The study protocols were approved by the Swedish Ethical Review Authority (diary numbers 2021-02993 and 2022-01369-02), and except for being registered in a database, the study was executed in agreement with the ethical standards defined in the Declaration of Helsinki.

Eighteen healthy human volunteers (self-reported biological sex; female/male, 50% females) were recruited for this study, twelve participants to an experimental group (50% females) and six to a control group. To be considered for enrollment, volunteers were required to be 20–40 y/o, non-smokers and free from medical conditions and medications that could potentially influence the study outcomes (e.g., antidepressants) as well as not performing more than low levels of physical exercise. All participants were regarded as untrained to recreationally trained, performing zero to two recreational training sessions per week, involving resistance- and aerobic-type exercises, yoga, and various sports activities, e.g., badminton and football. Participant characteristics for the experimental and the control group are presented in Table 1. All participants provided oral and written informed consent prior to enrollment.

For all blood parameters, a randomized crossover-controlled design was employed in which each participant in the experimental group rested in a supine position for approximately 4 h on two different occasions while receiving a 60-min venous infusion of either sodium lactate (Lactate trial) or isotonic saline (Saline trial). The two experimental trials were separated by seven to 30 days. Blood was sampled on both occasions. In addition to blood samples, muscle biopsies were taken from participants in the experimental group, but only in the Lactate trial. To reduce the number of muscle biopsies for the participants in the experimental group, a smaller control group was also recruited (n = 6). This was warranted out of ethical aspects, especially as there was very little, if any, rationale that a saline infusion and repeated biopsies would induce changes in skeletal muscle relevant to the aims of this study. Each participant in the control group took part in an experimental trial where they donated muscle biopsies while receiving a venous infusion of isotonic saline (Control trial). The muscle data from the Control trial were subsequently used to compare with the muscle-specific response in the Lactate trial of the experimental group. An illustrative overview of the general study design is provided in (Figure 1).

All volunteers visited the laboratory for an information meeting and health screening to determine eligibility. If criteria were met, volunteers had their maximal oxygen uptake (VO₂max) determined with a graded exercise test on a treadmill using COSMED Quark CPET breath-by-breath analysis with Omnia software (COSMED The Metabolic Company, Rome, Italy). Volunteers were instructed to refrain from eating within 3 h prior to testing and to arrive well-rested and in good health. The test protocol began with a 5-min warm-up at an individually selected light pace, immediately followed by two 5-min stages at two different moderate intensities (0.5°incline). After a few minutes of active rest, the maximal test began at an individually selected running speed. Treadmill velocity was subsequently increased by 0.5–1.0 km/h every minute until the participant reached their highest sustainable running speed without significant deterioration in running form. Thereafter, the treadmill incline was increased by 1° for 1 min, followed by an additional 0.5° increase for each subsequent minute. The test continued until volitional exhaustion and participants were asked to report their rating of perceived exertion (RPE) using the Borg scale. VO₂max was accepted as valid if at least two of the following criteria were met: (a) a plateau in VO₂ despite an increase in workload, (b) RPE ≥ 18, and (c) a respiratory exchange ratio (RER) ≥ 1.1. The highest 45-s average of VO₂ was considered the participant’s VO₂max. Enrolled participants came back in a fasted state and had their body composition measured with dual x-ray absorptiometry (DXA) to assess their fat-free mass (FFM). All participants were instructed to avoid increasing their habitual amount of physical exercise throughout the study, and 2 days before each trial participants were told to fully refrain from physical exercise.

On the trial days, participants reported to the laboratory at 7:30 AM in an overnight fasted state. After arrival, participants took a supine position and had a 20- or 22-gauge Teflon catheter inserted into the antecubital vein of each forearm, one for infusion and one for repeated blood sampling. Following a 30-min rest, a baseline blood sample was obtained to confirm that fasting glucose, resting lactate, acid-base balance and electrolyte status were within established normal reference values. In the Lactate and Control Trials, a baseline muscle biopsy was then taken from the vastus lateralis muscle. The biopsy procedure was performed as described in Liegnell et al. (2020). In short, samples were taken under local anesthesia (Carbocaine 20 mg, Astra Zeneca AB, Sweden) using a Bergström needle (Stille, Torshälla, Sweden) with applied suction (Evans et al., 1982). After baseline sampling, a venous infusion of sodium lactate (APL, Stockholm, Sweden) or a volume-matched isotonic saline (0.9% sodium chloride) infusion (Fresenius Kabi AB, Uppsala, Sweden) was initiated using an automated infusion pump. The concentration of the sodium lactate solution was 1 M (pH 6.4) and the infusion rate was 125 μmol × kgFFM^–1 × min^–1 with a total infusion time of 60 min. Throughout the infusion blood was drawn every 10 min. Immediately following the termination of the infusion, a blood sample was collected, and a second muscle biopsy (Lactate- and Control trial) was taken from the same leg. Thereafter, the participant stayed in a supine position and blood samples were drawn after 15, 30, 45, 60, 90, and 120 min. Blood that was not used for immediate analyses was collected in heparinized- and serum-separating tubes. To prevent between-sample contamination, the catheter used for blood sampling was flushed with a saline solution between each blood sample. In the Lactate- and Control trials, additional biopsies were taken from the same leg at 60- and 120 min post-infusion. An illustrative overview of the experimental trial protocols is presented in Figure 1.

Whole blood was analyzed for levels of pH, base excess, sodium, potassium and hematocrit (HCT) using the handheld i-STAT1 blood analyzer with EG6 + cassettes (Abbot Laboratories, Chicago, IL, USA). Due to availability reasons whole blood levels of lactate were analyzed with either the Biosen C-line analyzer (EKF-Diagnostics, Cardiff, UK) or the Lactate Scout 4 (EKF-Diagnostics, Leipzig, Germany). Whole blood lactate measurements were primarily used for monitoring infusion progress during the trials. In the Saline trial, whole blood was only analyzed for pH, base excess, sodium, potassium and HCT at baseline, 30 and 60 min after infusion start, and 15-, 30-, 60-, and 120 min after the end of infusion.

Whole blood sampled in heparinized tubes was stored on ice until centrifuged at 3000 g at 4°C for 10 min. The upper- and bottom half of the plasma obtained was then transferred to new separate tubes, the bottom half saved as plasma and the upper half spun again at 3000 g at 4°C for 10 min. The top 75% was then collected and stored as platelet-poor plasma (PPP). Serum-separating tubes were kept agitated at room temperature for a minimum of 15 min. After centrifuging at 2000 g at 4°C for 10 min serum was transferred to new tubes. All plasma- and serum samples were stored at −80°C until analysis.

Immediately after collection, muscle samples were quickly blotted, cleared of blood and snap-frozen in liquid nitrogen before subsequent storage at −80°C. Following lyophilizing, samples were thoroughly dissected free from blood and connective tissue under a light microscope (Carl Zeiss, Germany) and organized into small bundles of fibers. The fiber bundles were then mixed and separated into aliquots. For immunohistochemistry, a piece of the sampled muscle was instantly cleared from blood and connective tissue and placed on a flat surface with the fibers pointing vertically. The sample was then embedded in an O.C.T. medium (Tissue-Tek O.C.T. Compound) and frozen in isopentane pre-chilled in liquid nitrogen. Samples were stored at −80°C until cryosectioning.

Plasma and muscle lactate levels were analyzed spectrophotometrically (Bergmeyer, 1974) using a microplate reader (Infinate 200 Pro, Tecan, Männedorf, Switzerland). To determine plasma lactate concentrations 10 μl plasma was added to a 48-well microplate. A glycine-hydrazine buffer (pH 8.8) consisting of 0.5 M glycine, 0.4 M hydrazine hydrate, 10 mm ethylenediaminetetraacetic acid (EDTA) and 500 ml distilled water (dH2O) was mixed thoroughly with nicotinamide adenine dinucleotide (NAD) and lactate dehydrogenase (LDH) to form the reaction solution. 770 μl reaction solution was subsequently added to each well, and following a 60 min incubation at room temperature, the absorption of the samples was read at 340 nm, and the concentrations were calculated with the following equation: (Absorbance – Blank) × μl in well (780 μl)/6.22 × μl added sample (10 μl) = mmol/L.

To determine muscle lactate levels, 2 mg of lyophilized muscle was transferred into Eppendorf tubes using a Mettler Toledo XA105 Dual Range Scale (d = 0.01 mg/0.1 mg). All samples were homogenized in trichloroacetic acid (TCA) for five min using a BulletBlender™ (Next Advanced, New York, USA) with zirconium oxide beads. Tubes were then centrifuged at 3000 g (4°C) for five min after which the solution was transferred to new tubes and mixed with 1 M potassium chloride. Thereafter, 25 μl of the muscle homogenate was pipetted into a 48-well microplate, followed by a 500 μl reaction solution with the same content as described above. Enzymatic reactions were then run for 60 min at room temperature, followed by absorbance readings at 340 nm. Lactate levels were quantified from the absorbance readings using the following equation: (Absorbance – Blank) × μl in well (525 μl)/6.22 × μl added sample (25 μl) × 100/mg muscle × 1.33 = mmol/kg dry muscle. The muscle levels in mmol/kg dry muscle (dry weight; d. w.) were also recalculated to mmol/L intracellular (i.c.) water in agreement with the data from Sjøgaard et al. (1985). For the plasma- and muscle lactate analysis, blanks, controls, and samples were assayed in duplicates, and the average readings were used for calculations.

Serum levels of cortisol were analyzed with the Cortisol ELISA-kit (CO368S) and performed following the directions for analyses provided by the manufacturer (Calbiotech Inc., El Cajon, CA, USA). A Wellwash™ Microplate Washer (Thermo Fisher Scientific, Waltham, MA, USA) was used for all assays and assays were run with standards, controls and samples in duplicates. Each participant’s samples from both trial conditions (Lactate and Saline trial) were analyzed on the same ELISA plate and the intra-assay coefficients of variation (CV%) for all samples were <9% averaging 2.17% ± 1.71% (SD). All ELISA plates, reagents and standards were sourced from a single batch (same lot number).

Plasma levels of pro-BDNF (pg × ml^–1), PPP-levels of mBDNF (pg × ml^–1) and serum levels of mBDNF (pg × ml^–1) were quantified with enzyme-linked immunosorbent assay (ELISA). The Human Pro-BDNF DuoSet ELISA kit (DY3175), in combination with the DuoSet Ancillary Reagent Kit 2 (DY008) from R&D Systems (Minneapolis, MN, USA), was used for pro-BDNF analyses. Assays were performed in accordance with the manufacturer’s instructions apart from a few adjustments. Standards and samples were diluted in sterile PBS, standards with 10% FBS + 0.02% Tween-20 and samples with 1% FBS, and then incubated in the coated plate overnight (16 h) at 4°C. All plasma samples used for pro-BDNF analysis were diluted by 20% in 1% FBS. To quantify mBDNF in PPP- and serum samples the Human Free BDNF Quantikine Immunoassay kit (DBD00) from R&D Systems was used. The assays were run according to the manufacturer’s directions with serum samples diluted 20-fold in the diluent provided with the kit. A Wellwash™ Microplate Washer (Thermo Fisher Scientific, Waltham, MA, USA) was used for all assays and the assays were performed with standards, controls and samples in duplicates. All samples from each individual participant were analyzed on the same ELISA plate for each of the three assays. The mean intra-assay coefficients of variation were 2.1% for pro-BDNF, 2.3% for PPP mBDNF and 2.3% for serum mBDNF. All ELISA components for each of the three assays were from the same production lot.

Immunoblotting was performed in accordance with the procedure described in Liegnell et al. (2020) with antibodies validated by Edman et al. (2024).

Cryosectioning and the subsequent immunohistochemical procedures for determination of fiber type composition and fiber cross-sectional area (fCSA) were performed as described previously (Horwath et al., 2020; Horwath et al., 2021a; Horwath et al., 2021b) and the percentage fiber type cross-sectional area (fCSA%) was calculated as explained in Horwath et al. (2021a). The fiber type characteristics for participants in the experimental and the control group are presented in Table 1.

For genotyping the Val66Met polymorphism in the human BDNF gene, a one-step amplified refractory mutation system (ARMS) polymerase chain reaction (PCR) was performed in agreement with the method outlined in Sheikh et al. (2010) and the PCR amplification run in CFX96 touch real-time PCR detection system (Bio-Rad Laboratories, Sundbyberg, Sweden). Prior to the ARMS-PCR, isolation of total DNA was completed using approximately 3 mg of lyophilized and dissected skeletal muscle tissue. The extraction was performed using the DNeasy Blood and Tissue kit (Cat. No. 69504) from Qiagen (Hilden, Germany) and the spin-column protocol for purification of total DNA from animal tissues. Electrophoresis was run on a 5% TBE polyacrylamide precast gel (Bio-Rad Laboratories, Mississauga, Canada) and then soaked with SYBR™ Safe DNA gel stain concentrate (Invitrogen, Thermo Fisher Scientific, Carlsbad, CA, USA) diluted in TBE buffer. PCR amplicons were subsequently documented with the ChemiDoc MP Imaging system (Bio-Rad).

The sample size in the experimental group (n = 12) was based on power calculations from a pilot experiment in four individuals (data not presented). A minimum sample size of n = 8 was estimated based on a statistical power of 0.80 (α = 0.05), using Cohen’s d = 0.4, which corresponds to an anticipated mean difference of approximately 50% in mBDNF levels, based on mBDNF data from Edman et al. (2024). To account for potentially smaller effects, four additional participants were enrolled in the experimental group to ensure sufficient power. Data were analyzed using TIBCO Statistical 13 for Windows (TIBCO Software Inc., Palo Alto, CA). The data are presented as mean ± standard deviation (SD). Weighted means were used to replace two missing blood and plasma lactate values before analysis. The normal distribution was assessed with histograms and the Shapiro-Wilk test of normality. Following log transformation of the cortisol data, all variables were considered acceptable for parametric statistics. However, the results present the untransformed means to facilitate the interpretation of the data. A two-way repeated measures analysis of variance (ANOVA; trial × time) was used for the data analysis on circulating levels of plasma- and serum mBDNF, pro-BDNF, and cortisol (2 × 5), plasma lactate (2 × 11), whole blood lactate (2 × 13) as well as sodium, pH, base excess, potassium and HCT (2 × 7). If the ANOVA showed a significant main effect of time or interaction effect, a Tukey’s Honest Significant Difference (HSD) test was performed. Muscle BDNF levels were analyzed using a two-way ANOVA (group × time) with four time points (baseline, post-infusion, and 60- and 120 min post-infusion) to compare the muscle data from the Lactate trial of the experimental group (n = 12) with the muscle data from the control groups saline trial (n = 6). Correlations between fiber type and muscle pro-BDNF levels as well as between cortisol and plasma mBDNF were calculated with Pearson’s product-moment correlation (r). Welch’s t-test was conducted to assess differences in BDNF levels between participants with and without the BDNF Val66Met polymorphism. Additionally, Welch’s t-test was used to compare baseline descriptive variables between the experimental and control groups to confirm group homogeneity (Table 1). A significance level of α = 0.05 was employed for all statistical tests.

Lactate infusion continuously increased blood and plasma levels of lactate over the 60-min infusion protocol, peaking at 5.4 ± 0.36 and 5.9 ± 0.37 mmol × L^–1, respectively (Figures 2A, B). Immediately upon infusion cessation in the Lactate trial, blood and plasma lactate decreased from peak levels (p = 0.0002; 75 min vs. 60 min) and returned to baseline levels following 120 min of recovery (180 min vs. Pre, p > 0.8; Figures 2A, B). Blood and plasma lactate levels remained unaltered in the Saline trial throughout the protocol (p > 0.05 vs. Pre; Figures 2A, B). At baseline, average plasma lactate levels were 60% greater than mean whole blood levels; however, the difference was reduced to 10% when peak lactate levels from the infusion were reached. Lactate infusion increased muscle lactate levels by 29%, from 11.4 ± 4.88 to 14.8 ± 2.86 mmol × kg^–1 d. w. (Figure 2C), or 3.07 to 3.98 mmol/L i.c. water (p = 0.0044 vs. Pre). Muscle lactate levels remained unchanged with saline infusion (p = 0.5488 vs. Pre; Figure 2C).

Lactate infusion had an alkalizing effect on blood, increasing pH from baseline levels (7.35 ± 0.02) to 30 min into the infusion, and then remaining elevated throughout the trial (7.42 ± 0.03–7.49 ± 0.03, p < 0.0001 vs. Pre; Figure 3A), while saline infusion had no effect on pH (p > 0.0547 vs. Pre; Figure 3A). The blood base excess followed the changes in pH for both trials (Figure 3B). Blood levels of sodium increased 30 min into the lactate infusion and remained elevated throughout the trial (141.9 ± 1.0–144.0 ± 1.28 mmol × L^–1, p < 0.0001 vs. Pre, Figure 3C). Saline infusion did not affect blood sodium levels in the Saline trial (p > 0.9848 vs. Pre, Figure 3C). 15 min after the end of the lactate infusion (min 75) and throughout the recovery, blood potassium levels decreased below baseline levels (3.83 ± 0.15 mmol × L^–1) to between 3.55 ± 0.28 and 3.22 ± 0.13 (p < 0.0005 vs. Pre; Figure 3D). By contrast, saline infusion increased blood potassium levels from the same time point, to between 4.07 ± 0.4 and 4.14 ± 0.41 mmol × L^–1 (p < 0.0011 vs. Pre; Figure 3D), with levels having returned to baseline at the end of the trial (p = 0.1067; 180 min vs. Pre). The lactate infusion also lowered blood HCT levels to 36.0 ± 3.02% (p < 0.0001 vs. Pre; Figure 3E), partly recovering to 38.0 ± 3.36% during the subsequent recovery period (p = 0.1044 vs. Pre; Figure 3E). Blood HCT levels remained unchanged in the Saline trial (p > 0.3623 vs. Pre). Serum cortisol levels fell between 19- and 30% during the lactate and saline infusion, respectively (p < 0.0241 vs. Pre; Figure 3F). In the Saline trial, cortisol levels remained depressed throughout the recovery (p < 0.0002 vs. Pre), while cortisol returned to baseline levels in the Lactate trial at 15- and 30 min post infusion (p > 0.2615). At the same time-points, a difference between the trials was evident (min 75 and 90; p < 0.0159 for Lactate vs. Saline; Figure 3F).

In the Lactate trial, plasma pro-BDNF levels were unaltered during the infusion, but increased by 50% from 25.58 ± 16.6 to 38.58 ± 30.0 pg × ml^–1 an hour after infusion cessation (min 120 vs. Pre, p = 0.0361; Figure 4A). Plasma pro-BDNF levels were higher during the recovery period in the Lactate trial compared to the Saline trial (min 75-, 90 and 120; p < 0.0286; Figure 4A). Pro-BDNF levels remained unchanged in the Saline trial (p > 0.9999 vs. Pre; Figure 4A). Both infusion protocols affected PPP mBDNF levels similarly, with a distinct 31%–51% decrease from baseline levels seen at the end of the infusion (p < 0.0002 vs. Pre; Figure 4B). Post hoc analysis revealed no time-point specific differences between trials, despite a significant interaction effect (p = 0.0302; Figure 4B). The two infusion protocols also had a similar effect on serum mBDNF levels, with levels decreasing by 20%–23% from baseline to the end of infusion (p < 0.0014 vs. Pre; Figure 4C). Instead of remaining decreased during the recovery from infusion, as noticed for plasma mBDNF levels, serum mBDNF levels returned to baseline 15 min after infusion cessation and stayed there until trial end (p > 0.4932 vs. Pre for Lactate and Saline trial; Figure 4C). Neither infusion protocol affected muscle levels of pro-BDNF, as it remained unchanged during the infusion- and recovery period (p > 0.0895 for Main- and Interaction effects; Figure 4D). Muscle levels of mBDNF were non-detectable with immunoblotting (data not shown).

A significant positive correlation was seen between muscle pro-BDNF expression and type I fCSA% (n = 18, r = 0.6746, p = 0.0021, Figure 5). Genotyping results showed allele-specific bands for two genotypes, Val/Val and Val/Met, with no participant showing the mutation in both alleles (Met/Met). The BDNF Val66Met polymorphism was expressed in 7 out of 18 participants (39%), with a similar distribution in genotype seen within the experimental-and control group (42 and 33% respectively, Table 1). No differences in pre- or delta BDNF values were detectable between participants with and without the Val66Met polymorphism in the experimental group (n = 12, p > 0.05) for any of the BDNF variables. Likewise, when dividing all participants (n = 18) based on BDNF genotype, no difference was seen in muscle pro-BDNF levels between Val/Val- and Val/Met groups (data not presented).