Testing the reproducibility of metabolomic signatures in different populations is important to determine their utility in different populations. Using plasma metabolomic signatures representing a “healthy” and “unhealthy” dietary pattern that have been constructed using data from a randomised controlled crossover feeding study (DQFS)(1) this study aimed to map changes in diet-derived plasma metabolites in a sample of rural adults screened as at elevated-cardiovascular disease (CVD) risk who received 6-months of medical nutrition therapy (MNT)(2).
A small sample of MNT participants (n=11, 90% male) provided additional plasma samples for metabolomics analysis. Both the DQFS and MNT intervention plasma samples were analysed using ultra-high performance liquid chromatography-tandem mass spectrometry. Restricted maximum likelihood mixed-effect models were used to evaluate change in metabolites and dietary intake across the MNT intervention. Metabolanalyst was used to conduct principle component analysis (PCA) plots, PERMANOVA and partial least squares discriminant analysis (PLS-DA) plots were used to compare metabolite profiles between MNT participants and DQFS data.
Five metabolites significantly changed between baseline and the 3- or 6-month timepoints. Two metabolites [Ethylmalonate and Glycosyl-N-stearoyl-sphingosine (d18:1/18:0)] significantly reduced at both 3- and 6-months. Diet quality (%energy from nutrient-dense, core foods) significantly increased across the intervention (p<0.001). PERMANOVA results suggest a significant difference in metabolites between groups with the grouping variable explaining nearly half of the variation in the data (F-value: 19.7; R-squared: 0.46; p=0.001). PCA and PLS-DA plots identified that post MNT intervention 3- and 6-month metabolic signatures aligned more closely with the “healthy” dietary pattern signature. There was no significant difference between the “healthy” and the 6-month post MNT metabolomic signatures (p=0.14), suggesting stronger alignment with “healthy” metabolite signature at 6-months.
Although only a small sample size, findings demonstrated that the metabolomic signature identified in the controlled feeding study can be used to map changes in diet quality in response to an MNT intervention. This is promising for advancement of precision and personalised nutrition and potential inclusion of metabolomic data in clinical practice as a biomarker of diet quality and for the monitoring of intervention response.