Impact of Sprint Duration With Fixed Recovery on Acute Physiological and Performance Responses to Repeated Cycle Sprints in Hypoxia in Male and Female Sprinters

ABSTRACT

Repeated short sprints (≤ 10 s) in hypoxia are often used to enhance performance, but acute responses to longer sprints (> 10 s) performed in a repeated‐sprint format with short recovery periods (∼30 s) remain unclear. We examined performance and physiological strain during repeated short (10‐s) and long (15‐s) sprints with identical recovery (30 s) comparing acute responses between sexes. Sixteen highly trained sprinters (7 females) performed 7 × 10‐s (10:30) and 15‐s (15:30) sprints with 30‐s rest in normobaric hypoxia (FiO₂: 0.15), using a randomized cross‐over design. Average mean power output across the seven sprints was lower in 15:30 than 10:30 (6.30 ± 0.44 vs. 6.85 ± 0.82 W/kg; p < 0.001). Sprint decrement score (S_dec) was larger in 15:30 than 10:30 (−17.9 ± 4.8% vs. −13.1 ± 4.6%; p < 0.001). Average pulse oximetry‐derived peripheral oxygen saturation (SpO₂) across the seven sprints was lower in 15:30 than 10:30 (87.1 ± 4.2% vs. 89.7 ± 2.7%; p < 0.001), whereas the magnitude of change from sprint 1 to sprint 7 did not clearly differ between conditions. Heart rate increased across sprints (Δsprint1‐7: +9.2 ± 5.3% and p < 0.001), independently of condition (p = 0.652). ΔSpO₂ correlated with ΔS_dec (r = 0.609 and p = 0.012). Female athletes experienced smaller performance declines than males in both conditions (p = 0.001). Longer repeated sprints with identical rest in hypoxia impair mechanical output while inducing greater physiological strain, with comparatively smaller performance reductions in females.

Highlights

• Extending sprint duration from 10 to 15 s with fixed 30‐s recovery in normobaric hypoxia (FiO₂ = 0.15) led to greater performance decrement and lower SpO₂, indicating reduced external mechanical stimulus but greater physiological strain.

• The between‐condition change in SpO₂ correlated positively with the between‐condition change in S_dec, suggesting that a greater peripheral deoxygenation might be associated with impaired repeated sprint ability when sprint duration is increased while recovery duration is held constant fixed under hypoxic conditions.

• Females showed greater fatigue resistance than males in both 10‐ and 15‐s hypoxic repeated‐sprint protocols, despite similar SpO₂ and heart rate responses, indicating that factors other than cardiorespiratory and arterial oxygenation responses may underlie the sex differences in fatigue development.

1. Introduction

Many team and racket sports require athletes to execute repeated maximal or near‐maximal efforts with incomplete recovery, known as repeated sprint ability (RSA) (Girard et al. 2011). Repeated sprint training in hypoxia (RSH) is one of the most effective interventions to boost RSA (Faiss et al. 2013; Brocherie et al. 2017; Faiss et al. 2025; Kasai et al. 2019). Faiss et al. (2013) first showed that RSH improves RSA after eight sessions (3 × 5 × 10‐s all‐out sprints with 20‐s recoveries) compared to equivalent normoxic training (Faiss et al. 2013). A meta‐analysis later confirmed that RSH provides additional benefits for peak (SMD = 0.31) and mean (SMD = 0.46) power outputs during repeated sprints relative to near sea‐level training (Brocherie et al. 2017).

Most previous RSH studies have employed short sprints lasting 5–10 seconds (Faiss et al. 2025). However, some have incorporated 15–20‐s sprints (Kasai et al. 2017, 2019), and Brocherie et al. (2017) recommended repeated efforts of 4–15 s with a 1:2 to 1:5 work‐to‐rest ratio (Brocherie et al. 2017). Nevertheless, compared to the numerous investigations examining the acute and long‐term effects of RSH using ≤ 10 s sprints, research on > 10 s sprints remains very limited. The dominance of ≤ 10 s protocols in RSH studies likely reflects their practical relevance to team‐sport athletes (Faiss et al. 2025), who repeatedly perform short‐duration sprints (< 10 s) during competition (Spencer et al. 2005). Conversely, RSH training has also been applied to cyclists (Faiss et al. 2013; Faiss and Rapillard 2020), who are sometimes required to perform repeated longer sprints (> 10 s) under fatigue, often during the decisive final moments of a race. For instance, cyclists perform over 70 sprints during a ∼90 min criterium, with > 10‐s efforts accounting for > 15% of all sprints; notably, most of these longer efforts fall within the 11–15‐s range (Ebert et al. 2006). Therefore, incorporating repeated long sprints (> 10 s) into RSH may be a valuable strategy for some athletes (e.g., cyclists). However, evidence comparing the acute physiological and performance responses to short (≤ 10 s) versus long (> 10 s) repeated sprints in hypoxia is lacking.

The use of longer sprints in RSH can substantially alter acute physiological and performance responses. For example, Raberin et al. (2023) compared repeated 5‐, 10‐, and 20‐s sprints using fixed exercise‐to‐rest ratio (1:2) under normoxia or hypoxia (FiO₂ = 0.13) (Raberin, Elmer, et al. 2023). They reported that sprint duration influenced the oxidative‐glycolytic balance, blood lactate accumulation, and muscle oxygenation responses, resulting in a reduced number of sprints completed to exhaustion with longer sprint durations (Raberin, Elmer, et al. 2023). Thus, extending sprint duration from 10 to 15 s is expected to increase the glycolytic contribution and metabolite accumulation. However, differences between normoxia and hypoxia occurred only during the shorter (5‐ or 10‐s) sprints, where hypoxia reduced performance (Raberin, Elmer, et al. 2023). This likely reflect the shorter absolute recovery duration rather than the sprint duration itself. Supporting this, another study using the same sprint durations (5, 10, or 20 s) but a longer recovery period (exercise‐to‐rest ratio = 1:6) reported no performance differences between normoxia and hypoxia (FiO₂ = 0.136) (Raberin, Willis, et al. 2023). Because PCr resynthesis occurs rapidly (Bishop et al. 2011; Haseler et al. 1999) and PCr is the primary energy source during repeated sprints (Girard et al. 2011), even small reduction in absolute recovery duration could alter PCr recovery kinetics and influence RSA (Haseler et al. 1999; Dawson et al. 1997; Mendez‐Villanueva et al. 2012). Furthermore, RSH has been shown to increase intramuscular creatine stores (Kasai et al. 2017, 2019), suggesting that short absolute recovery durations which limit complete PCr resynthesis, may help drive these adaptations. Collectively, these findings suggest that simply adjusting the exercise‐to‐rest ratio when using longer sprints may not replicate the intended physiological stimulus of RSH. This study investigated the acute physiological and performance responses to repeated‐sprint protocol, characterized by increased glycolytic dependence and incomplete phosphocreatine recovery, by deliberately extending sprint duration from 10 to 15 seconds while keeping the absolute rest time fixed.

Although most RSH studies have typically recruited male participants (Faiss et al. 2025), recent studies have begun to address this gap (Piperi et al. 2024; Devantay et al. 2025; Ait et al. 2024). Previous studies reported that RSH can improve repeated‐sprint ability to exhaustion to a similar extent in active males and females (Piperi et al. 2024). In recent years, studies have examined both the acute physiological responses and training adaptations to RSH induced by voluntary hypoventilation at low lung volume (RSH‐VHL) in women, although male participants were not included (Devantay et al. 2025; Ait et al. 2024). One of these studies reported beneficial training adaptations following RSH‐VHL, whereas the other showed that the intended level of arterial oxygen desaturation was not consistently achieved, thereby highlighting the need for further investigation. Together, these findings reflect a growing interest in female responses to RSH and RSH‐VHL. However, evidence regarding the physiological and performance responses to RSH and RSH‐VHL in women remains limited, and it remains unclear whether these responses differ between males and females. Females generally have a higher proportion of type I muscle fibers and rely less on glycolysis, with comparatively greater fat oxidation capacity (Ansdell et al. 2020). They also have smaller airways and lungs relative to stature, which may increase susceptibility to exercise‐induced arterial hypoxemia (Ansdell et al. 2020). In addition, a recent review highlighted that females and males may differ across several components of the oxygen transport and utilization cascade during hypoxic exercise, including ventilatory responses, vascular regulation, hematological characteristics, muscle metabolism, and autonomic modulation (Raberin et al. 2024). During high‐intensity repeated exercise, such as RSH, females' anatomically smaller lungs and airways may increase ventilatory constraints and the work of breathing during strenuous exercise in hypoxia. Conversely, females may also exhibit distinct peripheral characteristics, including a more oxidative muscle profile and lower reliance on glycolytic metabolism, which could influence fatigue development during repeated high‐intensity efforts. These biological and anatomical differences suggest that acute performance and physiological responses to RSH may vary by sex, particularly when sprint duration is extended and metabolic demand is increased. In addition, in a normoxic repeated‐sprint protocol (5 × 20‐s all‐out sprints with 25‐s rest), females exhibited smaller performance decrements and greater fatigue resistance than males (Billaut and Smith 2009). Conversely, Smith and Billaut (2012) reported that, during repeated‐sprint exercise under hypoxic conditions (FiO₂ = 0.13 and 10 × 10‐s all‐out sprints with 30‐s rest), there was no sex difference in performance when the mechanical work of the initial sprint was matched between sexes (equalizing the absolute performance level) (Smith and Billaut 2012). Evidence for sex differences in repeated‐sprint ability appears protocol‐dependent and inconsistent. Notably, no study has examined how increasing sprint duration in hypoxia (i.e., an adjustment that amplifies metabolic demand) modifies acute responses between sexes. Consequently, the influence of longer sprint intervals on sex‐based differences during RSH remains unknown.

Therefore, this study aimed to test whether repeated longer sprints (15‐s efforts with 30‐s recovery) would result in lower mechanical output and lower pulse oximetry‐derived SpO₂ compared with shorter sprints (10‐s efforts with 30‐s recovery) performed in normobaric hypoxia (FiO₂ = 0.15). A secondary aim was to quantify sex differences in acute physiological and performance responses between the 10‐s and 15‐s repeated‐sprint formats under identical hypoxic and recovery conditions. Accordingly, we hypothesized that, compared with the 10‐s sprint protocol, the 15‐s sprint protocol would result in lower mean power output, greater sprint decrement, and lower pulse oximetry‐derived SpO₂. As a secondary hypothesis, given the limited and partly inconsistent evidence regarding sex‐related responses to repeated‐sprint exercise in hypoxia, we expected that acute performance and physiological responses would differ between male and female athletes.

2. Methods

2.1. Participants

The sample size was estimated using a power analysis software (G*Power 3.1.9.7, Heinrich‐Heine‐Universität Düsseldorf, Germany) based on the effect size (d = 0.705) reported for within‐condition differences in performance (fatigue index) between the two different sprint duration (10‐s vs. 20‐s sprints) during RSH (exercise‐to‐rest ratio = 1:2 and FiO₂: 0.13) (Raberin, Elmer, et al. 2023). For sex‐based comparisons, the estimated sample size was derived from a previously reported effect size (d = 1.104) for between‐sex differences in performance decline (work decrement from the first to the 20th repetition) during repeated‐sprint exercise in normoxia (20 × 5‐s sprint with 25‐s rest and FiO₂: 0.21) (Billaut and Smith 2009). Power analyses indicated that 14 participants (7 per sex) would achieve adequate statistical power. Therefore, we recruited 16 participants (7 females and 9 males) to account for potential dropouts. Sixteen highly trained sprint runners (as described in Table 1), categorized as Tier 3 based on established criteria, volunteered to participate after providing written informed consent (McKay et al. 2022). All participants had not been exposed to a hypoxic environment for at least 3 months prior. The study adhered to the Declaration of Helsinki and received approval from the Research Ethics Committee of the Surugadai University (No. 03‐8‐1).

TABLE 1.

Characteristics of participants.

	Age (year)	Weight (kg)	Height (m)	Personal best records
				100‐m (s)	200‐m (s)	400‐m (s)
Female (n = 7)	20 ± 2	53.6 ± 3.1	1.59 ± 0.05	12.17 ± 0.29	25.18 ± 0.43	56.40 ± 0.50
Range	18–22	50.0–57.0	1.53–1.64	11.84–12.52	24.39–25.65	55.90–56.98
Male (n = 9)	21 ± 1	63.5 ± 6.9	1.72 ± 0.07	10.72 ± 0.15	21.74 ± 0.51	48.70 ± 1.48
Range	19–22	54.1–76.0	1.63–1.87	10.56–10.92	20.86–22.19	46.64–51.19
All (n = 16)	21 ± 1	59.2 ± 7.4	1.67 ± 0.09	11.38 ± 0.79	23.46 ± 1.85	51.07 ± 3.90
Range	18–22	50.0–76.0	1.53–1.87	10.56–12.52	20.86–25.65	46.64–56.98

Note: Values are expressed as mean ± SD.

2.2. Experimental Trial

This pilot study was performed as a randomized cross‐over trial. On separate visits, participants performed either short repeated sprints (7 × 10‐s ‘all‐out’ cycle sprints with 30‐s passive recoveries; 10:30) or long repeated sprints (7 × 15‐s ‘all‐out’ cycle sprints with 30‐s passive recoveries; 15:30) in normobaric hypoxia (FiO₂: 0.15). These two exercise tests were separated by 4–7 days. After a self‐selected warm‐up (e.g., walking, jogging, and dynamic stretching), participants performed 3 × 10‐s familiarization sprints (inter‐sprint recovery = 50 s) at 60%, 80%, and 100% of maximal effort in hypoxia (FiO₂: 0.15). The peak cadence achieved during the maximal reference sprint was recorded as an individual benchmark value. Before accepting the repeated‐sprint trial, we confirmed that the peak cadence during the first sprint was within 95% of this reference value for all participants (achieved in all instances). Strong verbal encouragement was also provided during every sprint to encourage maximal effort throughout the protocol. Cycling workload was set at 5.0% of body weight, and all participants regularly performed sprint cycling as part of their training routines. Although 7.5% of body mass is commonly used for cycle ergometer sprint exercise, a slightly lower load (5.0% body mass) was adopted in the present study to ensure sprint repeatability and avoid excessive reductions in pedal cadence during the repeated 15‐s sprint protocol. This load was selected because the protocol involved 15‐s sprints, which are longer than those typically used in repeated‐sprint exercise, and was considered more appropriate for sprint‐trained athletes to maintain maximal pedaling effort across repeated bouts. Handlebar and seat positions were replicated across visits. During the repeated sprint tests, mean power output (MPO), pulse oximetry‐derived peripheral oxygen saturation (SpO₂), and heart rate were measured. Participants were asked to avoid ergogenic substances (e.g., energy drinks and supplements) for 24 h and refrain from heavy physical activity for 48 h before each test. All sessions were conducted at the same time of day (± 1 h) for each participant to mitigate circadian effects.

2.3. Measures

The hypoxic chamber (FCC‐5000S, Fuji Medical Science, Japan) maintained normobaric hypoxia (15% O₂ and < 1% CO₂) through a nitrogen‐dilution technique. Testing utilized an electrically braked cycle ergometer (PowerMax VIII, Konami, Japan) to measure MPO for each sprint. The sprint decrement score (S_dec) was calculated by the formula below (Girard et al. 2011).

$${\mathit{S}}_{\mathit{d}\mathit{e}\mathit{c}}(\%)=\left\{\frac{\left({\mathit{S}}_{\mathbf{1}}+{\mathit{S}}_{\mathbf{2}}+{\mathit{S}}_{\mathbf{3}}+\text{…}+{\mathit{S}}_{\mathit{f}\mathit{i}\mathit{n}\mathit{a}\mathit{l}}\right)}{\left({\mathit{S}}_{\mathit{b}\mathit{e}\mathit{s}\mathit{t}}\times \mathit{n}\mathit{u}\mathit{m}\mathit{b}\mathit{e}\mathit{r}\mathit{o}\mathit{f}\mathit{s}\mathit{p}\mathit{r}\mathit{i}\mathit{n}\mathit{t}\mathit{s}\right)}-\mathbf{1}\right\}\times \mathbf{100}$$

where S refers to sprint performance (i.e., MPO). The S_dec assesses fatigue by evaluating an individual's actual performance in relation to their “ideal performance” (i.e., where the best effort would be replicated in each sprint). Heart rate (Polar H10, Polar, Finland) was measured throughout the experiment. SpO₂ were measured using a pulse oximeter (SAT‐2200, Nihonkohden, Japan). The sensor was attached to the index finger of the left hand immediately after sprint and SpO₂ was recorded approximately 10 s after sprint completion during passive recovery. This timing was selected to minimize motion‐related artifacts during exercise and to allow several seconds for the displayed SpO₂ value to stabilize after sensor placement. A timekeeper monitored the time from the end of each sprint and informed the examiner when 10 s had elapsed. The examiner then recorded the single SpO₂ value displayed on the monitor at that time. ΔSpO₂, ΔMPO, and ΔHeart rate were calculated as the average difference between the 15:30 and 10:30 conditions, and ΔS_dec was computed as the difference in S_dec between conditions. In addition, pooled data from both conditions were analyzed to examine the overall association between average SpO₂ and S_dec across protocols. Δsprint1‐7 was calculated as the percentage change from sprint 1 to sprint 7. This variable was used as a descriptive index of the magnitude of change across the repeated‐sprint protocol. To investigate sex differences in the progression of physiological responses, changes in SpO₂ and heart rate across sprints were calculated as the difference between the final and initial sprint. In addition, the S_dec was used to assess sex differences in fatigue development between conditions.

2.4. Statistical Analysis

Statistical analysis was carried out in GraphPad Prism (v10.2.2; GraphPad Software, USA). Normality was assessed using the Shapiro–Wilk test. Paired t‐tests and two‐way repeated measures ANOVA [time (sprint 1, 2, 3, … and 7) or sex (male and female) × condition (10:30 and 15:30)] were used to compare dependent variables, followed by Tukey's multiple comparisons. To assess the assumptions of variance, Mauchly's sphericity test was employed for all ANOVA results, with a Greenhouse–Geisser correction applied if assumptions were violated. Pearson's correlation coefficients were computed to investigate relationships between variables. Effect sizes were determined using Cohen's d (0.20–0.49 = small effect; 0.50–0.79 = moderate effect; and ≥ 0.80 = large effect) or eta squared (η ²; 0.010–0.059 = small effect; 0.060–0.139 = moderate effect; and ≥ 0.140 = large effect). All values are presented as mean ± standard deviation, with statistical significance set at p < 0.05.

3. Results

MPO decreased significantly across sprints (p < 0.001 and η ² = 0.393), with a greater decline (p = 0.023 and η ² = 0.081) observed in 15:30 compared with 10:30 (Δsprint1‐7: −25.6 ± 6.6% vs. −20.4 ± 7.6%; Figure 1A). Average MPO across the seven sprints was significantly lower (p < 0.001 and d = 1.19) in 15:30 than 10:30 (6.30 ± 0.44 vs. 6.85 ± 0.82 W/kg; Figure 1B). S_dec was significantly larger (p < 0.001 and d = 1.43) in 15:30 than 10:30 (−17.9 ± 4.8 vs. −13.1 ± 4.6%; Figure 2).

FIGURE 1.

Changes in mean power outputs across sprint repetitions (A) and the averaged power output of seven sprints (B) between conditions. Blue circle markers and line indicate the 10‐s sprint condition (10:30), whereas orange triangles and line indicate 15‐s sprint condition (15:30). Markers represent individual values, and lines represent mean values. Gray connecting line indicate individual changes between conditions. ^### p < 0.001, ^## p < 0.01, significantly different from the previous sprint (i.e., sprint 1 vs. 2 and sprint 2 vs. 3). ***p < 0.001, **p < 0.01, significantly different between conditions (10:30 vs. 15:30). p value and effect size are expressed as p value (effect size).

FIGURE 2.

Sprint decrement score between conditions. Blue circle markers and line indicate the 10‐s sprint condition (10:30), whereas orange triangles and line indicate the 15‐s sprint condition (15:30). Markers represent individual values, and lines represent mean values. Gray connecting lines indicate individual changes between conditions. ***(p < 0.001) denotes a significant difference between conditions.

SpO₂ decreased significantly across sprint repetitions (p < 0.001 andη ² = 0.336), with a greater decline (p = 0.022 and η ² = 0.074) noted in 15:30 compared with 10:30 (Δsprint1‐7: −9.4 ± 4.9% vs. −9.1 ± 5.3%) (Figure 3A). Average SpO₂ across the seven sprints was significantly lower (p < 0.001 and d = 1.21) in 15:30 than 10:30 (87.1 ± 4.2 vs. 89.7 ± 2.7%; Figure 3B). Heart rate increased significantly across repetitions (p < 0.001 and η ² = 0.178; Δsprint1‐7: +9.2 ± 5.3%), independently of condition (p = 0.652 and η ² = 0.004) (Figure 4A). Average heart rate did not differ (p = 0.548 and d = 0.228) between 10:30 (170.2 ± 8.2 bpm) and 15:30 (171.7 ± 9.9 bpm; Figure 4B).

FIGURE 3.

Changes in peripheral oxygen saturation (SpO₂) across repetitions (A) and averaged arterial oxygen saturation of seven sprints (B) between conditions. Blue circle markers and line indicate the 10‐s sprint condition (10:30), whereas orange triangles and line indicate 15‐s sprint condition (15:30). Markers represent individual values, and lines represent mean values. Gray connecting line indicate individual changes between conditions. ^# p < 0.05, significantly different from the previous sprint (i.e., sprint 2 vs. 3 and sprint 4 vs. 5). **p < 0.01, *p < 0.05, significantly different between conditions (10:30 vs. 15:30). p value and effect size are expressed as p value (effect size).

FIGURE 4.

Changes in heart rate across repetitions (A) and averaged heart rate of seven sprints (B) between conditions. Blue circle markers and line indicate the 10‐s sprint condition (10:30), whereas orange triangles and line indicate 15‐s sprint condition (15:30). Markers represent individual values, and lines represent mean values. Gray connecting line indicate individual changes between conditions. ^# p < 0.05, significantly different from the previous sprint (i.e., sprint 1 vs. 2 and sprint 3 vs. 4). The n.s. indicates ‘not significantly different’ between conditions. p value and effect size are expressed as p value (effect size).

There was a significant correlation between ΔSpO₂ and ΔS_dec (r = 0.609 and p = 0.012; Figure 5A), but not between ΔSpO₂ and ΔMPO (r = 0.132 andp = 0.472; Figure 5B), or ΔSpO₂ and ΔHeart rate (r = −0.095 and p = 0.726; Figure 5C). Additional exploratory analyses were performed to examine whether average SpO₂ was associated with S_dec in pooled data. The relationship between average SpO₂ and S_dec was not significant (r = 0.256 and p = 0.157; Fig. S1). These findings suggest that the between‐condition association between ΔSpO₂ and ΔS_dec may be influenced by factors other than oxygen saturation alone and should therefore be interpreted cautiously.

FIGURE 5.

Relationships between Δarterial oxygen saturation and Δsprint decrement score (A), Δmean power output (B), and Δheart rate (C). Pulse oximetry‐derived oxygen saturation, SpO₂; sprint decrement score, S_dec; and mean power output, MPO. Blue cross (male) and orange square (female) markers indicate individual measured values, and solid lines indicate approximate straight lines. All Δ variables represent between condition differences calculated for each participant as the average value across seven sprints in the 15:30 condition minus the average value across the seven sprints in the 10:30 condition. The gray area represents the 95% confidence interval.

Regarding sex differences, S_dec was significantly higher (p < 0.001 and η ² = 0.228) in females than males in both the 10:30 (−9.5 ± 2.1 vs. −15.8 ± 3.8%) and 15:30 conditions (−14.8 ± 2.6 vs. −20.0 ± 4.4%; Figure 6C). In addition, S_dec was lower in the 15:30 condition than in the 10:30 condition in both males (−15.8 ± 3.8 vs. −20.0 ± 4.4%) and females (−9.5 ± 2.1 vs. −14.8 ± 2.6%; Figure 6C). In contrast, when sex‐related responses were examined using percent changes from sprint 1 to sprint 7, SpO₂ and heart rate did not differ between sexes or conditions (Figure 6A, B).

FIGURE 6.

Sex differences in changes in SpO₂ (A) and heart rate (B) across repeated sprints and in sprint decrement score (C). Pulse oximetry‐derived oxygen saturation, SpO₂; sprint decrement score, S_dec. Panels (A) and (B) show percent changes from sprint 1 to sprint 7. Blue bar and circle markers indicate 10:30 condition, and orange bar and triangles indicate 15:30 condition. **p < 0.01, ***p < 0.001 indicates significant differences between condition or sex. p value and effect size are expressed as p value (effect size).

4. Discussion

This study examined acute physiological and performance responses to repeated cycle sprints of 10 versus 15 s with 30‐s rest in moderate normobaric hypoxia. The main findings indicated that longer sprints with identical rest duration significantly reduce external workload (lower MPO and greater S_dec) and result in lower SpO₂ values, whereas heart rate readings did not differ between conditions. Additionally, the significant correlation between ΔSpO₂ and ΔS_dec suggests that a greater between‐condition reduction in SpO₂ was associated with a greater between‐condition impairment in repeated sprint ability. Regarding sex differences, females demonstrated greater fatigue resistance (i.e., lower S_dec) than males in both conditions, despite similar SpO₂ and heart rate responses between sexes.

As hypothesized, MPO decreased significantly across repetitions, with greater alterations for 15:30 than 10:30 (Δsprint1‐7: −25.6 ± 6.6% vs. −20.3 ± 7.6%). These decrements are slightly lower than the −31.2%–36.2% noted previously during 10 × 10‐s sprints with 30‐s recovery, likely because those protocols included three additional sprints (Billaut and Smith 2009; Smith and Billaut 2012). In our study, the MPO decline was significantly larger in 15:30 for sprints 3–7 with a moderate effect size (p = 0.023 and η ² = 0.081; Figure 1A). S_dec was also higher in 15:30 than 10:30. Overall, this suggests that longer sprints with identical rest impair exercise capacity during RSH, leading to a decreased mechanical training stimulus.

Another key observation was the significant correlation (r = 0.609 and p = 0.012; Figure 5A) between the performance reduction (ΔS_dec) and the decrease in peripheral oxygen saturation (ΔSpO₂) with longer sprint durations. This suggests that, during repeated cycle sprints of 10 s versus 15 s with the identical 30‐s recovery, greater reduction in pulse oximetry‐derived SpO₂ with comparable heart rate may be related to larger performance decline. However, this relationship should be interpreted cautiously. The additional pooled analysis showed no significant association between average SpO₂ and S_dec (Fig. S1). Therefore, although the Δ‐based association suggests that changes in postsprint SpO₂ and sprint decrement occurred in parallel when sprint duration was extended, the present findings do not establish that lower SpO₂ directly explains the greater performance decrement. Previous research comparing 5‐, 10‐, and 20‐s sprints under normoxia or hypoxia at a fixed 1:2 exercise‐to‐rest ratio reported performance decrements only in the shorter sprint conditions (5 or 10 s) (Raberin, Elmer, et al. 2023). Despite longer sprints being more demanding, the detrimental effect of hypoxia disappeared in the longer (20 s) sprint condition when the absolute recovery duration increased proportionally (Raberin, Elmer, et al. 2023). This has been attributed to the rapid second‐by‐second kinetics of PCr resynthesis (Bishop et al. 2011; Haseler et al. 1999), such that longer absolute recovery periods (inherent to longer sprints performed with a matched work‐to‐rest ratio) may allow PCr levels, even under hypoxia, to recover to a greater extent, thereby reducing performance differences (Haseler et al. 1999; Dawson et al. 1997; Mendez‐Villanueva et al. 2012).

In this study, we intentionally employed same absolute rest duration (30 s) for both sprint durations. Under these circumstances, longer sprints (15 s vs. 10 s) produced greater performance decline, lower postsprint SpO₂ values, and identical heart rate response. Although PCr resynthesis was not assessed, the combination of reduced oxygen delivery (lower SpO₂ with similar heart rate) and longer sprint duration with same absolute rest duration likely impaired PCr recovery in the 15:30 condition. Previous studies have shown that RSH can increase PCr stores compared with normoxic training (Kasai et al. 2017, 2019). Although the mechanisms remain unclear, it is plausible that repeated sprints performed under hypoxia, where PCr resynthesis is restricted and PCr levels remain lower, serve as a key stimulus triggering these adaptations. From this perspective, when implementing longer sprints, it may be crucial to prescribe short absolute rather than relative recovery durations to intentionally limit phosphocreatine PCr resynthesis and maintain the desired training stimulus. However, the allocation of a fixed 30‐s recovery period may have contributed to the larger performance decrement observed in the 15:30 condition. This interpretation remains speculative because PCr kinetics were not measured. Additionally, because metabolite accumulation was not assessed, the contribution of other underlying mechanisms cannot be excluded. Further research is warranted to examine PCr kinetics and metabolite responses during longer sprint protocols employing either relative or absolute recovery structures.

As hypothesized, SpO₂ after sprints was lower in 15:30 than 10:30 (average: 87.1 ± 3.4 vs. 89.7 ± 2.7%), indicating greater physiological stimulus in 15:30. These readings align with previous research showing SpO₂ of ∼84%‐92% during repeated sprint exercise with exercise‐to‐rest ratios of 1:3 to 1:5 under hypoxia (FiO₂: 0.13–0.17) (Billaut and Buchheit 2013; Smith and Billaut 2010; Townsend et al. 2021). In our study, the lowest SpO₂ remained above 80% in all participants in the 10:30 condition, whereas in 15:30, 5 of 16 participants (∼30%) fell below 80%, despite most RSH protocols targeting 80%–90%. At the between‐condition level, the greater SpO2 decrease observed in the 15:30 condition coincided with a larger performance decline, suggesting a potential association between these responses rather than a causal relationship. Previous studies reported that decreased SpO₂ can led to reductions in cerebral tissue saturation index, accompanied by suboptimal muscle activation, contributing to impaired RSA (Smith and Billaut 2010; Billaut, Kerris, et al. 2013). Moreover, longer sprint durations (5, 10, or 20 s) increase muscle deoxygenation irrespective of recovery duration (exercise‐to‐rest ratio 1:2–1:6) (Raberin, Elmer, et al. 2023; Raberin, Willis, et al. 2023). Although a marked reduction in SpO₂ may contribute to inducing hypoxia‐related physiological adaptations, excessive desaturation levels during RSH has also been linked to reduced training load and attenuated adaptations (Takei et al. 2024, 2025). These observations highlight the importance of real‐time and continuous SpO₂ monitoring to prevent excessive desaturation and maintain desired SpO₂ target range when opting for longer sprints (> 10 s) during RSH sessions, preserving both safety and training quality (Soo et al. 2020).

Heart rate increased significantly across repetitions in both conditions, averaging ∼170 bpm. This is comparable to previous research involving 8 × 5‐s sprints (exercise‐to‐rest ratio 1:5) in normoxia or hypoxia (FiO₂: 0.17 or 0.13) (Townsend et al. 2021). Consistence with our hypothesis, heart rate responses did not differ between the 10:30 and 15:30 conditions. This suggests that heart rate was not sufficiently sensitive to distinguish the physiological demands of the two protocols, given their similar maximal repeated‐sprint nature and the relatively small difference in total exercise duration (35 s). Therefore, the absence of differences in heart rate should not be interpreted as evidence of similar physiological stress between conditions. Townsend et al. (2021) similarly reported that heart rate during repeated‐sprint exercise (8 × 5‐s all‐out sprints with 25‐s recovery) did not differ across oxygen levels (FiO₂: 0.21 vs. 0.17 vs. 0.13), despite large differences in SpO₂ (∼97% vs. ∼90%‐95% vs. ∼80%–89%) (Townsend et al. 2021). Thus, SpO₂ changes are not necessarily coupled with heart rate responses, consistent with our finding of no significant correlation between ΔSpO2 and ΔHR between conditions. A previous study also observed greater metabolic demand (i.e., higher blood lactate concentrations) when sprint duration were increased (5, 10, or 20 s) at the same exercise‐to‐rest ratio (1:2) (Raberin, Elmer, et al. 2023). Taken together, these findings indicate that longer sprints may affect anaerobic energy metabolism (i.e., increased glycolysis) without additional cardiovascular solicitation (aerobic systems).

Regarding sex differences, females exhibited smaller performance decrements than males in both the 10:30 and 15:30 conditions (Figure 6C). These findings align with a previous study (20 × 5‐s all‐out sprint with 25‐s rest protocol) reporting greater fatigue resistance in females (Billaut and Smith 2009). It is well established that, compared with males, females possess a higher proportion of type I muscle fibers, lower glycolysis reliance, and greater oxidative capacity (Haizlip et al. 2015; Lundsgaard and Kiens 2014; Cardinale et al. 2018). These sex‐based differences may have influenced peripheral fatigue progression during repeated sprints (Ansdell et al. 2020; Raberin et al. 2024). Although 15‐s sprint bouts are expected to increase glycolytic demand (i.e., an adjustment theoretically less advantageous for females) (Haizlip et al. 2015; Lundsgaard and Kiens 2014; Cardinale et al. 2018), no condition × sex interaction was observed for S_dec (Figure 6C). This suggests that superior fatigue resistance in females persists even when sprint duration is extended. A likely explanation is that although longer sprint bouts increase early glycolytic demand, PCr resynthesis—an oxidative‐dependent process—becomes a dominant determinant of performance maintenance as sprints are repeated (Girard et al. 2011). In contrast, prior work reported that sex differences in hypoxic repeated‐sprint performance disappear when initial‐sprint workloads are matched and normalized to lean mass (Smith and Billaut 2012), controlling by preliminary screening. Although the present study expressed power output relative to body weight (W/kg), this approach does not fully account for fat‐free mass differences and may allow residual confounding from adiposity. Furthermore, baseline performance differences between sexes may have also influenced the observed results.

Females generally possess smaller airways and lungs than males, which may increase vulnerability to exercise‐induced arterial hypoxemia (Ansdell et al. 2020; Raberin et al. 2024). It has also shown that the oxygen cost of exercise hyperpnea is greater in women than in men (Dominelli et al. 2015). However, we observed no sex differences in SpO₂ or heart rate responses (Figure 6A, B). This finding is partly consistent with, and extends, previous observations by Smith and Billaut (Smith and Billaut 2012), who reported that when initial sprint work was matched between males and females, sex differences in fatigue and systemic, cerebral, and peripheral oxygenation responses during repeated‐sprint exercise were largely minimized under both normoxic and hypoxic conditions. Importantly, the absence of sex differences in postsprint SpO₂ and heart rate should not be interpreted as evidence of comparable oxygen delivery or oxygen utilization between sexes. Previous studies have shown that, compared with young men, young women exhibit a greater compensatory vasodilatory response during exercise in hypoxia (Casey et al. 2014). However, because ventilation, respiratory mechanics, blood lactate, acid–base balance, muscle oxygenation, and blood flow were not assessed, the mechanisms underlying the observed sex‐related differences in S_dec remain unclear. Future studies should include these measurements to clarify how pulmonary, metabolic, and peripheral factors contribute to sex‐related responses during longer repeated‐sprint protocols in hypoxia.

This investigation examined the acute responses to repeated‐sprint exercise lasting longer than 10 seconds in hypoxia—a topic that has remains largely unexplored—and therefore has several limitations. First, we did not include a normoxic condition, preventing acute physiological responses between oxygen conditions. Nevertheless, because our primary aim was to clarify the acute responses to longer sprint durations during RSH, the present findings offer new insights for future RSH protocol design. Second, we fixed the absolute recovery duration to standardize the time available for PCr resynthesis. Although including relative recovery conditions (e.g., 10:20 or 15:45) may have provided a deeper understanding of recovery manipulation, we deliberately minimized experimental load given that participants were competitive athletes. Another limitation was that SpO₂ was measured intermittently after each sprint rather than continuously throughout exercise and recovery, because substantial body movement during sprint exercise introduced considerable motion artifacts. Continuous SpO₂ monitoring during all‐out sprint cycling would require a wearable or telemetry‐based device capable of providing stable measurements while minimizing susceptibility to motion artifacts. Such a device was not available for the present protocol. Therefore, the present SpO₂ data should be interpreted as standardized post‐sprint measurements, and the nadir SpO₂ response may have been missed. In addition, SpO₂ does not directly reflect active muscle oxygenation or oxygen delivery to the working muscles. Because muscle oxygenation, muscle blood flow, blood lactate, and acid‐base markers were not measured, the physiological mechanisms underlying the greater performance decrement in the 15:30 condition remain unclear. Future studies should incorporate multiple recovery structures and include continuous measurements of SpO₂, tissue oxygenation, blood flow, and metabolic responses to better characterize how recovery duration influences responses to longer sprints in hypoxia. Furthermore, to investigate sex differences in performance across RSH protocols, future studies should normalize workload to lean body mass and apply initial‐sprint performance matching between sexes.

5. Conclusion

In highly trained sprint runners, extending exercise duration from 10 to 15 s (with fixed 30‐s recovery) during hypoxic repeated cycle sprints significantly reduced exercise performance (lower MPO and greater S_dec) and induced greater peripheral deoxygenation (lower SpO₂), whereas heart rate responses did not differ. Although the between‐condition difference in SpO₂ was associated with the between‐condition difference in S_dec, this relationship should be interpreted cautiously and does not provide evidence of a causal link between lower SpO₂ and impaired repeated‐sprint performance. Regarding sex comparisons, females exhibited greater fatigue resistance than males across both conditions, despite similar SpO₂ and heart rate responses. These results suggest that longer sprints with identical rest intervals reduced external workload (lower mechanical stimulus) yet impose greater physiological strain (greater physiological stimulus) when incorporating RSH.

Funding

This study was financially supported by Japan Society for the Promotion of Science—JSPS KAKENHI Grant Nos. JP22K17693 and JP25K20999.

Ethics Statement

This study adhered to the Declaration of Helsinki and received approval from the Research Ethics Committee of the Surugadai University (No.03‐8–1). All participants volunteered to participate after providing written informed consent.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Relationships between average SpO₂ and Sdec in pooled data. Pulse oximetry‐derived oxygen saturation, SpO2; sprint decrement score, Sdec. Data from both the 10:30 and 15:30 conditions were pooled for this analysis. Blue cross (male) and orange square (female) markers indicate individual measured values.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.