Can the Metabolome Be Used to Assess Dietary Pattern Consumption? A Systematic Review of Evidence from Observational Studies ^†

Accurate dietary assessment is difficult yet fundamental for deriving meaningful findings from diet-related studies. Commonly used dietary assessment methods such as food diaries, recalls, and questionnaires rely on self-reporting and are often subject to bias. Furthermore, nutrition research often focuses on individual foods and nutrients, which has several limitations. Dietary pattern analysis considers the total diet and combinations of food eaten and can be derived empirically (e.g., factor analysis) using a diet index or a combination of both (e.g., reduced rank regression). Metabolomics, the study of low-molecular-weight compounds (metabolites), has been used for disease biomarker detection, and the technologies employed have advanced rapidly in recent years. As such, studies exploring the application of metabolomics for identifying objective biomarkers of dietary intake have also increased. The aim of our systematic review was to investigate observational studies exploring the relationship between dietary pattern consumption and the metabolome. A comprehensive search strategy was employed, using five search databases. Of the 14,328 studies screened, 35 met the pre-defined inclusion criteria and were included. All studies found a significant correlation between metabolites measured and dietary pattern scores. This demonstrates the potential for using the metabolome as a source of objective biomarkers of dietary pattern consumption. However, similar dietary patterns did not always result in similar metabolomic profiles across different study populations. This is likely due to the wide range of metabolomics methodologies employed, different biospecimens analysed, and a wide variety of populations and dietary patterns explored. In conclusion, although there is evidence that dietary pattern consumption can be reflected in the metabolomic profile of individuals, robust markers for specific dietary patterns have not yet been identified. Future studies should consider performing discovery and validation studies within a similar population, replicating the same methodology, to determine robust biomarkers of dietary pattern consumption.