Meeting-in-the-middle using metabolic profiling – a strategy for the identification of intermediate biomarkers in cohort studies

Predictive disease risk biomarkers that can be linked to exposure have proved difficult to identify in case-control studies.Parallel statistical analysis of the correlation between (1)H NMR profiles from plasma samples collected before disease onset (EPIC cohort), versus exposure to dietary compounds, and follow-up disease endpoints (colon and breast cancer) was performed.Metabonomic signatures associated with colon cancer and dietary fiber intake (a protective factor according to epidemiological studies) were identified.This implementation of the novel "meet-in-the-middle" analytical strategy indicates how case-control studies nested in prospectively collected cohorts may reveal intermediate biomarkers linking exposure and disease.