Prospective study of circulating metabolomic profiles and breast cancer incidence among predominantly premenopausal women.

Background: Associations between premenopausal plasma metabolites and breast cancer incidence are largely unknown.

Methods: We conducted a prospective, matched case-control study in which we measured pre-diagnostic metabolomic profiles among predominantly premenopausal women in the Nurses' Health Study II (n = 2010). Lipids, carbohydrates, and organic acid-related metabolites (n = 218) were profiled via liquid chromatography-tandem mass spectrometry. Conditional logistic regression was used to estimate odds ratios (OR) for associations between individual metabolites and breast cancer incidence. Associations with metabolite groups were assessed using metabolite set enrichment analysis (MSEA).

Results: Six individual lipid-related metabolites were nominally associated with breast cancer incidence (taurodeoxycholate [OR for per 1 standard deviation increase in metabolite level = 1.15, 95% CI = 1.04-1.28]; C16:1 cholesteryl ester [OR = 0.88, 95% CI = 0.79-0.97]; three phosphocholine (PC)-related metabolites, C34:1 PC [OR = 0.87, 95% CI = 0.78-0.98], C34:3 PC [OR = 0.88, 95% CI = 0.79-0.98], C32:1 PC [OR = 0.88, 95% CI = 0.79-0.98]; indoxyl sulfate [OR = 0.90, 95% CI = 0.82-1.00]). In MSEA analyses, triglycerides (TAGs) with <3 double bonds (normalized enrichment score (NES) = -2.54) and PCs (NES = -2.12) were inversely associated with breast cancer incidence overall and across subgroups. Phosphatidylethanolamine (PE) plasmalogens (NES = 1.83) and PC plasmalogens (NES = 2.23) were positively associated with breast cancer incidence.

Conclusions: Premenopausal plasma TAGs, PCs, and plasmalogen metabolites were associated with breast cancer incidence. Further validation in independent cohorts is warranted.