost).

Main Results:

• Among 316,624 participants, 2914 developed CRC.
• European ancestry showed higher odds of CRC (OR, 1.50) compared to other ancestries.
• American admixed-Latino and East Asian ancestries had higher age-specific CRC hazards than European ancestry; the multiethnic XGBoost model achieved an AUC of 0.898.

Conclusions:

• Genetic ancestry is significantly associated with variations in CRC burden and age-specific risk.
• A multiethnic XGBoost model demonstrates strong performance in predicting CRC risk.
• This model can potentially enhance CRC screening strategies by identifying high-risk individuals.