dc.description.abstract |
Background and Objectives: Diet and its role in long-term health and disease continues to be recognised and investigated. Determining these relationships requires an accurate assessment of dietary intake, habitual exposure, and the many variables that influence and impact an individual’s dietary choices. The current approaches to dietary assessment rely heavily on self-report methods and therefore have a number of flaws and systematic biases. However, nutritional biomarkers are minimally affected by these behavioural factors. Dietary assessment self report tools, although useful, can therefore benefit from objective measures of intake. Identifying, validating, and establishing dietary metabolites and their use as intake biomarkers using metabolomic analysis and technologies, is thus necessary and beneficial in nutrition and health research. Methods and Study Design: Secondary analysis was performed on a primary randomised controlled, parallel-group trial involving 31 healthy men aged ≥70yrs. Participants were randomised into two groups (N=16,15), where they received a controlled whole-food diet for 10 w designed to achieve either 0.8 protein kg-1·d-1 (RDA) or 1.6g protein kg-1·d-1 (2RDA). Both diets were matched for food variety but varied in portions to adjust for protein intake. Fasting biofluid samples (plasma) were collected pre- and post- intervention and analysed using non-targeted polar metabolomics - Hydrophilic Interaction Liquid Chromatography High Performance Liquid Chromatography-Mass Spectometry (HILIC HPLC-MS) to profile the circulating plasma metabolome by matching to an in-house database. 7 d of weighted dietary data, collected by investigators, was used from the 10 w primary trial for secondary analysis, subsequent statistical analyses was performed to determine associations between dietary intake and the change (r) in metabolite profiles. Results: Participants in the 2RDA consumed, on average, more fish, chicken, meat, dairy products, and eggs than the RDA diet. 22 plasma metabolites were identified as significantly different between the RDA and 2RDA diets post-intervention (p<0.05). These included nitrogen- balance metabolites: creatine and urea, and carnitine-family metabolite glutarylcarnitine, that were positively correlated with increased protein-rich food intake. Several unidentified compounds possessed correlations with fruit, plant-based proteins, and carbohydrates (M165T801-, M84T736-, M128T828-). Positive correlations were also found, specifically, glutamine and 3-dehydrocarnitine, with protein rich-foods. Conclusions: Nitrogen, although a validated indicator of protein intake, is unlikely to be specific to particular foods. The other metabolites linked to increased protein intake were also controlled by endogenous processes to some magnitiude. Fragmentation of compounds also meant that many remained unnamed, despite possessing good associations with intake of particular foods. Thus, the present study was unable to clearly identify metabolites that could be used as distinct biomarkers of habitual intake, as many of those identified are also involved in endogenous processes, making it unclear as to the influence that dietary intake had. It is proposed that future investigations into metabolites that are impacted by both exogenous intake and endogenous processes would be of benefit in establishing their use as biomarkers of intake. Along with identifying metabolic profiles linked to dietary patterns, instead of single foods alone. |
en |