Analytical method validation was performed following the main requirements of the U.S. Food and Drug Administration (FDA) guidelines for bioanalytical method validation (13). These include parameters like linearity, precision, accuracy, carryover, matrix effects, recovery, selectivity, stability, limit of detection (LOD) and limit of quantification (LOQ). For precise description on the method development including synthesis of the internal standard (14) and validation process, see Supplementary Material.

Fifteen male and female triathletes (n = 15, nine males and six females) from the North Rhine Westphalian (regional) triathlon youth squad were included in the study (mean + sd, age: 16 ± 2 years, bodyweight: 60 ± 10 kg, height: 171 ± 8 cm, V˙O2max: 59 ± 6 ml/min/kg) and tested on two occasions (Table 1 for sex-specific and occasion-specific data). Two subjects dropped out of the post-measurement period. One subject, that showed implausible lac-phe data due to a methodological error, was excluded from the analysis. The participants and their legal guardians were informed about the benefits and risks of the study and provided written informed consent prior to participation. The study was approved and based on the Declaration of Helsinki by the university's local ethics committee (011/2023) before the start of data collection.

The subjects arrived at the laboratory on two different occasions, separated by a preparatory training phase prior to the season between their two test days. Subjects were educated on maintaining an adequate carbohydrate (CHO) intake of 5–8 g/kg body weight (BW) in the two days before the test day and consumed their habitual breakfast and lunch on the test day prior to the exercise protocol. The participants were asked to report their food intake and exercise activity using a validated reporting protocol (15). Before starting the test, bioimpedance analysis was performed to determine the body composition (SECA Medical Body Composition Analyzer, Hamburg, Germany). On both test days, the subjects performed their tests at the same time in the afternoon. During the diagnostic test, the subjects drank water ad libitum but did not consume any energy-containing foods or beverages.

Data were collected before, during and after the diagnostic test (Figure 1). The comprehensive test protocol has been described elsewhere (16). Concerning this study, in brief, the first blood sample was collected before the test for a baseline (BL) in a seated and resting position. After the warm-up (running at 2.4 m/s + 0.4 m/s every five minutes for 15 min) on a treadmill (h/p/cosmos pulsar® 3p, Traunstein, Germany), a 15 s all-out sprint-test on a bike ergometer (SRM ErgoMeter, Juelich, Germany) followed. Blood samples were taken before (Pre-Sprint) and two minutes after the test (Post-Sprint +2 min). After 9 min of passive recovery, active recovery was performed until lactate concentration decreased below 1.5 mmol/L (10–15 min). Afterwards, a first incremental step test on a bike ergometer (1.5 W/kg + 20 W every three minutes) was performed till volitional exhaustion. Blood samples were collected before (IT-1) and at time-to-exhaustion (TTE-1). A second incremental step test on the treadmill followed and was also carried out until exhaustion. Blood samples were taken before (IT-2), directly after (TTE-2) as well as five (TTE-2 + 5 min) and 15 min (TTE-2 + 15 min) after the test. The incremental bike and treadmill tests were separated by 8 min of passive rest. Spirometric data (Metalyzer 3B Cortex Biophysik GmbH, Leipzig, Germany) were collected during the warm-up and both incremental tests every second. The spirometer was calibrated on each testing day before the first test with a reference gas and before each test section with ambient air and a 3 L-syringe according to the manufacturer's instructions.

Capillary blood samples were collected from the earlobe by trained researchers using a 20 µl capillary tube. The tube was emptied on a dbs filter card (QIAcard FTA DMPK-C) for lac-phe analysis. Simultaneously, using another 20 µl capillary tube, capillary lactate and glucose samples were taken and directly analyzed using Biosen C-Line (EKF diagnostics Holding, Cardiff, UK). The dbs cards were dried for at least three hours at room temperature in the dark. Subsequently, the dbs cards were stored individually in Ziploc plastic bags with a desiccant (silica gel) at −80°C until analysis. The detailed sample preparation is explained in the Supplementary Material. Briefly, the entire dried blood spot was punched from the card, cut, and extracted into 600 µl of methanol (100%) in a tube. To quantify lac-phe, an internal standard (d₅-lactoyl-phenylalanine) was added to the tubes. After sonication for 30 min, the tubes were centrifuged at 17,000 × g for 10 min (19°C). The supernatant was dried in a vacuum centrifuge for 90 min (45°C) and reconstituted with a 10% acetonitrile solution. After a second centrifugation at 17,000 × g for 10 min (19°C), the supernatant was transferred into a vial and analyzed using LC-MS/MS (for instrumental conditions, please see Supplementary Material). Lac-phe concentrations exceeding the validated linear working range of 0–100 ng/ml were calculated by extrapolation.

Statistical data analysis was conducted using R version 4.1.3 (The R Foundation for Statistical Computing) and Excel 2019 (Microsoft Corp., Redmond, WA, USA). Data are presented as mean ± standard deviation. Coefficients of variations were calculated. Before performing parametric or non-parametric tests, normality of residuals was tested using the Shapiro–Wilk test and graphical analysis by QQ plots and histograms. Wilcoxon sum rank test or independent t-test for independent samples and Wilcoxon or paired t-test was performed for dependent samples. To quantify the magnitude of changes between pre vs. post occasion and male vs. female subjects, respectively, Cohen's effect sizes (ESs) were calculated and classified as trivial (ES <0.2), small (0.2 ≤ ES <0.6), moderate (0.6 ≤ ES <1.2), and large (1.2 ≤ ES <2.0) (17). Since some participants completed two tests at different times (pre and post), associations between variables were assessed using mixed linear models with the independent variable as a fixed effect and a random intercept for participant ID to account for repeated measurements. Estimated slopes were classified as trivial (|r| < 0.1), low (0.1 ≤ |r| <µ 0.3), moderate (0.3 ≤ |r| < 0.5), high (0.5 ≤ |r| < 0.7), very high (0.7 ≤ |r| < 0.9), and nearly perfect (|r| ≥ 0.9) (17). Only data within the 99% confidence interval were considered in the final statistical analysis, which led to exclusion of three values above this interval.

A summary of all the validation results is presented in the Supplementary Material (Supplementary Table S1).

There was no statistical significant difference in subjects' nutrition within the two days before the first (pre) and the second (post) diagnostic, considering energy intake (pre 2,582.7 ± 1,180.2 kcal/day, post 2,639.8 ± 1,230.7 kcal/day; p = 0.727), relative energy intake (pre 44.2 ± 16.6 kcal/kg BW/day, post 43.8 ± 16.9 g/kg BW/day; p = 0.844), CHO intake (pre 335.3 ± 151.6 g/day, post 341.2 ± 136.4 g/day; p = 0.78) and relative CHO intake (pre 5.8 ± 2.2 g/kg BW/day, post 5.7 ± 2.1 g/kg BW/day; p = 0.972).

Throughout the performance test and subsequent recovery period of 15 min, the mean capillary glucose concentration was 4.9 ± 0.8 mmol/L, the capillary lactate concentration was 4.6 ± 3.3 mmol/L, and the dbs lac-phe concentration in dried whole-blood spot samples was 67.9 ± 37.3 nmol/L. Individual curves are displayed in the Supplementary Material (Supplementary Figures S1a–c). Individual maximum dbs lac-phe concentrations were reached before, immediately after, and five minutes after the second incremental test (IT-2: n_max = 11, TTE-2: n_max = 9, TTE-2 + 5 min: n_max = 8). The mean maximum dbs lac-phe concentration was 107 ± 37 nmol/L.

Capillary glucose concentrations were significantly lower at IT-2 (p = 0.043) and TTE-2 + 5 min (p = 0.017) but significantly higher at TTE-2 + 15 min (p = 0.014) in the post-test than in the pre-test. Capillary lactate concentrations were significantly lower for IT-1 (p = 0.028), TTE-1 (p = 0.004), TTE-2 (p = 0.021), TTE-2 + 5 min (p = 0.021), and TTE-2 + 15 min (p = 0.025) in the post-test than in the pre-test. Dbs lac-phe concentrations did not show significant differences for any timepoint when comparing pre- and post-test (Figure 2). The mean concentrations, standard deviations, coefficients of variation coefficients, and statistical p-values for each timepoint are displayed in Table 2.

Capillary glucose concentration was significantly higher at TTE-1 (p = 0.006) and at IT-2 (p = 0.044) and lower at Pre-Sprint (p = 0.021) and at Post-Sprint + 2 min (p = 0.032) in females than in males. No significant differences in capillary lactate concentrations were found between sexes. Dbs lac-phe concentrations at IT-2 (p = 0.041) and TTE-2 (p = 0.015) were significantly higher in female subjects. In general, mean female capillary lactate and dbs lac-phe concentrations tended to be higher than mean male concentrations, especially during incremental tests (Figure 3). The mean concentrations, standard deviations, coefficients of variation, and statistical p-values for each timepoint are displayed in Table 3.

Linear mixed effect models indicated statistical significance between lac-phe and glucose concentrations at IT-2 (β = 0.016, p = 0.002), TTE-2 (β = 0.012, p = 0.003), TTE-2 5 min (β = 0.011, p = 0.032), and TTE-2 15 min (β = 0.015, p = 0.025) (Figure 4A). For lac-phe and lactate concentrations, statistical significance is indicated at Post-Sprint + 2 min (β = 0.016, p = 0.034), TTE-1 (β = 0.027, p = 0.032), IT-2 (β = 0.030, p = 0.010), TTE-2 (β = 0.018, p = 0.006), and TTE-2 15 min (β = 0.013, p = 0.047) (Figure 4B). For lactate concentration at timepoint t and lac-phe concentration at timepoint t + 1, statistical significance is indicated at TTE-1 (β = 0.027, p = 0.016), IT-2 (β = 0.016, p = 0.048), TTE-2 (β = 0.022, p = 0.007) and TTE-2 + 5 min (β = 0.021, p = 0.045) (Figure 4C). Although statistically significant, the estimated slopes indicate only weak associations. Non-significant results were found for the AUC (t4-t5) of lac-phe and the AUC of lactate (t5-t4) (β = 0.002, p = 0.766), and time to exhaustion (β = 0.712 p = 0.169), as well as for the AUC (t4-t7) of lac-phe and the AUC (t4-t7) of lactate (β = 0.016, p = 0.082), the accumulated time to exhaustion (β = 0.954, p = 0.302), CHO oxidation (β = 0.036, p = 0.700), fat oxidation (β = −0.007, p = 0.594), energy expenditure (EEx) (β = 0.196, p = 0.552), V˙O2max (β = 0.005, p = 0.557), body mass index (BMI) (β = 0.001, p = 0.440), body fat percentage (BFP) (β = 0.007, p = 0.434), muscle mass (MM) (β = −0.007, p = 0.191), fat mass (FM) (β = 0.003, p = 0.672), fat-free mass (FFM) (β = 0.000, p = 0.949) and bodyweight (BW) (β = 0.001, p = 0.831).

While overeating is a major contributor to obesity, athletes often face the opposite challenge meeting their high energy demands. Prolonged or frequent appetite suppression resulting in (long-term) reduced energy intake can lead to energy and nutrient deficiency (28). Chronic underfuelling, defined as energy availability below 30 kcal/kg FFM, has physiological and psychological consequences summarised as relative energy deficiency in sports (RED-S) (7, 8). Low energy intake in athletes is often assumed to be present due to tight schedules, lack of nutritional knowledge, body image, disordered eating behaviour, and external and internal pressure (29–31). Although there is recently some doubt about the conceptualisation of RED–S as a syndrome and its cause (32), Brooks et al. state: “Anecdotally, hunger disappeared in studies on 12-h fasted men given exogenous lactate infusion. Of note, athletes competing in 400–1,500 m runs that result in extraordinary lactatemia are seldom hungry immediately after hard training or competition” (33). Exercise–induced appetite suppression is generally assumed to occur at intensities above 60% V˙O2peak (23). The identification of lac-phe as a potential appetite–suppressing metabolite adds another possible explanation for the underfueling in athletes. Energy homeostasis and therefore energy intake is regulated by a complex interplay of physiological, behavioral, and environmental factors (10, 23). However, the causal relevance for humans remains uncertain. For instance, diet-induced obese mice lost weight after receiving an injection of 50 µg/kg BW lac-phe (5), which increased plasma concentrations more than 100-fold – far above physiological concentrations observed after exercise (34). The gastrointestinal tract alone secretes more than 20 peptide hormones that influence digestion, satiety, and energy balance, many of which are sensitive to nutrient content and regulate short-term hunger (35). The appetite-suppressing mechanism of lac-phe is still unclear. There are some hypotheses on lac-phe affecting the hypothalamus by passing through the blood-brain barrier whose permeability is increased by obesity and exercise and affects G-coupled receptors in the brain (34). As phenylalanine is affecting G-protein coupled receptors, lac-phe might just have a role as a blood-brain barrier crossing metabolite and, once having crossed dissolve into lactate and phenylalanine and might therefore be a factor just enriching cerebral concentrations of lactate and phenylalanine. Identifying the specific receptor(s) and target cell types is therefore essential to understand the potential role in appetite regulation of lac-phe (23). Exercise-induced increases in lactate are also known to influence gut hormones activated by the hypothalamic-pituitary axis, including acylated ghrelin, glucagon-like-peptide-1 (GLP-1), peptide YY (PYY), pancreatic polypeptide (PP) and cholecystokinin (CKK) (9, 10, 23, 33, 34). Previous research also shows suppression of acylated ghrelin and increasing levels of PYY, GLP-1, and PP for 2–9 h after exercise (36, 37). Among these, ghrelin appears to play a particularly pivotal role in mediating lactate-induced appetite suppression, especially within the first within 1–2 h after exercise (9, 10, 38, 39). To clarify lac-phe's relevance for exercise-induced appetite-suppression, and thus its potential contribution to low energy availability in athletes, future studies should assess gut hormone concentrations, feeling of satiety, energy expenditure, and energy intake simultaneously.