The Effects of In-Season Complex Contrast Training on Anaerobic Performance Metrics in Elite Female Soccer Players

Abstract

Gepfert, M, Terbalyan, A, Roczniok, R, and Gołaś, A. The effects of in-season complex contrast training on anaerobic performance metrics in elite female soccer players. J Strength Cond Res 40(5): e491–e499, 2026—The aim of this study was to evaluate the effects of in-season complex contrast training (CCT) on speed, power, and anaerobic capacity metrics in elite female soccer players. Twenty-one players participated in a 3-phase study conducted across the postpreparatory period, midseason, and postseason. Testing included countermovement jump, single-leg jump, 5 and 30 m sprints, and the repeated anaerobic sprint test (RAST). Weekly complex training sessions combined resistance exercises with plyometric drills. Significant improvements were observed in anaerobic performance midseason compared with the preparatory phase (RAST 1, p = 0.0073, d [Cohen's d ES] = 0.45; 30 m sprint, p = 0.0073, d = 0.81), followed by performance declines at the end of the season. Peak RAST metrics improved significantly midseason (e.g., RAST 2, p = 0.025, d = 0.81; RAST 3, p = 0.0083, d = 0.95), but no changes were observed for concentric peak velocity (p = 0.98) or rate of force development (p = 0.37). The findings demonstrate that complex contrast training effectively enhances power and anaerobic performance during the midseason but may require strategic load management to mitigate late-season fatigue-related declines. A single weekly session of CCT seems sufficient to maintain, and in some instances improve, physical performance across the season. However, without a comparison group, it is unclear whether it is more or less effective than other training methods.

Introduction

Soccer is a high-intensity, intermittent team sport characterized by repeated explosive actions, including sprints, accelerations, decelerations, jumps, directional changes, and physical duels (18). These movements rely heavily on the stretch-shortening cycle and the capacity to rapidly generate force and power (31). Key determinants of soccer performance include strength, power, agility, and the ability to sustain repeated high-intensity efforts, all of which depend on neuromuscular efficiency and adequate muscle strength (12,14). The in-season period poses additional challenges, because frequent competitions and congested schedules limit the time available for supplemental physical training. Therefore, coaches must select training methods that are both time efficient and effective in maintaining or enhancing player performance while minimizing the risk of overload or injury. Importantly, elite women's soccer remains comparatively underrepresented in the literature, underscoring the need for research addressing the specific demands and recovery profiles of female athletes (2,37).

Within this context, contrast training (CT)—and, specifically, its complex form known as complex-CT (CCT)—has emerged as a promising strategy to target both strength and power within a single session (4). Complex-CT is recognized as an effective method for enhancing athletic performance by combining high-load resistance exercises with biomechanically similar explosive movements (8,30). This sequencing is intended to stimulate adaptations across the force–velocity spectrum and to leverage neuromuscular mechanisms that acutely enhance performance (11). Evidence suggests that CCT can surpass traditional resistance or plyometric training performed in isolation, producing improvements in maximal strength, sprinting, vertical jump performance, and change-of-direction speed (8,20).

The central physiologic mechanism underpinning CT is postactivation performance enhancement (PAPE). Postactivation performance enhancement refers to the temporary improvement in muscle performance—particularly force and power—after heavy-load conditioning activity (CA) when paired with an explosive exercise sharing a similar biomechanical pattern (24,30). Mechanistically, PAPE has been attributed to increased phosphorylation of myosin regulatory light chains, heightened calcium sensitivity, enhanced recruitment of high-threshold motor units, and improved synaptic conduction (3,36). These processes collectively improve rate of force development (RFD) and power output, which are crucial determinants of performance in explosive actions such as sprinting, jumping, and rapid changes in direction (16,34). Importantly, the effectiveness of PAPE is highly time dependent; optimal rest intervals between the CA and subsequent explosive activity are essential to balance potentiation and fatigue, with studies reporting effective windows ranging from <1 to >20 minutes (6,19). Although acute improvements are well documented, the extent to which these transient effects translate into chronic adaptations has required careful evaluation (8).

Recent long-term training studies using volume- and/or intensity-matched designs have begun to clarify this issue in team-sport athletes. For example, matched comparisons indicate that complex/contrast configurations can produce comparable improvements in strength- and power-related outcomes to traditional progressive resistance approaches when total work is controlled (25,28). Notably, in the latter, CCT sessions were completed on average ∼7 minutes faster than traditional training with similar session RPE, underscoring its time efficiency without greater perceived effort (25). This aligns with earlier season-long or mesocycle studies reporting beneficial or at least maintained performance with low-frequency CCT when exercise sequencing and recovery are well managed (4,8,20). Collectively, these data support the pragmatic appeal of CCT for in-season contexts—where time efficiency and careful load management are paramount—and may further encourage adoption by coaches in team-sport environments.

Therefore, the aim of this study was to investigate the effects of a 6-week complex-contrast training (CCT) program performed once per week during the in-season period in female soccer players. We specifically examined whether this low-frequency, time-efficient approach could maintain or enhance key performance indicators—speed, agility, jump performance—without increasing the overall training volume of the program. We hypothesized that the implementation of CCT would improve or preserve physical performance capacity despite the constraints of in-season training loads.

Methods

Experimental Approach to the Problem

The study was conducted over summer calendar during the 2024 competitive season and focused on the preparatory period and subsequent in-season phases. The research was designed to evaluate the effects of complex training on speed, power, and anaerobic performance metrics at distinct points in the season. Testing was performed at 3 specific time intervals:

• Postpreparation period (August): Testing occurred immediately after the conclusion of the preparation phase, just before the first league match.

• Midseason (early October): Testing took place midway through the summer calendar, after 6 league matches, coinciding with a break for the national team training camp.

• Postseason (late November): The final testing session was conducted at the end of the summer calendar, after the completion of all league matches.

Each experimental session was conducted on the same day of the training microcycle, precisely 3 days after a match (match day [MD] + 3). This timing was chosen to ensure consistency and minimize the potential influence of match-related fatigue. All testing sessions were performed under identical conditions. Strength assessments were performed in the university gym, while sprint, jump, and anaerobic capacity tests were conducted in an indoor hall equipped with a synthetic track. This ensured that environmental factors such as weather, wind, or temperature did not affect the results. All tests were performed at a standardized time of day, between 8:00 and 10:00 am, to control for diurnal variations in performance. A standardized warm-up protocol, identical equipment, and the same testing staff were used at all sessions to ensure methodological consistency.

Moreover, CCT protocol was implemented once per week and represented the only structured gym-based strength session during the in-season period. All remaining training sessions were field based and followed the standard competitive microcycle structure of the team. Specifically, the weekly schedule included:

• MD: official game.

• MD + 1: day off (full recovery).

• MD + 2: on-field recovery session.

• MD + 3: strength training session (CCT) combined with on-field small-sided games (reduced pitch size, fewer players).

• MD − 3: medium- to large-sided games with higher on-field physical demands.

• MD − 2: tactical session (set pieces, reduced overall load).

• MD − 1: final match preparation (game fragments with integrated plyometric elements).

Subjects

At the beginning, in January, 23 female soccer players from a professional team (Polish highest division) participated in the study. Two subjects were excluded from the analysis because they were injured during the study and did not participate in training and competitions. Finally, 21 players (age: 22.1 ± 2.4 years; height: 164.3 ± 3.8 cm; body mass: 58.7 ± 4.2 kg; body fat 16.3 ± 3.8%; back squat 1 repetition maximum [1RM]): 83.3 ± 9.2 kg) completed all stages of the research. Before signing the consent to take part in the research, all subjects were informed about the procedures, benefits, discomforts, and possible risks of the study. The study protocol was approved by The Research Ethics Committee for Scientific Research at the Academy of Physical Education in Katowice (Approval Number: 3/2021). All procedures were conducted in accordance with the Declaration of Helsinki.

Procedures

Testing sessions were conducted at 3 predetermined time points: after the preparation period, midseason, and at the end of the season. All testing took place on the same microcycle day—3 days postmatch (MD + 3)—and at a consistent time of day between 8:00 and 10:00 am The testing protocol included the following sequence of assessments: countermovement jump (CMJ), single-leg jump (SL), speed tests for 5 and 30 m, and the repeated anaerobic sprint test (RAST). All assessments were performed in a controlled indoor environment; ambient temperature was 20–23° C and relative humidity 45–55% to minimize variability due to environmental conditions.

Countermovement Jump and Single-Leg Jump

A bilateral CMJ with arm swing was performed on force plates (Forcedecks, Vald Performance, Newstead, Queensland, Australia). All athletes performed 3 CMJs with a 30-second rest interval between each attempt. The CMJ starting position was a standing position with a straight torso and the knees fully extended, with the feet shoulder-width apart and the hands free to move. The subjects were instructed to perform a quick downward movement (approximately 90° of knee flexion), and afterward a fast upward movement to jump as high as possible. Moreover, athletes were instructed to jump with both feet and knees fully extended (no leg tucking was allowed). The following variables were evaluated: jump height (based on the flight time measurement), relative peak power and concentric peak force. The best attempt in terms of jump height was preserved for further analysis.

The SL with arm swing was performed on force plates (Forcedecks, Vald Performance). All athletes performed 3 SL with a 30-second rest interval between each attempt. The SL starting position was a 1 leg standing position with a straight torso and the 1 knee fully extended, and the other leg was bent at 90°. The subjects were instructed to perform a quick downward movement (approximately 90° of knee flexion), and afterward a fast upward movement to jump as high as possible. Moreover, athletes were instructed to jump with feet and knee fully extended (no leg tucking was allowed). The following variables were evaluated: jump height (based on the flight time measurement), relative peak power. The best attempt in terms of jump height was preserved for further analysis.

Sprint Test

The tests were conducted with Microgate Witty photocell timing gates (Bolzano, Italy) positioned at 5 and 30 m. Short-distance acceleration (0–5 m) was timed using gates at the start and 5 m, and maximal sprint speed (0–30 m) was timed using gates at the start and 30 m. To standardize conditions and eliminate weather effects, all trials were performed in indoor facility on a turf-surfaced running track, with ambient temperature maintained at 20–22° C and relative humidity at 45–55%. Each subject completed 2 trials per test; the fastest time was retained for analysis.

Repeated Anaerobic Sprint Test

The test was conducted on the field using Microgate photocells (Photocells Witty Gate, Bolzano, Italy). Repeated anaerobic sprint test is a physical performance test designed to measure anaerobic power, fatigue, and endurance. The test consists of 6 maximal sprints for a 30-m distance with 25 seconds active rest to the starting point, at a jog. A stopwatch was used for the duration of the rest break, while the coach counted down from 5 so that the players could prepare. The short rest interval challenges the body's ability to recover and perform under anaerobic conditions, simulating the demands of repeated high-intensity sports activities. All assessments were performed in a controlled indoor environment; ambient temperature was 20–23° C and relative humidity 45–55% to minimize variability due to environmental conditions. All RAST time points (6) were used for further analysis.

Training Protocol

The complex method was used every week on the same day in a microcycle 3 days after match day (MD + 3). At the beginning, the players performed standard motor preparation including exercises for mobilization of the hip joint, ankle, hamstring muscle group, activation, and exercises for injury prevention. Then, general development exercises were performed, dynamic stretching, several exercises for the upper and lower body parts, such as squats with one's own body weight and split squats, double-leg jumps, single-leg jumps, wall drills, 4 × 5 m running accelerations, 2× accelerations with a change of direction and decelerations (Figure 1). The 1-minute CA plyometric transition and 2-minute interset rest were selected for in-season feasibility (MD + 3), to keep the main complex block <12 minutes and maintain neuromuscular temperature, acknowledging that longer rests can elicit greater acute PAPE in some contexts (3). Rest periods were not individualized.

Figure 1.

General training protocol.

The main training block used the complex method, in which each strength exercise was immediately followed by a biomechanically similar plyometric drill. Athletes completed 2 sets of 3 strength–plyometric pairs: (a) back squat, 3 repetitions at 80% 1RM, followed immediately by 5 box jumps to 75 cm; (b) barbell hip thrust, 3 repetitions at 80% 1RM, followed by 6 alternating split jumps (scissor jumps); and (c) Bulgarian split squat with dumbbells, 3 repetitions at ∼50% of back-squat 1RM, followed by 3 Bulgarian split jumps performed with body weight. Rest between the strength and plyometric actions was 1 minute, with 2 minutes between sets. The same protocol was applied throughout the 14-week in-season period (Figure 2). Each player's external load during the CCT protocol was individualized based on predetermined percentages of their 1RM, ensuring that the stimulus was matched to individual strength capacity rather than being applied uniformly across the group. Importantly, these loads remained constant throughout the study. This was a deliberate decision, because the primary objective of the study was to examine the feasibility and seasonal effects of embedding a standardized, low-frequency CCT protocol under realistic in-season conditions rather than to test progressive overload. In elite soccer, load progression during the competitive period is often limited by fixture congestion, recovery demands, and the need to avoid excessive fatigue or injury. Therefore, maintaining stable loads provided both methodological control—ensuring that observed performance changes could be attributed to the training method itself—and practical alignment with the constraints of applied high-performance environments.

Figure 2.

Complex training method.

Statistical Analyses

The G*power (v3.1.9.6, Kiel University, Kiel, Germany) software was used to determine a priori sample size. To achieve an F effect size level of 0.7 (α = 0.05; power = 0.95), it was necessary to recruit 20 subjects. The analysis began with the calculation of basic descriptive statistics, including means, standard deviations, and confidence intervals, for all variables. The Shapiro–Wilk test was used to assess the normality of distributions. For variables that met the assumption of normality (p > 0.05), 1-way repeated measures analysis of variance (ANOVA) was used to evaluate differences across the 3 time points, followed by Tukey's post hoc multiple comparison tests to identify specific group differences. For non-normally distributed variables (p < 0.05, Shapiro–Wilk), the Friedman ANOVA was used as a nonparametric alternative. Statistical significance was set at p ≤ 0.05 for all analyses. Effect sizes were interpreted using conventional thresholds: for Cohen's d (standardized mean difference), trivial <0.20, small 0.20–0.49, medium 0.50–0.79, and large ≥0.80; for partial eta squared (${\eta }_{\mathrm{p}}^2$) in ANOVA, small ≈ 0.01, medium ≈ 0.06, and large ≈ 0.14 (7). Analysis was performed using STATISTICA 13.1 (TIBCO Software Inc., Palo Alto, CA) and data visualized through PRISM (9.4.1.; GraphPad, San Diego, CA).

Tests Reliability

Previously published work indicates good–excellent test–retest reliability for bilateral CMJ when jump height is derived from flight time and when propulsive force–plate metrics are used (e.g., peak force r ≈ 0.99, CV [coefficent of variation] ≈ 2–5%), whereas RFD shows notably poorer reliability (CV ≈ 13%) (32). For unilateral CMJ, between-session reliability of performance variables is generally acceptable to high (CV 4.05–9.98%), although unilateral jump height can be less stable (often CV ≥ 11% depending on processing thresholds) (26). Electronic timing of linear sprints shows good–excellent reproducibility at short and longer splits: 5-m sprint time typically demonstrates good reliability (ICC [intraclass correlation] ≈ 0.71; CV ≈ 2.8%) (13), while a 20–30-m split recorded with Witty gates can reach ICC ≈ 0.94 (13). For the Running-based Anaerobic Sprint Test (RAST; classically 6 × 35 m), mean and peak power and impulse display good relative reliability (e.g., ICC = 0.88, 0.72, and 0.93; CV ≈ 5.9–10.2%) (10).

Results

Descriptive statistics for all dependent variables are provided in Table 1. Normality test for the analyzed variables by measurement time shown in Table 2.

Table 1.

Basic descriptive statistics for the analyzed variables by measurement time.^*

Variable	Postpreparation	Midseason	Postseason	Within-time effect
	M ± SD (±95% CI)
Concentric peak velocity (m·s−1)	2.79 ± 0.30 (2.66–2.93)	2.79 ± 0.29 (2.66–2.92)	2.84 ± 0.40 (2.66–3.02)	χ2 = 0.024; p = 0.98
Concentric peak force (N·kg−1)	26.00 ± 3.11 (24.58–27.41)	26.13 ± 3.33 (24.62–27.65)	26.36 ± 3.48 (24.77–27.94)	χ2 = −0.38; p = 0.83
RFD (N·s−1)	5,587.29 ± 2,808.19 (4,309.01–6,865.56)	5,242.05 ± 2,682.03 (4,021.20–6,462.90)	6,397.05 ± 2,973.85 (5,043.37–7,750.73)	χ2 = 2.00; p = 0.37
Height (cm)	36.08 ± 4.59 (33.99–38.17)	36.48 ± 4.46 (34.45–38.51)	35.99 ± 4.09 (34.12–37.85)	F = 0.08; p = 0.99
Left leg (LL)	20.00 ± 2.44 (18.89–21.11)	21.78 ± 3.07 (20.38–23.18)	21.16 ± 3.87 (19.40–22.92)	F = 1.69 p = 0.19
Right leg (RL)	19.49 ± 2.78 (18.22–20.75)	21.81 ± 2.69 (20.59–23.04)	22.34 ± 2.78 (21.07–23.60)	F = 6.39; p = 0.0031; ${\eta }_{\mathrm{p}}^2$ = 0.18
1 (s)	4.73 ± 0.16 (4.66–4.80)	4.63 ± 0.27 (4.49–4.77)	4.78 ± 0.23 (4.67–4.88)	χ2 = 9.83; p < 0.01
2 (s)	4.83 ± 0.17 (4.75–4.91)	4.71 ± 0.22 (4.60–4.81)	4.90 ± 0.25 (4.78–5.01)	F = 3.93; p = 0.025; ${\eta }_{\mathrm{p}}^2$ = 0.12
3 (s)	4.93 ± 0.19 (4.84–5.01)	4.77 ± 0.22 (4.66–4.88)	4.98 ± 0.22 (4.88–5.08)	F = 5.21; p = 0.0083; ${\eta }_{\mathrm{p}}^2$ = 0.15
4 (s)	5.00 ± 0.19 (4.92–5.09)	4.83 ± 0.21 (4.72–4.94)	5.05 ± 0.22 (4.95–5.15)	F = 5.90; p = 0.0047; ${\eta }_{\mathrm{p}}^2$ = 0.17
5 (s)	5.07 ± 0.21 (4.98–5.17)	4.93 ± 0.24 (4.81–5.05)	5.09 ± 0.22 (4.99–5.19)	F = 2.82; p = 0.068
6 (s)	5.10 ± 0.21 (5.00–5.19)	4.95 ± 0.25 (4.82–5.07)	5.09 ± 0.23 (4.99–5.20)	F = 2.70; p = 0.076
RAST (s)	29.67 ± 1.05 (29.19–30.15)	28.81 ± 1.33 (28.15–29.47)	29.88 ± 1.32 (29.28–30.48)	F = 3.99; p = 0.024; ${\eta }_{\mathrm{p}}^2$ = 0.12
5 m (s)	1.19 ± 0.08 (1.15–1.23)	1.15 ± 0.07 (1.12–1.19)	1.21 ± 0.07 (1.18–1.25)	F = 2.97; p = 0.059
30 m (s)	4.66 ± 0.19 (4.57–4.74)	4.60 ± 0.27 (4.47–4.73)	4.78 ± 0.23 (4.67–4.88)	χ2 = 9.83; p < 0.01

^*M = mean; SD = standard deviation; CI = confidence interval.

Table 2.

Normality test results.^*

Variable	Postpreparation	Midseason	Postseason
	P (S-W)
Concentric peak velocity (m·s−1)	0.03	0.80	0.37
Concentric peak force (N·kg−1)	0.02	0.13	0.41
RFD (N·s−1)	0.002	0.009	0.16
Height (cm)	0.81	0.70	0.10
Left leg (LL)	0.12	0.33	0.63
Right leg (RL)	0.37	0.88	0.43
1 (s)	0.10	0.011	0.52
2 (s)	0.82	0.07	0.17
3 (s)	0.82	0.26	0.55
4 (s)	0.10	0.52	0.28
5 (s)	0.39	0.39	0.61
6 (s)	0.19	0.68	0.45
RAST (s)	0.13	0.17	0.35
5 m (s)	0.67	0.29	0.61
30 m (s)	0.62	0.023	0.52

^*RFD = rate of force development; P S-W = p value for Shapiro–Wilk normality test.

Countermovement Jump

Concentric peak velocity (CPV; m·s⁻¹), concentric peak force (CPF; N·kg⁻¹), and (RFD; N·s⁻¹) did not differ across measurement time (Friedman tests: CPV, χ²(2, N = 21) = 0.024, p = 0.98; CPF, χ²(2, N = 21) = 0.38, p = 0.83; RFD, χ²(2, N = 21) = 2.00, p = 0.37). Countermovement jump height was likewise unchanged (F = 0.08, p = 0.99; Table 1).

Single-Leg Jump

Single-leg jump height for the left leg showed no effect of time (F = 1.69, p = 0.19). In contrast, right-leg performance differed by time (F = 6.39, p = 0.0031, ${\eta }_{\mathrm{p}}^2$ = 0.18): midseason (21.81 ± 2.69 cm; p = 0.022, ES = 0.85) and end of season (22.34 ± 2.78 cm; p = 0.004, ES = 1.03) were both higher than postpreparation (19.49 ± 2.78 cm), with no difference between mid- and end season (p = 0.81; Figure 3).

Figure 3.

Comparison of results for the RL variable with respect to measurement time. RL = right leg.

Speed

Five-meter (5 m) sprint time did not differ across time (F = 2.97, p = 0.059). The 30-m sprint time did differ (χ² [2, N = 21] = 9.83, p = 0.0073), with midseason faster than end of season (4.60 ± 0.27 vs. 4.78 ± 0.23 seconds; Figure 4).

Figure 4.

Comparison of results for 30-m variable with respect to measurement time.

Repeated Anaerobic Sprint Test

RAST1 differed by time (χ² [2, N = 21] = 9.83, p = 0.0073), with midseason faster than end of season (4.63 ± 0.27 vs. 4.78 ± 0.23 seconds). RAST2 (F = 3.93, p = 0.025, ${\eta }_{\mathrm{p}}^2$ = 0.12) and RAST3 (F = 5.21, p = 0.0083, ${\eta }_{\mathrm{p}}^2$ = 0.15) were also faster midseason than end of season (4.71 ± 0.22 vs. 4.90 ± 0.25 seconds, p = 0.020, ES = 0.81; 4.77 ± 0.22 vs. 4.98 ± 0.22 seconds, p = 0.008, ES = 0.95). RAST4 differed (F = 5.90, p = 0.0047, ${\eta }_{\mathrm{p}}^2$ = 0.17), being faster midseason than end of season (4.83 ± 0.21 vs. 5.05 ± 0.22 seconds, p = 0.005, ES = 1.02) and faster midseason than postpreparation (5.00 ± 0.19 seconds, p = 0.029, ES = 0.85), with no difference between postpreparation and end season (p = 0.77). RAST5 (F = 2.82, p = 0.068) and RAST6 (F = 2.70, p = 0.076) showed no time effect. Total time (RAST_total) differed by time (F = 3.99, p = 0.024, ${\eta }_{\mathrm{p}}^2$ = 0.12), with midseason faster than end season (28.81 ± 1.33 vs. 29.88 ± 1.32 seconds, p = 0.025, ES = 0.81) and no differences involving postpreparation (vs. midseason p = 0.088; vs. end season p = 0.84; Figures. 5‒7).

Figure 5.

Comparison of results for variable 2, 3, 4 (RAST) because of measurement time.

Figure 7.

Comparison of results for variable 1 (RAST) with respect to measurement time.

Figure 6.

Comparison of results for the RAST variable with respect to measurement time.

Discussion

This study investigated the feasibility and seasonal effects of embedding once-weekly CCT sessions within the competitive microcycle of elite female soccer players. Without a control group or acute potentiation measures, the findings do not establish comparative effectiveness or confirm PAPE. Instead, they provide descriptive evidence of performance trajectories under a pragmatic, in-season program.

Sprint (30 m) and repeated-sprint ability (RAST) improved by midseason but declined toward season end, whereas CMJ-derived indices (CPV, CPF, RFD, and CMJ height) remained stable. A unilateral advantage was observed in the right (dominant) leg. These results suggest that low-frequency CCT can selectively support speed-related outcomes during the season, whereas force-related metrics may require a greater training frequency or volume. The late-season return toward baseline likely reflects cumulative in-season fatigue and constrained recovery associated with fixture density and a shift in training time toward tactical/match demands. Importantly, maintaining CMJ-derived force characteristics on just 1 structured CCT session per week represents a practically meaningful outcome in congested in-season schedules, while further improvements in CPF and RFD may necessitate higher weekly exposure to heavy loading.

Complex-contrast training combines high-load resistance and explosive plyometric movements, spanning the force–velocity spectrum and offering an efficient transfer to match actions requiring force and speed (8,30). Systematic reviews indicate that CCT can yield moderate-to-large effects on sprint, jump, and agility performance (35). In soccer, where repeated accelerations, decelerations, and vertical duels are decisive (18,31), CCT provides a time-efficient, in-season stimulus. Our observation of midseason gains in sprint and RAST mirrors previous team-sport evidence showing preferential transfer of contrast protocols to velocity-specific capacities (5,20).

Although we did not directly measure potentiation, CCT is commonly justified by PAPE, in which a heavy conditioning contraction temporarily enhances the subsequent explosive output by increasing motor unit recruitment, myosin phosphorylation, and Ca²⁺ sensitivity (3,36). Reviews have confirmed its relevance for explosive actions, such as sprinting and jumping (15,25). However, the potentiation response depends strongly on the rest interval, exercise selection, and strength status (6,25). In this study, 1-minute rest intervals were chosen for feasibility within in-season schedules, but such short transitions may not maximize the potentiation effect. This trade-off emphasizes the tension between scientific optimization and applied practicality in elite football settings.

To our knowledge, this is the first study to evaluate a comprehensive in-season complex-training regimen in elite female soccer players. In elite male soccer, Jarosz et al. (23) demonstrated an in-season PAPE response at mid-season for a 4-week intervention. Although CMJ height improved, the change did not correlate with Global Positioning System (GPS)-derived external load, suggesting that load magnitude alone is not the primary driver of in-season PAPE adaptations. However, these interventions remain untested in women.

Such interventions remain largely untested in women. This gap is significant because elite female players face not only similar match demands but also a greater injury burden, especially noncontact anterior cruciate ligament (ACL) ruptures, which occur at a rate 2–8 times higher than in men (17,21). Anatomical and biomechanical differences, including a wider pelvis-to-femur angle, greater dynamic knee valgus, and altered neuromuscular activation patterns, are thought to underpin this disparity (21). Consequently, sex-specific adaptations of strength and power training methods are urgently needed to sustain performance and mitigate injury risk. The current findings, demonstrating the feasibility and benefits of CCT during the competitive season in women, contribute novel insights into this under-researched area.

Evidence from other team sports indicates that strength status moderates the response to complex training, and this helps explain our pattern of results. In male handball/basketball samples, complex training produced clearer gains in weaker athletes and uncertain effects in stronger athletes (1). The deterioration in sprint and RAST performance by season end likely reflects cumulative fatigue from dense competition and reduced recovery, patterns also reported in female leagues (22,27). Observational studies in Finnish female players have indicated that performance maintenance depends more on progressive intensity in very high-intensity running and accelerations/decelerations than on the total training load (27). This underscores the importance of managing both load distribution and recovery strategies to preserve neuromuscular performance at the end of the season.

Converging evidence from rugby league shows that both variable-resistance and traditional complex training improve strength, power, and short-sprint performance with comparable gains in back-squat 1RM (29). Our data align with this transfer pattern: the complex stimulus mapped onto sprint-relevant outcomes (30 m, RAST_total) more readily than onto CMJ-derived force characteristics. Practically, this supports retaining a weekly complex session during in-season microcycles to target speed and repeated-sprint capacity, while in parallel progressing individualized strength work (and managing late-season load) to sustain adaptations beyond the midseason peak.

Midseason performance enhancements followed by end-of-season declines underscore the need to balance training stimulus, match exposure, and recovery. Without deliberate load management, the benefits of complex training can be offset by cumulative fatigue, particularly in anaerobic and sprint-specific capacities. Furthermore, strength-based programs are essential for injury prevention in women's soccer. Darragi et al. (9) showed that a 12-week in-season ST program in elite young female players improved neuromuscular fitness and reduced noncontact injury incidence compared with controls. Moreover, study indicated that noncontact injuries during the season were significantly lower (p = 0.003, d = 1.31) in the STG (0.48/1,000 hours of exposure) than in the CG (2.62/1,000 hours of exposure) (9). Although our study did not measure injuries, the stability of CMJ indices suggests that even low-frequency CCT can help maintain neuromuscular readiness, which is relevant for injury mitigation. Given the high incidence of ACL injuries in women, combining CCT with dedicated strength training may offer a dual benefit of preserving performance while addressing one of the sport's most pressing health challenges.

Across the season, we observed an upward trend in power-related indicators from the preparatory phase to midseason—evident in height (CMJ/SLJ), concentric peak velocity (CPV; m·s⁻¹), concentric peak force (CPF; N·kg⁻¹), and rate of force development (RFD; N·s⁻¹)—followed by a plateau from midseason to season's end. This pattern suggests that, under in-season constraints, the training load was sufficient to drive early adaptation but insufficient to sustain further gains once accumulated fatigue increased.

Unilateral analyses showed a persistent advantage of the right limb over time. Given that all athletes designated the right leg as dominant, the larger gains in right-leg SLJ are best interpreted as a dominance-dependent, local PAPE effect—consistent with evidence that unilateral conditioning primarily benefits the activated limb and yields at most limited, task-specific contralateral changes—rather than as a generalized bilateral adaptation (33). This finding supports routine monitoring of interlimb differences and the inclusion of unilateral strength and plyometric work to manage asymmetry. The right-leg advantage observed here is consistent with dominant limb reliance and highlights the need for unilateral strength and plyometric training to address asymmetries, which can otherwise exacerbate injury risk (38).

The ability to draw conclusions is restricted by the lack of a control group, small sample size, and absence of measures for acute potentiation. Future studies should directly compare CCT with alternative in-season methods, including acute mechanistic assessments, and integrate external and internal load monitoring. Expanding such research in male cohorts is crucial for developing sex-specific, evidence-based training strategies that balance performance with injury prevention.

Rest intervals were fixed and short, and we did not individualize them by strength status nor verify potentiation acutely; thus, the protocol likely underestimates the maximal PAPE that longer or tailored rests might produce. Prior work shows substantial interindividual variability and context dependence, with effective windows ranging from <1 to >20 minutes, moderated by strength status, exercise selection, and fatigue–potentiation balance (19,30,36). We purposefully prioritized ecological validity within an in-season MD + 3 microcycle: keeping the complex block brief to fit staff and facility constraints, maintain session density, and preserve warm-up temperature without extending total training time. We also did not individualize rest based on strength or on acute potentiation testing, which likely attenuated immediate potentiation in some players. Consequently, our outcomes should be interpreted as the training effects of a feasible, time-efficient CCT configuration, not as evidence for optimal PAPE timing. Future studies should compare individualized vs. fixed rests and incorporate within-session potentiation diagnostics to titrate recovery windows.

Practical Applications

Embedding 1 weekly CCT session (MD + 3) during the in-season period was feasible within the existing workload and was accompanied by short-term improvements in sprint and Running-based Anaerobic Sprint Test (RAST) performance for the first 6 weeks. At the late-season assessment, these improvements were no longer evident. Given our study design, we cannot attribute this pattern to specific recovery or progression strategies. Moreover, because we did not compare CCT with alternative training methods, our findings indicate that once-weekly CCT can be sufficient to elicit in-season gains in selected performance measures, but do not establish whether it is better or worse than other approaches.