Association of serum metabolites and breast cancer risk: A population‐based case–control study in black urban <scp>South African</scp> women

The incidence of breast cancer has been steadily increasing in South Africa over the past decades, but this rise can only be partially attributed to changes in known modifiable risk factors. Metabolomics may help to elucidate novel biological pathways and identify potential biomarkers associated with cancer. We investigated the association between serum metabolites and breast cancer risk in black women from Soweto, South Africa. An untargeted ultra-high-performance LC-MS method was used to measure molecular features in serum samples from a total of 396 breast cancer cases and 396 population-based controls matched on age and demographic settings, enrolled in the South African Breast Cancer study. A total of 5820 features were detected from metabolomics analyses and 1732 were retained for statistical analysis after data pre-processing and imputation. Multivariable conditional logistic regression was used to estimate odds ratios (ORs) and false discovery rate-adjusted (FDR) confidence interval (CI) for the association of metabolite features and breast cancer risk. Overall, 12 molecular features were significantly associated with odds of breast cancer (FDR <0.05); 11 features were associated with increased odds of breast cancer, and 1 feature was associated with decreased odds of breast cancer. Of these, 7 metabolic features corresponding to 3 individual metabolites were identified. Serum levels of cortisol (OR = 1.63, 95% CI = 1.18-2.25 per 1 SD increase), kynurenine (OR = 1.38, 95% CI = 1.01-1.88), and octenoylcarnitine (OR = 1.46, 95% CI = 1.02-2.08) were associated with higher odds of breast cancer. This study suggests that metabolic pathways related to cortisol, kynurenine, and carnitine metabolism may play a role in black African women with breast cancer. These results need to be explored in prospective studies.