Locomotion-specific metabolic demands in 30-s anaerobic efforts

Abstract

Purpose

This study aimed to examine whether energy system contributions differ across locomotion modalities during 30-s all-out exercise.

Methods

Sixteen trained male team-sport athletes completed a Wingate Anaerobic Test (WAnT), a 30-s continuous jumping test (CJ30), and a 30-s all-out running test (RUN30) in randomized order. Aerobic, glycolytic, and phosphagen energy contributions were estimated using the PCr–La–O₂ method based on oxygen uptake, blood lactate accumulation, and excess post-exercise oxygen consumption.

Results

Total energy expenditure was higher in RUN30 (130.8 ± 18.0 kJ) and WAnT (125.6 ± 14.9 kJ) than in CJ30 (95.4 ± 19.6 kJ), and glycolytic energy contribution was also greater in RUN30 (55.2 ± 9.4 kJ) and WAnT (54.3 ± 6.3 kJ) than in CJ30 (26.7 ± 7.9 kJ). Phosphagen energy was higher in RUN30 (49.6 ± 7.5 kJ) than in CJ30 (43.3 ± 10.1 kJ; ηp² = 0.240), with no difference between RUN30 and WAnT. Aerobic energy contribution did not differ across modalities. Blood lactate changes were markedly greater after RUN30 (13.0 ± 1.6 mmol L⁻¹) and WAnT (12.8 ± 1.2 mmol L⁻¹) than after CJ30 (5.9 ± 1.7 mmol L⁻¹). Overall, significant modality effects were observed for total energy expenditure, glycolytic and phosphagen energy contributions, and blood lactate response (all p < 0.001 except phosphagen, p = 0.029), whereas aerobic energy contribution did not differ between modalities (p = 0.425).

Conclusion

These findings indicate modality-associated differences in absolute energy demands, with greater glycolytic and total energy expenditure in running and cycling than jumping.

1. Introduction

Short-duration maximal efforts of approximately 30 s represent a commonly used experimental model to examine the interaction of anaerobic and aerobic metabolism under conditions relevant to intermittent team sports, where repeated high-intensity actions are performed with limited recovery.^1,2 During such efforts, energy provision emerges from a dynamic interaction between phosphagen degradation, rapid glycolytic flux, and progressively increasing oxidative metabolism.^3,4 Although anaerobic pathways dominate the initial phase, aerobic contribution becomes increasingly relevant as exercise duration approaches 30 s due to the finite capacity of phosphocreatine stores and accelerated glycolytic activation.^5,6 Importantly, oxygen uptake kinetics during short-duration maximal exercise are constrained by time constants and measurement resolution, potentially limiting the observable magnitude of oxidative contribution despite meaningful metabolic involvement.^7,8 Beyond exercise duration alone, locomotion modality may further modulate metabolic demand, as differences in muscle recruitment patterns, contraction type, and mechanical constraints have been shown to influence oxygen uptake, lactate accumulation, and total energy expenditure across exercise.9, 10, 11 Accordingly, the metabolic profile of a 30-s maximal effort cannot be assumed to be uniform across different locomotor tasks.

Among laboratory-based assessments, the Wingate Anaerobic Test (WAnT) and the Bosco Continuous Jumping Test (CJ30) are widely used to characterize anaerobic performance in cycling and jumping modalities, respectively.^12,13 Their metabolic profiles have been extensively investigated using multi-component models integrating oxygen uptake, blood lactate accumulation, and post-exercise oxygen kinetics.^5,6,14,15 These studies consistently demonstrate substantial glycolytic and phosphagen contributions during 30-s maximal efforts, alongside a measurable oxidative component that varies according to mechanical constraints and contraction characteristics. Notably, modality-specific differences have been reported even under time-matched conditions, suggesting that external load configuration and movement structure may meaningfully alter pathway partitioning.^6,10,15 In contrast, running-based anaerobic assessments are typically structured around fixed-distance sprints (e.g., 100 m or 200 m), where exercise duration varies across individuals and performance levels.^1,16 This variability complicates direct metabolic comparisons with duration-standardized protocols such as WAnT and CJ30 and limits the ability to isolate locomotion-specific energetic characteristics under controlled temporal conditions. Consequently, despite the ecological relevance of running in team sports, direct comparative data on the metabolic profile of duration-matched all-out running relative to established cycling and jumping protocols appear to remain limited. Accordingly, the metabolic profile of a time-matched 30-s all-out running effort remains insufficiently characterized within a cross-modal comparative framework, and, to our knowledge, no previous study has applied PCr–La–O₂ modeling to a duration-matched 30-s running effort alongside established cycling and jumping protocols.

From a biomechanical perspective, time-matched locomotor tasks may differ substantially in force application patterns, contraction types, duty cycles, and external constraints. Cycling during WAnT is performed at fixed resistance, with predominantly concentric muscle actions and a force–velocity relationship constrained by the ergometer.⁹ In contrast, CJ30 consists of repeated stretch–shortening cycle (SSC) actions with brief ground contact phases and partial reutilization of elastic energy, potentially altering the relative reliance on glycolytic and phosphagen pathways.^6,17 Running represents a cyclic SSC-based locomotion that alternates eccentric and concentric contractions across a larger active muscle mass and is mechanically less constrained than ergometer-based cycling.^9,18 Moreover, acceleration dynamics and pacing freedom during overground running may influence early phosphagen utilization and subsequent glycolytic engagement. Taken together, these modality-specific characteristics suggest that differences in energy system partitioning may reflect both locomotion modality and inherent mechanical constraints.

To address this gap, the present study introduces a 30-s all-out running protocol (RUN30) aligned with the duration of WAnT and CJ30, thereby enabling comparison of metabolic responses across mechanically distinct locomotion modalities under standardized temporal conditions. Energy system contributions were examined within the framework of the PCr–La–O₂ model, which integrates oxygen uptake, blood lactate accumulation, and excess post-exercise oxygen consumption to partition phosphagen, glycolytic, and aerobic components.^5,14,19,20 This model has demonstrated acceptable validity and reliability in cycling- and jumping-based exercise,²¹ providing a structured framework for cross-modal comparison. The primary aim of this study was to compare absolute and relative energy system contributions during 30-s maximal efforts performed in running, cycling, and jumping modalities.

We hypothesized that (i) total energy expenditure would differ between modalities due to differences in contraction characteristics, particularly stretch–shortening cycle (SSC)-dominant versus predominantly concentric actions, together with differences in mechanical constraint and muscle recruitment patterns; (ii) RUN30 would elicit greater absolute phosphagen contribution than CJ30 given its continuous force application characteristics and acceleration-related mechanical demand; and (iii) aerobic contribution would remain comparable across modalities, reflecting shared oxygen uptake kinetics during short-duration maximal exercise.⁴ By characterizing metabolic responses under duration-matched but mechanically distinct conditions, this investigation aims to contribute to the interpretation of anaerobic testing across both laboratory and field-based settings.

2. Methods

2.1. Study design

This study utilized a cross-sectional design to compare the metabolic profiles of 30-s all-out running, cycling, and jumping in trained male team-sport athletes, conducted following their competitive season. All participants performed WAnT,¹² CJ30,¹³ and RUN30, with tests administered in a randomized, counterbalanced sequence and separated by 3–4 days of recovery to minimize fatigue effects. Unlike traditional running assessments based on distance covered, RUN30 was developed for this study. The protocol was designed as a fixed-duration, overground maximal running effort in which participants were instructed to cover the greatest possible distance within 30 s. The 30-s duration was selected not only to permit direct temporal matching with the established WAnT and CJ30 protocols, but also because this duration represents a commonly used experimental window for examining the interaction of phosphagen, glycolytic, and oxidative energy provision during short-duration maximal exercise.4 In contrast to fixed-distance sprint tests, in which exercise duration varies according to individual performance, the fixed-duration design of RUN30 was intended to standardize exposure time across participants and modalities. This time-matched approach therefore enabled a more controlled comparison of energy system contributions across running, cycling, and jumping while reducing the confounding effect of inter-individual differences in exercise duration.⁴

Participants were instructed to refrain from strenuous physical activity for at least 48 h prior to each testing session, avoid alcohol for 24 h, and abstain from caffeine intake for at least 12 h before testing, in accordance with established exercise testing guidelines. Nutritional intake was standardized by instructing participants to consume their final meal at least 3 h before testing and to replicate their habitual dietary pattern across sessions. Water intake was permitted ad libitum, while no ergogenic supplements or medications known to influence metabolic responses were allowed during the study period.

A familiarization session was conducted at least 48 h before the first experimental trial to accustom participants to all procedures and equipment, including the gas analyzer mask and maximal-effort protocols. Each experimental session began with a standardized 10-min warm-up consisting of light jogging and dynamic stretching, followed by brief task-specific efforts (5-m sprint, three maximal jumps, and a 3-s cycling sprint). Tests commenced 3 min after warm-up completion, and strong verbal encouragement was provided to ensure maximal effort.

All laboratory-based tests (WAnT and CJ30) were performed indoors under controlled environmental conditions (∼22°C; ∼45% relative humidity), whereas RUN30 was conducted on a standard synthetic 400-m outdoor track under stable weather conditions (∼20°C; ∼50% relative humidity). All tests were scheduled between 15:00 and 18:00 to minimize circadian variation in physiological responses.

Participants wore standardized athletic clothing and the same type of running shoes across sessions to reduce biomechanical variability. For metabolic measurements, a properly fitted face mask was secured according to manufacturer guidelines to ensure airtight sealing.

Anthropometric measurements were collected on the first test day: height using a portable stadiometer (Holtain Ltd., UK) and body mass/composition via air displacement plethysmography (Bod Pod, Cosmed, USA; CV = 0.15%, ICC = 0.996, SEM = 0.001 kg L⁻¹; calibrated per manufacturer guidelines.²² Blood lactate concentration ([BLa]) was assessed from fingertip capillary samples (Lactate Pro 2, Arkray, Japan; CV = 3.3%) at rest and 1, 3, 5, 7, and 10 min post-exercise.

$\dot{\mathrm{V}}$ O₂ was measured breath-by-breath with a portable metabolic gas analyzer (K5, Cosmed, Italy; error < 1.5%)²³ 10 min pre-exercise, during the test, and 15 min post-exercise. The portable gas analyzer was calibrated before each session according to the manufacturer's instructions using a 3-L syringe for flow/volume calibration and standard reference gases of known concentration for O₂ and CO₂ calibration.

All testing procedures were supervised by two PhD-level exercise physiologists with more than five years of experience in metabolic and performance testing. Blood sampling and data collection were performed by trained research personnel, and the same investigator oversaw all sessions to ensure procedural consistency (see Fig. 1).

Fig. 1.

Overview of the study design, randomization procedure, and testing timeline.

2.2. Participants

Eighteen male athletes classified as Trained/Developmental (Tier 2) according to McKay et al.‘s (2021) Participant Classification Framework were recruited from competitive team sports, including football (n = 6), handball (n = 4), ice hockey (n = 4), and basketball (n = 4). This classification reflects athletes engaged in structured, coach-supervised training and regular competition, performing 6–9 h of sport-specific training per week for a minimum of three consecutive years, with experience in regional and national-level leagues but not competing at an elite international level.²⁴ Participants had a mean age of 21.8 ± 1.9 years, height of 179.4 ± 6.3 cm, body mass of 75.2 ± 7.8 kg, and BMI of 23.4 ± 1.6 kg/m².

Inclusion criteria required participants to (i) be free from musculoskeletal injury within the previous six months, (ii) have no diagnosed cardiovascular, metabolic, or respiratory disorders, and (iii) be accustomed to maximal-intensity exercise testing. Exclusion criteria included contraindications to maximal effort testing, current orthopedic injuries, use of supplements that affect metabolism, or ongoing medical treatment.

All participants were recruited through sports clubs in collaboration with our institution and provided written informed consent after being informed of the study procedures, risks, and benefits. The study received approval from the Ethics Committee of Atatürk University, Sport Sciences Faculty (Approval No: E−70400699-050.02.04-2200126915/2022/04 Date: 21.04.2022), ensuring compliance with the ethical standards stipulated by the Declaration of Helsinki.

Two participants (one football, one handball) withdrew for personal reasons, resulting in a final cohort of 16 athletes (football: n = 5; handball: n = 3; ice hockey: n = 4; basketball: n = 4). A per-protocol analysis was applied to ensure accurate representation of metabolic and performance outcomes.

An a priori power analysis was conducted using G*Power (version 3.1) to determine the required sample size. A repeated-measures ANOVA (within factors) design was selected with one group and three measurements (WAnT, CJ30, RUN30). The median effect size reported by Kaufmann et al. (2021) was converted to an effect size of f = 0.41 for the within-subject model. With α = 0.05, desired power (1−β) = 0.90, correlation among repeated measures assumed at 0.50, and nonsphericity correction ε = 1.0, the analysis yielded a required total sample size of 15 participants (critical F = 3.34; numerator df = 2; denominator df = 28; actual power = 0.92). To account for potential attrition due to the physically demanding protocol, we oversampled by approximately 20% and recruited 18 participants. Sixteen athletes completed the study, exceeding the minimum required sample size and maintaining the predetermined statistical power.

2.3. Anaerobic tests

Three 30-s all-out tests were administered to assess anaerobic performance across distinct locomotion types: cycling, jumping, and running. The protocols and measurement techniques for each test are detailed below.

• •Wingate Anaerobic Test (WAnT): Conducted on a cycle ergometer (Peak Bike 894e, Monark, Sweden), with a resistance load set at 10% of body mass. This higher load, calculated prior to testing, is recommended for trained athletes to optimize power output.^25,26 Peak power and mean power were recorded using Monark software.
• •Bosco Continuous Jumping Test (CJ30): Performed using photoelectric cell technology (Optojump Next, Microgate, Italy), which provides high reliability (ICC = 0.982–0.989, CV = 2.7%, random error ±2.81 cm).²⁷ Participants executed continuous maximal jumps for 30 s, with jump height, peak power, and mean power measured per Bosco et al. (1983).
• •30-Second All-Out Running Test (RUN30): Conducted on a 400-m running track (polyurethane surface), participants were instructed to cover the maximal distance possible in 30 s. Starting 50 cm behind a photocell gate (Fusion Sport SmartSpeed, Brisbane, Australia), timing began automatically upon crossing, and distance was determined via video analysis by marking the foot's position at the 30-s mark (version 2.3.2, Kinovea open-source project). Velocity–time profiles and split distances were not recorded; therefore, RUN30 performance was quantified only as the total distance covered within 30 s. Consequently, pacing strategy and acceleration characteristics could not be analyzed. Given that RUN30 was initiated from a standing start and inherently included an acceleration phase, the absence of velocity–time and split-distance data represents an important methodological limitation, as the contribution of acceleration dynamics and self-selected pacing to the observed metabolic responses could not be quantified.

2.4. Calculation of the energy pathway contributions

Contributions from aerobic, glycolytic, and phosphagen energy pathways were estimated using the PCr-La-O₂ method.^5,14,21 $\dot{\mathrm{V}}$ O₂ data for excess post-exercise oxygen consumption (EPOC) modeling were preprocessed as described.⁷ [BLa] was measured from 20-μL fingertip capillary samples (left hand) using a portable analyzer at rest and 1, 3, 5, 7, and 10 min post-exercise, with the fingertip cleaned with alcohol and dried to avoid perspiration effects. Delta blood lactate concentration (Δ[BLa]) was calculated as the difference between peak and resting values.

Aerobic Energy Contribution: Derived from $\dot{\mathrm{V}}$ O₂ above resting levels during exercise. Resting $\dot{\mathrm{V}}$ O₂ was averaged from the last 5 min of the 10-min pre-exercise measurement. Aerobic energy was calculated as the area under the curve of exercise $\dot{\mathrm{V}}$ O₂ minus resting $\dot{\mathrm{V}}$ O₂, using the equation:

$$\text{Aerobic}\text{Energy}\left(\text{kJ}\right)=\dot{\mathrm{V}}{\mathrm{O}}_2(\text{ml}·{\text{kg}}^{-1})·\text{caloric}\text{equivalent}(\mathrm{J}·{\text{ml}}^{-1})·\text{body}\text{mass}\left(\text{kg}\right)·{10}^{-3},\text{where}\text{the}\text{caloric}\text{equivalent}\text{was}20.9\mathrm{J}·{\text{ml}}^{-14}\text{.}$$

Glycolytic Energy Contribution: Estimated from Δ[BLa] using an O₂-lactate equivalent of 3.0 ml kg⁻¹ mM⁻¹, assuming 1 mM lactate accumulation equates to 3 ml O₂ per kg body mass and a lactate distribution space of ∼45% of body mass.³ These coefficients represent established assumptions within the PCr–La–O₂ approach rather than direct measurements of glycolytic energy production, and their use should therefore be interpreted within the recognized methodological limitations of indirectly estimating glycolytic contribution from blood lactate accumulation. Calculations followed:

Δ[BLa] (mM) = Peak [BLa] (mM) – Resting [BLa] (mM)

Glycolytic Energy (kJ) = Δ[BLa] (mmol·L⁻¹) × O₂-lactate equivalent (ml·kg⁻¹·mmol⁻¹) × caloric equivalent (J·ml⁻¹) × body mass (kg) × 10⁻³, with the caloric equivalent set at 20.9 J ml⁻¹.

Phosphagen Energy Contribution: The estimation of phosphagen-derived energy was based on the rapid component of excess post-exercise oxygen consumption (EPOC), which represents the oxygen cost associated with ATP and phosphocreatine (PCr) restoration during the initial phase of recovery. To characterize this response, post-exercise $\dot{\mathrm{V}}$ O₂ data were fitted using a bi-exponential model. Curve fitting was carried out primarily through a custom Python routine developed for this study, while OriginPro software (v9.0, OriginLab, USA) was used as a secondary analytical reference 14,28.

$\dot{\mathrm{V}}$ O_{2 (t)} = $\dot{\mathrm{V}}$ O_2baseline + A₁ [e^{– (t – td)/τ1}] + A₂ [e^{– (t – td)/τ2}], where $\dot{\mathrm{V}}$ O_2(t) is oxygen uptake at time t, $\dot{\mathrm{V}}$ O_2baseline is the asymptotic baseline, A is amplitude, td is time delay, τ is the time constant, and subscripts 1 and 2 denote fast and slow components, respectively. Fast EPOC was calculated as EPOCfast = A₁ × τ₁, where A₁ represents the amplitude of the fast recovery component and τ₁ represents its time constant, and phosphagen energy as:

Phosphagen Energy (kJ) = EPOCfast (ml·kg⁻¹) × caloric equivalent (J·ml⁻¹) × body mass (kg) × 10⁻³, using 20.9 J ml⁻¹.

Total energy expenditure was the sum of aerobic, glycolytic, and phosphagen contributions, with anaerobic energy defined as the sum of glycolytic and phosphagen components. A constant caloric equivalent of 20.9 J ml⁻¹ O₂ was applied across modalities, consistent with previous PCr–La–O₂ modeling studies, and based on the assumption that high-intensity, short-duration maximal efforts are predominantly supported by carbohydrate oxidation, resulting in a respiratory exchange ratio approximating 1.0.

2.5. Statistical analysis

Exponential models for EPOC kinetics were generated using OriginPro as a reference tool, complemented by a custom-developed Python-based algorithm that incorporated user-defined $\dot{\mathrm{V}}$ O₂baseline values for enhanced modeling accuracy.²⁸ All other statistical analyses were performed with SPSS (v25.0, IBM Corp, Chicago, IL, USA).^28,29 An alpha level of 0.05 was set for significance. Data normality was assessed using the Shapiro–Wilk test, which confirmed that the data were suitable for parametric analysis, and results are presented as mean ± standard deviation. Sphericity was assessed via Mauchly's test; violations were adjusted with Greenhouse-Geisser correction (ε < 0.75) or Huynh-Feldt correction (ε > 0.75) applied to degrees of freedom. A one-way repeated-measures ANOVA with Bonferroni correction (adjusted α = 0.0167 for three-way comparisons) compared energy contributions and performance variables across exercise modalities. Effect sizes were estimated using partial eta-squared, interpreted as small (0.01), medium (0.06), or large (0.14).³⁰

3. Results

WAnT exhibited a relative peak power of 11.9 ± 1.08 W kg⁻¹ and a mean power of 8.62 ± 0.69 W kg⁻¹, while CJ30 showed a higher relative peak power of 25.3 ± 3.26 W kg⁻¹ and mean power of 20.6 ± 2.94 W kg⁻¹, alongside a mean jump height of 24.3 ± 4.25 cm. RUN30 recorded a covered distance of 204.4 ± 11.9 m.

Absolute energy contributions are shown in Table 1. Total energy expenditure differed significantly between modalities (p < 0.001), with higher values in RUN30 and WAnT compared to CJ30.

Table 1.

Physiological and metabolic responses during 30-second maximal running, cycling, and jumping.

	CJ30	WAnT	RUN30	F	p	${\mathit{\eta }}_{\mathit{p}}^{\mathbf{2}}$
Aerobic energy (kJ)	25.4 ± 4.1	26.5 ± 2.4	26.3 ± 3.6	0.881	0.425	0.055
Glycolytic energy (kJ)	26.7 ± 7.9	54.3 ± 6.3*	55.2 ± 9.4*	145.2	p < 0.001	0.906
Phosphagen energy (kJ)	43.3 ± 10.1	44.7 ± 10.1	49.6 ± 7.5*	4.729	0.029	0.240
Aerobic energy (%)	27.0 ± 4.2	21.3 ± 2.7*	20.1 ± 2.0*	42.15	p < 0.001	0.738
Glycolytic energy (%)	27.9 ± 4.6	43.4 ± 3.6*	42.0 ± 2.9*	102.6	p < 0.001	0.872
Phosphagen energy (%)	45.0 ± 5.7	35.3 ± 4.2*	37.9 ± 2.8*	27.91	p < 0.001	0.650
Total energy (kJ)	95.4 ± 19.6	125.6 ± 14.9*	130.8 ± 18.0*	61.63	p < 0.001	0.804
Δ[BLa] (mmol·L−1)	5.9 ± 1.7	12.8 ± 1.2*	13.0 ± 1.6*	217.6	p < 0.001	0.936

Abbreviations: WAnT, Wingate Anaerobic Test; CJ30, Bosco Continuous Jumping Test; RUN30, 30-s all-out running test; Δ[BLa], change in blood lactate concentration. Note: Values are mean ± SD. *p < 0.05 compared to CJ30 (Bonferroni-adjusted post-hoc test). No significant differences were observed between WAnT and RUN30 in any variable.

Glycolytic energy contribution also differed significantly between modalities (p < 0.001), with RUN30 and WAnT showing higher values than CJ30. Phosphagen energy contribution was significantly greater in RUN30 than in CJ30 (p = 0.029), whereas no difference was observed between RUN30 and WAnT. There were no significant differences in aerobic energy contributions across modalities (p = 0.425). Blood lactate changes (Δ[BLa]) were significantly greater in RUN30 and WAnT than in CJ30 (p < 0.001).

Relative energy contributions (% of total) are also shown in Table 1, indicating higher glycolytic share in WAnT and RUN30, while CJ30 exhibited a relatively higher phosphagen and aerobic percentage. However, due to differences in total energy output, these relative patterns should be interpreted with caution.

Fig. 2 displays the absolute contributions of aerobic and anaerobic (glycolytic + phosphagen) systems. Anaerobic energy was greater in RUN30 and WAnT than in CJ30, whereas aerobic energy contribution was similar across modalities.

Fig. 2.

The absolute energy contributions of aerobic and anaerobic (Glycolytic + Phosphagen) systems (in kilojoules) for different exercise types of the same duration (30 s).

Note: *p < 0.05 significantly different from CJ30.

4. Discussion

This study compared absolute and relative energy system contributions during 30-s maximal efforts performed in running, cycling, and jumping modalities under duration-matched conditions. Consistent with our first hypothesis, total energy expenditure differed between modalities, with RUN30 and WAnT eliciting greater absolute energy output than CJ30. In line with our second hypothesis, RUN30 demonstrated a higher absolute phosphagen contribution than CJ30, whereas no difference was observed between RUN30 and WAnT. As hypothesized, aerobic energy contribution did not differ significantly across modalities, supporting the premise that oxygen uptake kinetics during short-duration maximal exercise may exhibit comparable overall magnitude despite mechanical differences. Collectively, these findings indicate that mechanically distinct locomotion modalities are associated with different absolute anaerobic demands even when exercise duration is standardized.

Absolute and relative energy contributions provide complementary but distinct information. CJ30's lower total energy expenditure resulted in proportionally higher phosphagen contribution despite its lower absolute phosphagen energy compared with RUN30, illustrating the dependence of percentage-based values on total energetic output. Thus, absolute values primarily reflect the magnitude of energetic demand, whereas relative values describe the proportional distribution of individual pathways within that demand. RUN30, with higher absolute glycolytic and phosphagen contributions, aligns with sports requiring high total energy turnover over 30 s, such as football, handball, or rugby.^2,31,32 Accordingly, its greater absolute phosphagen contribution compared with CJ30 should be interpreted as a higher overall anaerobic demand rather than simply as a difference in proportional pathway contribution.

RUN30 may have differed metabolically from the other protocols partly because running engages a large active muscle mass and involves cyclic SSC actions across consecutive strides. In this cycle, muscles first undergo eccentric lengthening followed by rapid concentric shortening, which characterizes the alternating braking–propulsive phases of ground contact during running. This makes running different from cycling, which mainly relies on muscles shortening without much stretching, and jumping, which depends heavily on the elastic energy stored in muscles and tendons to spring back. Previous studies back up these differences.^9,18,33 Moreover, biopsy-based evidence from sprint exercise demonstrates that phosphocreatine (PCr) availability is a critical determinant of power output during the initial phase of maximal efforts, and incomplete PCr resynthesis substantially impairs early sprint performance.³⁴ Collectively, these mechanical and neuromuscular characteristics may influence the magnitude and distribution of energy pathway contributions, suggesting that muscle recruitment patterns and contraction dynamics are important determinants of metabolic partitioning.

The greater glycolytic energy contribution observed in WAnT and RUN30 compared to CJ30 aligns with previous comparisons between cycling- and jumping-based 30-s protocols.^6,35 Glycolytic metabolism constitutes a substantial component of high-intensity, short-duration exercise, and its magnitude appears to be influenced by task-specific mechanical characteristics. In the present study, both WAnT and RUN30 required relatively continuous force production across the full 30-s duration, which may have contributed to the higher glycolytic estimates, as reflected in the higher post-exercise blood lactate concentrations.

In contrast, CJ30 involves repeated SSC actions, where elastic energy storage and reutilization during ground contact may partially offset the metabolic demand of force generation.^13,17 The intermittent nature of these contractions, combined with elastic recoil mechanisms, could contribute to the relatively lower glycolytic contribution observed in this modality. Taken together, these findings suggest that anaerobic energy partitioning during time-matched maximal efforts may be influenced by locomotion modality, task-specific mechanical constraints, contraction dynamics, and pacing characteristics, although further investigation across different populations and performance contexts is warranted.

Aerobic energy contribution did not differ significantly among WAnT, CJ30, and RUN30, which is consistent with previous findings in ∼30-s maximal exercise. Previous research indicates that during such efforts, oxygen uptake kinetics are influenced more by individual aerobic capacity than by locomotion modality.⁴ Although aerobic metabolism represents a smaller fraction of total energy supply in these conditions, it provides a measurable oxidative contribution even during maximal intensity. In the present study, absolute aerobic energy contribution was approximately 26 kJ across all modalities, aligning with values reported in comparable protocols.⁵ This consistency may reflect the trained status of the participants, who were accustomed to intermittent high-intensity efforts.

However, given the brief 30-s duration and the known time constant of pulmonary $\dot{\mathrm{V}}$ O₂ kinetics, the measurable oxidative contribution may have been partially constrained by the inherent delay between intramuscular oxygen utilization and pulmonary gas exchange responses.^7,28,36 In addition, breath-by-breath measurement and subsequent signal smoothing may attenuate rapid fluctuations in $\dot{\mathrm{V}}$ O₂, potentially reducing sensitivity to subtle modality-specific differences. Therefore, the observed similarity in aerobic contribution should be interpreted not solely as physiological equivalence, but also within the methodological limits of $\dot{\mathrm{V}}$ O₂ resolution during brief maximal efforts.

While absolute aerobic contribution remained similar, relative percentages differed due to variations in total energy expenditure between modalities. Specifically, CJ30 exhibited lower total energy output, thereby increasing the relative share of aerobic energy. This pattern reflects the dependency of percentage-based values on total energetic magnitude rather than a true increase in oxidative involvement. Importantly, biopsy-based sprint studies have demonstrated that aerobic metabolism can account for a substantial proportion of total energy supply during 30-s efforts, particularly when anaerobic contribution declines,³⁴ supporting the physiological relevance of oxidative pathways even in short-duration maximal exercise. Nevertheless, because pulmonary $\dot{\mathrm{V}}$ O₂ responses lag behind intramuscular metabolic events, small early differences in oxidative engagement may not be fully detectable within a 30-s measurement window.

Although the PCr–La–O₂ model provides an integrated framework for partitioning aerobic, glycolytic, and phosphagen energy contributions, its application relies on several physiological assumptions that warrant consideration.^14,15,37 The estimation of glycolytic energy from Δ[BLa] assumes a fixed O₂–lactate equivalent and a standardized lactate distribution volume (∼45% of body mass), despite known inter-individual variability in lactate kinetics, buffering capacity, and clearance rates.^3,14 Similarly, the attribution of the fast component of EPOC primarily to phosphocreatine resynthesis is based on established physiological rationale, yet rapid oxygen stores replenishment, temperature-related processes, and other recovery dynamics may also contribute to early post-exercise oxygen uptake.^5,28 While these assumptions are consistent with prior methodological frameworks and have demonstrated acceptable reliability in repeated-measures designs,^5,14,28 they introduce potential sources of estimation error—particularly when applied to relatively small samples. Importantly, however, the present study employed a within-subject, cross-modal comparison using identical analytical procedures, thereby reducing systematic bias across conditions. Therefore, while absolute values should be interpreted within the methodological boundaries of indirect modeling, the relative differences observed between locomotion modalities remain physiologically informative and consistent with prior energetic investigations.^6,15,18

From an exploratory perspective, these findings may contribute to the interpretation of sport-specific energetic characteristics across time-matched maximal efforts, rather than directly informing test selection or athlete monitoring. If the goal is to examine absolute phosphagen and glycolytic energy turnover within a 30-s framework—relevant to many high-intensity team sports—WAnT and RUN30 elicited greater total anaerobic demand than CJ30 in the present cohort. However, the present study did not evaluate reliability, responsiveness to training adaptations, or predictive validity for sport performance; therefore, these findings should not be interpreted as establishing diagnostic superiority. Accordingly, RUN30 should currently be regarded as an exploratory running-based protocol rather than a practically established assessment tool. Its higher absolute phosphagen contribution compared to CJ30 supports further investigation of the protocol but does not, by itself, justify practical implementation. Future studies examining test–retest reliability, sensitivity to training-induced changes, and associations with match performance are required before practical implementation can be recommended. Additionally, combining electromyography and advanced metabolic modeling may help clarify the mechanistic basis of the observed differences.

While the present findings provide insight into locomotion-specific metabolic demands, certain considerations should be acknowledged. The sample consisted exclusively of trained male team-sport athletes (Tier 2), which may limit generalizability to female athletes, elite competitors, or untrained populations. Given well-documented sex-related differences in muscle fiber distribution, glycolytic capacity, and lactate kinetics, the present results should not be directly extrapolated to female athletes without dedicated validation.38, 39, 40 In addition, although the PCr–La–O₂ model offers an integrated estimation of energy pathway contributions, it remains an indirect approach based on physiological assumptions rather than direct intramuscular measurements. Moreover, typical error, test–retest reliability of the RUN30 protocol, and model-fitting uncertainty for pathway-specific estimates were not directly quantified in the present study. Given that RUN30 is a newly proposed protocol, the absence of reliability and measurement-error estimates represents a particularly important limitation, as it remains unclear to what extent repeated measurements would reflect true performance differences rather than within-subject variability or measurement error. Consequently, the present findings cannot establish the precision, reproducibility, or practical utility of RUN30 as an assessment tool. Model-fitting uncertainty may also affect the precision of estimated energy contributions, particularly the phosphagen component derived from the fast EPOC response. This consideration is particularly relevant for the phosphagen contribution, where the between-modality effect, although statistically significant, was more modest (ηp² = 0.240) than those observed for total energy expenditure, the glycolytic contribution, and the blood lactate response. Running performance was assessed under controlled but field-based conditions, which may introduce minor environmental variability despite standardized scheduling. Although ambient temperatures and humidity were comparable between indoor (∼22°C) and outdoor (∼20°C) sessions, even small environmental differences may influence $\dot{\mathrm{V}}$ O₂ kinetics and lactate responses, and therefore cannot be entirely excluded as a contributing factor.

Furthermore, the RUN30 protocol was performed as a self-paced, overground maximal effort initiated from a standing start, inherently including an acceleration phase. Velocity–time profiles or split distances were not recorded; therefore, pacing dynamics and acceleration characteristics could not be quantified. This limitation is particularly important because pacing and acceleration may directly influence the temporal distribution of phosphagen and glycolytic energy provision during a 30-s maximal effort. Accordingly, the present design does not allow the independent effects of locomotion modality to be fully disentangled from those of acceleration dynamics and self-selected pacing. As a result, differences in energy system partitioning between RUN30 and the laboratory-based protocols may partly reflect pacing freedom and progressive force development rather than locomotion modality alone. Future investigations incorporating velocity profiling, acceleration-phase analysis, test–retest reliability assessment, typical error estimation, model-fitting uncertainty analysis, and complementary physiological measurements such as muscle activation profiling or alternative metabolic modeling techniques may further clarify the interaction between mechanical characteristics and energy system distribution during short-duration maximal exercise.

5. Conclusions

This study shows that during 30-s maximal efforts, running and cycling produce greater absolute contributions to glycolytic energy and total energy expenditure than jumping in trained male team-sport athletes. RUN30 also demonstrated a higher absolute phosphagen contribution than CJ30, whereas the aerobic energy contribution did not differ between modalities. These findings indicate modality-associated differences in anaerobic energy partitioning under time-matched conditions; however, these differences should not be attributed solely to locomotion modality, as task-specific mechanical constraints and pacing opportunities may also have contributed to the observed responses. The findings also highlight the importance of interpreting absolute energy values alongside relative percentages. RUN30 should be considered a preliminary running-based protocol for assessing short-duration anaerobic performance; however, further evidence regarding its reliability, validity, and practical applicability is required before its implementation can be recommended.

Availability of data and material

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Author contributions

Conceptualization: S.U., S.Ö., C.G., N.F.K. and F.K.; Methodology: S.U., S.Ö., C.G., N.F.K. and F.K.; Data Curation: S.U., H.H.Y., K.S., F.Ö. and G.A.; Formal Analysis: S.U., S.Ö., C.G. and S.Ç.; Writing – Original Draft: S.U., S.Ö., N.F.K. and F.K.; Writing – Review & Editing: S.U.; Supervision: S.U., S.Ö., N.F.K., F.K., O.H. and S.K.

Ethics approval and consent to participate

The study received approval from the Ethics Committee of Atatürk University, Sport Sciences Faculty (Approval No: E−70400699-050.02.04-2200126915/2022/04 Date: 21.04.2022), ensuring compliance with the ethical standards stipulated by the Declaration of Helsinki.

Consent for publication

All participants provided written informed consent for the publication of anonymized data.

Availability of data and material

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This publication was supported by the Scientific Research Projects Coordination Unit of Istanbul Yeni Yuzyil University.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.