Author(s): NUSSER, V., MURPHY, C., WASSERFURTH, P., KOEHLER, K., Institution: TECHNICAL UNIVERSITY OF MUNICH, Country: GERMANY, Abstract-ID: 2556

INTRODUCTION:
The impact of low energy availability (LEA) on metabolic processes has been widely documented in the literature, with notable alterations observed in various metabolic, endocrine and physiological pathways, e.g., sex hormones as well as indicators of bone and iron metabolism. However, a comprehensive understanding of the metabolic perturbations associated with LEA remains elusive. Metabolomics, capable of analyzing a vast array of metabolites at once, provides a unique opportunity to uncover the potentially complex metabolic signature of LEA, which holds promise for improved detection and characterization of LEA status.
METHODS:
In this study, we employed nuclear magnetic resonance-based metabolomics to quantify 250 metabolites and metabolite ratios in post-intervention blood samples obtained following short-term exposure to LEA (15 kcal/kg fat-free mass (FFM)/day) and normal EA as control (CON; 40 kcal/kg FFM/day). Blood samples utilized in our analysis were sourced from two larger crossover design studies (n=13, 85% males, aged 23.2±3.5 years), one of which involved daily aerobic exercise across both conditions, expending 15 kcal/kg FFM/day. We used generalized estimating equations to evaluate the effects of LEA on metabolite concentrations, while employing multiple logistic regression to predict LEA status based on metabolic profiles.
RESULTS:
We observed significant condition effects in 120 out of 250 metabolites, independent of exercise. Notably, triglycerides (LEA vs. CON: 0.63±0.20 vs. 0.99±0.44 mmol/L, adjusted p<0.05), fatty acids (9.22±1.38 vs. 10.65±2.51 mmol/L, adjusted p<0.05), ketone bodies (0.30±0.25 vs. 0.03±0.02 mmol/L, adjusted p<.001) and very-low density lipoprotein (VLDL) sub-classes (adjusted p<0.05) exhibited significant differences. Furthermore, the stepwise inclusion of these variables into a logistic regression model demonstrated their ability in predicting LEA status (LEA ~ Acetoacetate + Total triglycerides + Ratio of saturated fatty acids to total fatty acids, AIC=18.3, p<.001).
CONCLUSION:
Our analysis revealed significant group differences across a broad spectrum of metabolites, indicative of a transition towards increased fat utilization, ketosis, VLDL lipolysis and lipid transfer to high-density lipoprotein particles. These findings underscore the potential of metabolomics for identifying the metabolic signature of LEA, which may in turn be used to identify individuals currently exposed to LEA.