Metabolic signature of short‐term low energy availability

Exposure to low energy availability (LEA) can potentially detrimentally affect athletes' health and performance. Timely identification is crucial, yet its detection is often delayed until severe symptoms emerge. Our objective was to identify characteristic differences in the serum metabolome as potential early LEA biomarkers. We performed large-scale metabolomics analyses of data from two highly controlled, randomized controlled trials, exposing trained adults to short-term (3-5 days) low or high energy availability (15 (LEA) versus 40 (HEA) kcal·kg FFM^-1·day^-1), which were achieved once with and once without daily aerobic exercise. Differences between LEA and HEA were prominent in triglycerides (0.66 ± 0.22 vs. 1.07 ± 0.47 mmol·L^-1), total fatty acids (9.46 ± 1.50 vs. 11.22 ± 2.59 mmol·L^-1), amino acids (e.g., alanine: 0.46 ± 0.10 vs. 0.58 ± 0.15 mmol·L^-1), very-low-density lipoproteins (0.57 ± 0.23 vs. 0.67 ± 0.25 mmol·L^-1), and ketone bodies (e.g., β-hydroxybutyrate: 364 ± 241 vs. 30 ± 17 μmol·L^-1; all FDR < 0.05). These patterns reveal a marked shift towards increased fat utilization, altered lipoprotein profiles, and enhanced ketogenesis in response to short-term LEA. Post-intervention β-hydroxybutyrate (>0.09 mmol·L^-1) best predicted LEA, regardless of whether LEA was achieved with or without exercise, supporting its candidacy for LEA screening. Overall, our findings provide new insight into the metabolomic signature of LEA and support metabolomics as a tool for early detection of LEA.