Low-dose caffeine enhances cognitive processing but not physical performance in fatigued taekwondo athletes: a randomized crossover trial

ABSTRACT

Background

Caffeine is commonly used to combat fatigue and enhance both cognitive and physical performance. However, its effects on neurophysiological responses and sport-specific performance following fatigue induction remain unclear, particularly in combat sports such as Taekwondo. This study investigated the effects of a 200 mg caffeine dose on physiological markers, electroencephalographic (EEG) brainwave activity, auditory P300 event-related potentials (ERPs), and Taekwondo-specific performance following combined mental and physical fatigue.

Methods

Thirteen male Taekwondo athletes participated in a randomized, double-blind, crossover study with caffeine (CAF) and placebo (PLA) conditions. Measurements were taken at baseline (pre-supplementation), 30 minutes post-supplementation (post-Sup), and after fatigue induction (post-I). Physiological parameters (heart rate, blood glucose, blood lactate, and ratings of perceived exertion), EEG brainwave activity during resting eyes-open conditions, auditory P300 ERPs, and Taekwondo-specific agility (TSAT) were assessed at all time points.

Results

Caffeine significantly reduced delta wave power at frontal and parieto-occipital sites at post-Sup (p < 0.05), indicating decreased cortical drowsiness; however, this effect was not sustained at post-I (p > 0.05). P300 amplitude significantly increased in the CAF condition compared to PLA from post-Sup to post-I at the central and parietal electrode sites (p < 0.05), while P300 latency remained unchanged (p > 0.05). No significant differences were observed in reaction time, accuracy, or error rate in the auditory oddball task or TSAT performance across conditions (p > 0.05). Similarly, physiological parameters remained unchanged between groups (p > 0.05).

Conclusion

A single 200 mg dose of caffeine reduced central fatigue and enhanced cognitive processing, as reflected by suppressed delta wave activity at post-Sup and increased P300 amplitude at post-I. However, caffeine did not influence physiological responses or Taekwondo-specific performance. These findings suggest that low-dose caffeine primarily benefits cognitive function rather than physical performance in combat sports. Future studies should explore dose-response relationships and individual variability in caffeine metabolism to optimize its application in competitive settings.

1. Introduction

Taekwondo is a high‐intensity combat sport characterized by explosive movements, rapid decision‐making, and sustained cognitive engagement. Success in competition depends on an athlete’s ability to maintain peak physical and mental performance despite accumulating fatigue, which can impair reaction time, motor execution, and decision-making [1]. Taekwondo athletes experience both peripheral fatigue, resulting from energy depletion and impaired muscle contractility, and central fatigue, which stems from neurochemical changes in the central nervous system (CNS) that reduce alertness and cognitive efficiency [2,3]. Although fatigue has been widely studied in endurance and intermittent sports, its combined impact on cognitive functioning and performance in combat sports remains insufficiently understood. Identifying effective strategies to mitigate these impairments is essential for optimizing competitive performance.

Fatigue in sport is broadly classified into peripheral and central components. Peripheral fatigue arises from muscle dysfunction and metabolic disturbances, while central fatigue is linked to altered neurotransmitter activity, such as increased serotonin and decreased dopamine and norepinephrine, leading to reduced motivation and impaired motor control [2,4–7]. These physiological and cognitive disruptions are particularly detrimental in fast-paced sports like Taekwondo, where precision, reaction speed, and sustained focus are critical. Therefore, effective countermeasures must address both physical and mental fatigue in high-stress competitive settings.

Caffeine is one of the most widely studied ergogenic aids in sport, with extensive evidence demonstrating its ability to enhance endurance, strength, power, and cognitive function [8–10]. By antagonizing adenosine receptors, caffeine delays the onset of fatigue, heightens alertness, and improves neuromuscular efficiency. Doses between 3 and 6 mg/kg have been shown to increase VO₂max, peak power output, and reaction time in a variety of athletic populations, including cyclists, runners, and team‑sport athletes [11–14]. Meta-analyses have confirmed caffeine’s role in improving reaction time, attention, and working memory, not only in athletes but also in general populations under cognitively demanding conditions [8,15–17]. For instance, Connell et al. (2016) [18] reported that a 5 mg/kg dose significantly improved reaction times following exercise-induced fatigue, highlighting caffeine’s potential to counteract central fatigue and preserve motor performance.

While the ergogenic effects of moderate-to-high caffeine doses are well established, the efficacy of low doses (~3 mg/kg) remains less clear, especially in precision-based sports requiring fine motor control and coordination. Emerging evidence suggests that low doses may improve vigilance and reduce mental fatigue while minimizing side effects commonly associated with higher doses, such as jitteriness or gastrointestinal discomfort [12,19]. In combat sports like Taekwondo, where both technical execution and cognitive sharpness are paramount, it is important to evaluate whether low-dose caffeine can enhance performance under fatigue without compromising skill execution. This study, therefore, investigated whether a single 200 mg dose (~3 mg/kg) could enhance neurocognitive function and Taekwondo-specific performance following combined physical and mental fatigue.

Despite growing evidence, few studies have explored caffeine’s neurophysiological effects in sport-specific, fatigue-inducing contexts. Electroencephalography (EEG) provides a sensitive tool for assessing CNS responses under fatigue by capturing brainwave dynamics. Central fatigue is commonly associated with increased delta power and reduced theta, alpha, and beta activity – markers of decreased cortical arousal and cognitive readiness [20–22]. In addition, event-related potentials (ERPs), particularly the P300 component, a positive deflection around 300 ms post-stimulus, serve as reliable indices of attentional resource allocation and cognitive processing. Increased P300 latency and reduced amplitude are indicative of cognitive decline under stress, whereas caffeine has been shown to improve both, suggesting restoration of attentional control [23–25]. However, most studies to date have focused on general or non-athlete populations and lack sport-specific neurophysiological evaluation [26,27].

Given the unique physical and cognitive demands of Taekwondo, it remains unclear whether low-dose caffeine can effectively counteract fatigue and support performance. Therefore, this study aimed to investigate the effects of a single low-dose caffeine supplementation (approximately 3 mg/kg body weight, standardized as a 200 mg dose) on physiological markers, EEG brainwave activity, auditory P300 ERPs, and Taekwondo-specific performance following experimentally induced physical and central fatigue. By addressing this gap, the findings may offer valuable insight into the sport-specific use of caffeine as a strategy to preserve cognitive and motor performance under fatigue.

2. Materials and methods

2.1. Participants

This study employed a double-blind, randomized crossover design consisting of two experimental trials: a caffeine trial and a placebo trial, separated by a one-week washout period. Sample size calculation was based on Siepmann and Kirch [28] using a Cohen’s f effect size of 1.05, an alpha level of 0.05, and a statistical power of 0.95. After accounting for a 10% dropout rate, the required sample size was determined to be 13 participants.

Thirteen male university-level Taekwondo athletes, aged 18–25 years, were recruited and completed the study. Data collection occurred in multiple sessions between December 2019 and February 2023, depending on athlete availability during academic and training periods. Each participant’s age was recorded at the time of their involvement. All athletes competed in the under 58 kg, 63 kg, or 68 kg weight categories. Inclusion criteria included a VO₂max >40 ml/kg/min [29], 2–5 years of Taekwondo training experience, 9–15 hours of weekly training, collegiate competition experience, and low habitual caffeine intake (<150 mg/day). Athletes were excluded if they had a history of clinical disease, musculoskeletal disorders, adverse reactions to caffeine, or medication use in the past three months. All participants underwent a physical examination and provided written informed consent prior to participation. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Mahidol University Ethics Committee (MU-CIRB 2019/237.0909) and registered with the Thai Clinical Trials Registry (TCTR20250319009).

2.2. Experimental design

Participants were instructed to abstain from strenuous physical activity or sports for at least 24 hours, sleep for at least 8 hours, and abstain from CNS-acting substances (e.g. caffeine, alcohol, medications) for at least 24 hours before each experimental session. To minimize nutritional variability, they also maintained and replicated a 48-hour dietary log prior to both trials. The study included one preliminary visit and two experimental sessions: a caffeine (CAF) trial and a placebo (PLA) trial, conducted in a double-blind, randomized crossover design.

2.2.1. Preliminary visit

During the preliminary session, VO₂max was assessed using the Bruce treadmill protocol, performed until volitional exhaustion. The highest 30-second average VO₂max value was recorded to confirm eligibility, ensuring participants met the aerobic fitness requirement for Taekwondo athletes (VO₂max >40 ml/kg/min).

2.2.2. Experimental sessions

Each session included three assessment time points: baseline (pre-Sup), post-supplementation (post-Sup, ~60 min after ingestion), and post-fatigue induction (post-I, immediately after fatigue protocol).

At the start of each session, baseline (pre-Sup) data were collected, including1) physiological parameters: heart rate (HR), blood glucose (BG), blood lactate (BL), rating of perceived exertion (RPE), 2) neurophysiological measures: EEG recordings, auditory oddball P300 event-related potentials (P300-ERPs), and 3) performance assessments: Taekwondo-specific reaction time and kicking speed.

Following baseline assessments, participants ingested either 200 mg of caffeine (two 100 mg capsules) or a visually identical placebo, assigned in random order. They then rested in a quiet, temperature-controlled room for one hour to allow for caffeine absorption. After this resting period, post-Sup assessments were conducted, repeating the same physiological, neurophysiological, and performance measures.

Fatigue was then induced using a combined cognitive (Corsi Block-Tapping Test) and physical (Taekwondo-specific exercise protocol) challenge, immediately followed by post-I assessments (see Figure 1).

Figure 1.

The experimental procedure of each trial.

Abbreviations: HR = Heart rate, RPE = Rating perceived exertion scale, BGL = Blood glucose level, TSAT = Taekwondo-specific agility test, Electroencephalogram =EEG, ERP= Event-related potential, CBT = Corsi block-tapping

2.2.3. Central fatigue induction: Corsi Block-Tapping (CBT) test

Central fatigue was induced following the post-Sup assessments using the CBT test, as described by Zhu et al. [30]. Participants were seated with their heads supported by a chin rest to maintain a fixed 60 cm viewing distance from a 17-inch monitor (resolution: 1280 × 1024 pixels; Dell E1715S, China). During the test, participants replicated the sequence of illuminated blocks presented on the screen by clicking the corresponding blocks in the same order. The task began with a two-block sequence and increased in difficulty if the participant correctly recalled at least one of two trials. The test was terminated when a participant failed to recall the sequence in two consecutive trials at a given length. Each block remained illuminated for 500 milliseconds, with a 1000-millisecond inter-stimulus interval, following the standardized approach by Robinson and Brewer [31]. The CBT was administered immediately prior to the physical fatigue protocol and post-I assessments.

2.2.4. Physical fatigue induction: Taekwondo-specific exercise protocol

Immediately following the CBT test, physical fatigue was induced using a 10-minute Taekwondo-specific exercise protocol adapted from Bridge et al. [32], designed to simulate sport-specific physiological demands. The protocol was implemented approximately 65–70 minutes after caffeine or placebo ingestion, following all pre-fatigue assessments, including EEG and ERP recordings. The protocol consisted of three rounds (R1, R2, R3), each comprising a 2-minute warm-up and three 2-minute Taekwondo-specific exercise bouts, separated by 1-minute rest intervals. The exercise involved technical and skill-based drills, including turning kicks (left and right), rapid footwork, and combat-specific actions to replicate the physical and metabolic demands of competitive Taekwondo. This session was completed immediately before the post-I assessments.

2.2.5. Physiological responses

HR was monitored using a chest strap telemetry device (Polar FT1, Finland). Capillary blood samples were collected from the lateral side of the non-dominant ring finger using sterile fingertip lancets at baseline and all assessment time points. These samples were analyzed for blood glucose (BG) using an ACCU-CHEK glucose meter (Switzerland) and for blood lactate (BL) using a Lactate Scout analyzer (Germany) [33]. RPE was assessed using the 6–20 Borg scale, with participants reporting their perceived exertion at each time point [34]. RPE scores were categorized according to the American College of Sports Medicine (ACSM) guidelines: very light (<9), light (9–11), moderate (12–13), vigorous (14–17), and near-maximal to maximal (18–20).

2.2.6. Neurophysiological measures

2.2.6.1. Electroencephalogram (EEG)

EEG was recorded in a temperature-controlled (22–24°C), low-light (<150 lux), and quiet environment. To ensure signal quality, participants were required to wash their hair and refrain from using hair products. They were seated comfortably in front of a screen, and resting EEG was recorded for 5 minutes under eyes-open conditions. EEG data were recorded from 32 electrode sites, focusing on five midline electrode sites: frontopolar (Fpz), frontal (Fz), central (Cz), parietal (Pz), and occipital (Oz) [35].

Artifacts from eye movements, muscle activity, or electrode shifts were automatically cleaned using the artifact correction function of the asa ^TM erp software program. Pre-recording parameters were set to a bandpass frequency of 0.1–60 Hz, with a 50 Hz notch filter to remove noise. Post-recording parameters were set to a bandpass between 0.3 Hz (slope 12 dB/oct) and 30 Hz (slope 24 dB/oct). The absolute power (µV2) of the respective frequency bands, derived by fast-Fourier transforms (FFTs), was defined for the delta (0.5–4 Hz), theta (4.5–8 Hz), alpha (8.5–13 Hz), and beta (13.5–29.5 Hz) wave ranges. For FFT analysis, the software program used the time window of recording without overlapping and set a block length of 2 seconds [26,36].

2.2.6.2. Auditory P300 event-related potentials (ERPs)

Following EEG recording, P300 ERPs were assessed using an auditory oddball paradigm to evaluate attentional processing. Participants heard a randomized sequence of standard low-pitched tones (1000 Hz) and target high-pitched tones (2000 Hz) via headphones. They were instructed to press a button in response to the target tones. Each session lasted approximately 6 minutes and included 160 trials (80% standard, 20% target). Behavioral responses (accuracy and reaction time [RT]) were recorded [35]. ERP data were analyzed using Cognitech software to extract P300 peak latency and amplitude at Fz, Cz, and Pz. Peak amplitude was defined as the voltage difference from baseline to the maximum positive peak between 180–600 ms post-stimulus [37]. Latency was defined as the time from stimulus onset to the P300 peak.

2.2.7. Performance assessments

2.2.7.1. Taekwondo-Specific Agility Test (TSAT)

Agility performance was evaluated using the TSAT, a validated test for sport-specific movement patterns [38]. The test was conducted at three time points: pre-Sup, post-Sup, and post-I. Participants began each trial in a guard position behind the start/finish line. The movement sequence involved advancing to a center point, pivoting toward partner 1 to perform a left-leg roundhouse kick, then progressing to partner 2 for a right-leg kick. They returned to the center, then advanced again to deliver a double-roundhouse kick at partner 3, before retreating to the starting line (Figure 2).

Figure 2.

Graphic representation of the Taekwondo-Specific agility test (TSAT). Adapted from chaabene et al. [38].

Each participant completed two trials per time point, with a 2-minute rest between attempts. The faster trial was used for analysis. Quantitative performance was based solely on total completion time, measured with a stopwatch. While technique and target accuracy were visually monitored for consistency, they were not scored, consistent with TSAT use in prior sport-specific performance research [38].

2.3. Statistical analysis

The normality of the data was assessed using the Shapiro – Wilk test, while Levene’s test was used to verify the homogeneity of variances. For normally distributed variables, a repeated measures analysis of variance (ANOVA) was performed to examine the main effects of condition (CAF vs. PLA), time (pre-Sup, post-Sup, post-I), and condition × time interactions on physiological parameters (e.g. HR, RPE), EEG frequency bands, and Taekwondo-specific performance outcomes. When significant effects were identified, post hoc pairwise comparisons were conducted using Bonferroni correction.

To compare the percentage change (%∆) in P300 amplitude and latency from post-Sup to post-I between CAF and PLA conditions, independent samples t-tests were performed. For non-normally distributed data, the Friedman test was used, followed by Wilcoxon signed-rank tests for post hoc comparisons. Effect sizes were reported as partial eta-squared (η² p) for repeated measures ANOVA and Cohen’s d for independent t-tests. Statistical significance was set at p < 0.05. All analyses were conducted using SPSS version 28.0 (IBM, USA).

3. Results

A total of thirteen male Taekwondo athletes participated in the study, with a mean age of 20 ± 1 years. Participants had an average weight of 58.9 ± 6.4 kg, height of 1.73 ± 0.53 m, and body mass index (BMI) of 19.9 ± 2.2 kg/m². Their aerobic fitness level, indicated by VO₂max, averaged 50.3 ± 4.9 ml/kg/min, while resting heart rate (RHR) was 78 ± 15 bpm. Reported habitual caffeine consumption was 3.43 ± 0.35 mg/kg. Based on dietary logs collected during screening, most participants consumed caffeine between 10:00 AM and 2:00 PM, whereas caffeine ingestion in the experimental sessions was standardized at 08:30 AM. This timing ensured overnight caffeine clearance and minimized interference from habitual intake patterns. Participants were instructed to abstain from all caffeine-containing products for at least 12 hours prior to each trial to control for carry-over effects.

3.1. Physiological responses

Table 1 presents physiological parameters, including HR, RPE, BG, and BL, across all time points.

Table 1.

Mean ± SEM of physiological parameters, including heart rate (HR), rating of perceived exertion (RPE), blood glucose (BG), and blood lactate (BL), measured at baseline (pre-sup), post-supplementation (post-sup), during each round (R1, R2, R3) of the Taekwondo-specific exercise protocol, and post-fatigue induction (post-I) under caffeine (CAF) and placebo (PLA) conditions. Significant difference from previous time point, p < 0.05. (n = 13).

Variables		pre-Sup	post-Sup	Taekwondo exercise
				R1	R2	R3	post-I
HR (bpm)	CAF	78 ± 4	74 ± 4	139 ± 5*	151 ± 7	162 ± 5	89 ± 5*
	PLA	79 ± 5	74 ± 4	138 ± 5*	148 ± 5	146 ± 9	85 ± 5*
RPE (score)	CAF	6 ± 0	7 ± 0	10 ± 1*	11 ± 1	13 ± 1	7 ± 0*
	PLA	6 ± 0	7 ± 0	9 ± 1*	11 ± 1	12 ± 1	7 ± 0*
BG (mmol/L−1)	CAF	97 ± 2	93 ± 3	86 ± 3	86 ± 4	85 ± 4	99 ± 3
	PLA	104 ± 5	91 ± 3	84 ± 3	82 ± 3	82 ± 4	92 ± 2
BL (mmol/L−1)	CAF	5.4 ± 0.7	6.9 ± 0.1	10.6 ± 1.2	10.4 ± 1.2	11.6 ± 1.4	7.0 ± 1.2
	PLA	4.5 ± 0.8	5.9 ± 1.3	8.4 ± 1.4	10.9 ± 1.6	9.6 ± 1.2	6.4 ± 1.0

*Significant difference from the previous condition; p < 0.05.

For HR, repeated measures ANOVA revealed a significant main effect of time (F_(5,144) = 94.03, p < 0.0001, η² p = 0.77), but no significant condition × time interaction (p > 0.05). In the CAF condition, HR slightly decreased from pre-Sup to post-Sup, then significantly increased from post-Sup to R1 (p < 0.0001). HR further increased during R2 and R3, followed by a significant reduction from R3 to post-I (p < 0.0001). A similar pattern was observed in the PLA condition.

RPE also showed a significant effect of time (F_(5,144) = 45.28, p < 0.0001, η² p = 0.61), indicating increased perceived exertion during the exercise protocol. In the CAF condition, RPE significantly increased from post-Sup to R1 (p < 0.0001), continued rising through R2 and R3, then significantly decreased at post-I (p < 0.0001). The PLA condition showed a comparable trend.

No significant main effects or interactions were found for BG or BL (p > 0.05). BG levels remained relatively stable across all time points in both CAF and PLA conditions, with minor decreases during the exercise bouts and a slight rebound at post-I. Similarly, BL levels increased modestly during exercise but showed no significant differences across time or between conditions.

3.2. Neurophysiological responses

3.2.1. EEG brain wave activity during resting state with eyes open

The mean absolute power (μV²) of delta, theta, alpha, and beta brain waves was analyzed across pre-Sup, post-Sup, and post-I for both the caffeine (CAF) and placebo (PLA) conditions (Table 2).

Table 2.

Mean ± SEM of absolute brainwave power (μV²) during resting state with eyes open, including delta, theta, alpha, and beta bands at baseline (pre-sup), post-supplementation (post-sup), and post-fatigue induction (post-I), for both caffeine (CAF) and placebo (PLA) conditions (n = 13).

Electrode	Brain Wave	Resting Mean absolute power of brain area (µV 2)
		PLA			CAF
		pre-Sup	post-Sup	post-I	pre-Sup	post-Sup	post-I
Fpz	Delta	110.9 ± 7.9	102.2 ± 8.8	124.4 ± 7.6	118.7 ± 11.4	74.8 ± 5.6*#	97.3 ± 9.6
	Theta	11.4 ± 1.9	11.2 ± 1.3	10.3 ± 1.4	9.2 ± 0.7	8.2 ± 0.9	12.0 ± 4.2
	Alpha	16.7 ± 3.2	15.2 ± 2.0	19.9 ± 4.4	14.3 ± 2.1	14.3 ± 1.6	18.9 ± 3.5
	Beta	6.7 ± 0.7	8.3 ± 1.1	7.1 ± 0.8	8.4 ± 1.3	7.3 ± 1.1	7.8 ± 1.1
Fz	Delta	59.5 ± 4.3	55.8 ± 5.5	59.6 ± 5.3	63.1 ± 4.6	42.4 ± 2.0*	65.9 ± 10.7
	Theta	14.5 ± 2.9	12.1 ± 1.5	12.2 ± 1.7	10.8 ± 1.0	10.1 ± 1.2	12.9 ± 2.8
	Alpha	21.6 ± 3.7	18.8 ± 2.2	26.5 ± 5.5	18.6 ± 2.7	18.7 ± 1.9	24.3 ± 3.7
	Beta	7.1 ± 0.9	7.5 ± 1.1	7.3 ± 1.0	7.4 ± 1.0	6.9 ± 1.0	8.4 ± 1.4
Cz	Delta	53.2 ± 5.8	41.9 ± 3.0	57.5 ± 12.8	50.3 ± 3.2	36.1 ± 2.3*	59.1 ± 12.4
	Theta	12.7 ± 2.8	10.6 ± 1.7	11.4 ± 2.1	9.9 ± 1.2	10.1 ± 1.7	10.1 ± 1.8
	Alpha	24.9 ± 5.2	20.5 ± 2.2	28.3 ± 6.0	19.7 ± 2.8	19.6 ± 2.1	25.4 ± 4.0
	Beta	6.6 ± 0.7	7.3 ± 1.0	8.0 ± 1.3	6.9 ± 1.0	6.7 ± 1.0	6.7 ± 0.8
Pz	Delta	54.6 ± 7.0	35.6 ± 3.2*	46.1 ± 5.4	59.6 ± 4.7	34.9 ± 3.2*	54.7 ± 8.8
	Theta	11.2 ± 2.0	10.0 ± 2.3	11.0 ± 2.6	9.37 ± 1.5	10.3 ± 2.7	9.6 ± 2.2
	Alpha	28.7 ± 5.4	33.3 ± 5.2	37.9 ± 6.6	28.7 ± 4.3	31.9 ± 5.0	39.8 ± 7.0
	Beta	7.7 ± 1.0	7.9 ± 1.0	8.4 ± 1.2	7.5 ± 0.9	7.4 ± 1.1	7.8 ± 0.9
POz	Delta	57.8 ± 11.0	36.6 ± 5.4	53.5 ± 12.9	56.3 ± 5.6	34.8 ± 4.5*	47.0 ± 7.3
	Theta	9.3 ± 1.6	8.9 ± 1.9	9.2 ± 2.3	8.8 ± 2.0	9.0 ± 2.7	8.2 ± 2.0
	Alpha	33.2 ± 6.4	39.9 ± 8.5	36.8 ± 5.5	36.6 ± 7.3	42.6 ± 7.8	43.1 ± 8.0
	Beta	8.0 ± 1.2	8.1 ± 1.1	8.7 ± 1.3	8.0 ± 0.9	8.1 ± 1.2	8.6 ± 1.0
Oz	Delta	43.1 ± 9.6	24.0 ± 3.0	32.3 ± 4.5	45.6 ± 6.1	24.8 ± 3.0*	40.3 ± 8.2
	Theta	6.4 ± 1.4	5.2 ± 1.0	5.6 ± 1.2	5.3 ± 0.9	5.4 ± 1.4	4.9 ± 1.2
	Alpha	26.7 ± 5.4	25.0 ± 5.4	23.4 ± 3.9	23.9 ± 5.5	30.1 ± 6.4	39.0 ± 8.7
	Beta	6.0 ± 0.8	6.0 ± 0.9	6.4 ± 1.1	6.3 ± 0.7	6.5 ± 1.1	6.7 ± 0.8

Fpz = the midline frontopolar, Fz = the midline frontal, Cz = the midline central, Pz = the midline parietal and Oz = the midline occipital. * Significant difference from the previous condition; p < 0.05. ^# Significant between the PLA and CAF at post-Sup.

3.2.1.1. Delta waves

A significant condition × time interaction was found at the Fpz electrode (F_(2,48) = 4.730, p = 0.013, η² p = 0.16). Post-hoc analysis revealed a significant reduction in delta power from pre-Sup to post-Sup in the CAF condition (p = 0.004), and significantly lower delta power in CAF compared to PLA at post-Sup (p = 0.048).

A significant main effect of time was also observed at the Fz (F_(1,35) = 3.84, p = 0.04, η² p = 0.14), Cz (F_(1,28) = 3.55, p = 0.06, η² p = 0.13), Pz (F_(2,38) = 11.58, p = 0.0003, η² p = 0.33), POz (F_(1,31) = 5.34, p = 0.024, η² p = 0.18), and Oz (F_(2,43) = 7.86, p = 0.002, η² p = 0.25) electrodes. Post-hoc comparisons indicated significant reductions in delta power from pre-Sup to post-Sup in the CAF condition at Fz (p = 0.003), Cz (p = 0.029), Pz (p = 0.0004), POz (p = 0.001), and Oz (p = 0.009). A smaller but significant decrease was also observed in the PLA condition at Pz (p = 0.018).

At post-I, delta power showed a partial rebound in the CAF condition, increasing by approximately 15–25% across cortical sites but remaining below pre-Sup levels. However, no significant condition × time interaction was found at post-I, suggesting the caffeine-induced reduction in delta power was transient and not maintained after fatigue induction.

3.2.1.2. Theta, alpha, and beta waves

No significant main effects or condition × time interactions were observed for theta, alpha, or beta waves at any electrode. Theta power remained stable across all time points in both conditions, with minor non-significant fluctuations at Fpz and Pz.

3.2.2. P300 event-related potentials (ERPs) – auditory oddball paradigm

Analysis of the P300 components revealed significant differences in percentage change (%∆) in amplitude from post-supplementation (post-Sup) to post-fatigue induction (post-I) between the CAF and PLA conditions. Independent samples t-tests indicated a significantly greater increase in P300 amplitude in the CAF condition at both the Cz (95% CI: −1.62 to 118.10, p = 0.028, η² = 0.224) and Pz (95% CI: −0.59 to 40.60, p = 0.028, η² = 0.243) electrodes. Specifically, in the CAF condition, P300 amplitude increased significantly from post-Sup to post-I at Cz (+27 ± 9%) and Pz (+21 ± 9%), whereas in the PLA condition, amplitude decreased at Cz (−31 ± 27%) and remained relatively unchanged at Pz (+2 ± 2%) (Figure 3).

Figure 3.

Mean ± SEM of percentage change (%∆) from post-caffeine consumption (post-sup) to post-fatigue induction (post-I) of latency and amplitude of the P300 ERP components in auditory oddball paradigm (auditory P300 ERP) at the midline frontopolar (Fpz), the midline central (Cz), and the midline parietal (Pz) in the placebo and caffeine conditions (n = 13 for each condition).

* Different between conditions using independent samples t-tests, p < 0.05.

No significant differences in P300 latency were observed across conditions or electrode sites (p > 0.05). Latency remained stable, with small percentage changes at Fz (+2 ± 2% in PLA, −2 ± 2% in CAF), Cz (−3 ± 3% in PLA, −1 ± 2% in CAF), and Pz (−1 ± 2% in PLA, +2 ± 2% in CAF) (Figure 3).

3.2.3. Behavioral performance: reaction time and accuracy

There were no significant differences in reaction time (RT), accuracy rate, or error rate between the CAF and PLA conditions, or across time points (p > 0.05). RT values remained stable, ranging from 296 ± 8 ms to 325 ± 11 ms, with no main effects of time or condition. Accuracy rates were consistently high (96%–98%), while error rates remained low (2%–4%) in both conditions (Table 3).

Table 3.

Mean ± SEM of reaction time (RT), percent accuracy, and Taekwondo-specific agility test (TSAT) performance at pre-supplementation (pre-sup), post-supplementation (post-sup), and post-fatigue induction (post-I) in the caffeine (CAF) and placebo (PLA) conditions (n = 13 per condition).

Parameter
	PLA (n = 13)			CAF (n = 13)
	pre-Sup	post-Sup	post-I	pre-Sup	post-Sup	post-I
RT (millisecond)	307 ± 12	325 ± 11	322 ± 16	296 ± 8	297 ± 10	294 ± 9
Accuracy (%)	97 ± 1	96 ± 2	96 ± 1	97 ± 1	98 ± 1	98 ± 0
Error (%)	3 ± 1	4 ± 2	4 ± 1	3 ± 1	2 ± 0	2 ± 0
TSAT (millisecond)	542 ± 50	538 ± 77	524 ± 69	566 ± 76	555 ± 71	538 ± 66

3.3. Taekwondo-specific agility performance (TSAT)

No significant differences in TSAT performance were observed between conditions or across time points (p > 0.05). Agility times ranged from 524 ± 69 ms to 566 ± 76 ms in both CAF and PLA trials, with no significant main effects of time or condition (Table 3).

4. Discussion

This study investigated the effects of a 200 mg dose of caffeine (~3 mg/kg) on physiological responses, EEG brainwave activity, auditory P300 ERPs, and Taekwondo-specific performance following combined mental and physical fatigue. The findings indicate that although caffeine significantly influenced central nervous system (CNS) activity, evidenced by reductions in delta wave power and increases in P300 amplitude, no significant improvements were observed in physiological parameters or Taekwondo-specific performance outcomes.

The study revealed that a single 200 mg dose of caffeine did not significantly influence physiological markers, including HR, RPE, BG, or BL, nor did it enhance Taekwondo-specific performance measures such as reaction time or kick accuracy. These findings align with previous research indicating that low-to-moderate caffeine doses (~0.5–4 mg/kg) provide limited ergogenic benefits in activities emphasizing neuromuscular coordination, precision, and agility [8]. Although studies have shown clear performance benefits from low-dose caffeine supplementation (~3 mg/kg) in endurance-based activities, its effects in sports requiring technical precision and agility, such as soccer and tennis, remain less clear [9,20]. The lack of performance enhancement suggests that caffeine’s ergogenic effects are likely dose-dependent [8], with higher doses (≥5 mg/kg) demonstrating more pronounced benefits in endurance-based tasks by stimulating energy metabolism and reducing perceived exertion [10,39–41].

Furthermore, these results suggest that caffeine’s impact on fine motor execution and neuromuscular coordination may be minimal in elite Taekwondo athletes when administered at low doses. This is consistent with research in other skill-intensive sports such as archery and shooting, where caffeine supplementation has shown minimal or even potentially detrimental effects on motor precision [42,43]. Given that skill-based sports prioritize technique over absolute force production, caffeine’s ergogenic effects may be more evident in endurance or high-intensity power-based sports (e.g. running or swimming) than in technical combat sports like Taekwondo [13,14]. Additionally, the central effects of caffeine may not have been sufficient to counteract fatigue-related declines in reaction time, agility, and accuracy, suggesting that higher doses or individual caffeine sensitivity may play a critical role in performance outcomes.

Despite the lack of measurable improvements in physical performance, caffeine elicited significant neural effects, as evidenced by reductions in delta wave power across key cortical sites (Fpz, Pz, POz, and Oz). These reductions suggest enhanced alertness and decreased cortical drowsiness within 60 minutes of ingestion, reinforcing caffeine’s role as a CNS stimulant. This aligns with the adenosine receptor antagonism hypothesis, which posits that caffeine delays fatigue by blocking adenosine’s inhibitory effects on neuronal excitability and synaptic transmission [6,44]. By doing so, caffeine helps sustain arousal, vigilance, and cognitive readiness during recovery periods [45].

Interestingly, while delta activity was significantly suppressed, theta, alpha, and beta waves remained unchanged, suggesting that caffeine’s neurophysiological effects were selective. Delta waves are strongly associated with fatigue and cortical slowing [46], therefore, their suppression aligns with improved cognitive alertness. Notably, beta waves, which are typically linked to cognitive engagement and mental effort, did not significantly change in the caffeine condition. This contrasts with findings from Siepmann and Kirch [28], who reported reductions in alpha and beta waves following caffeine ingestion. The discrepancy may stem from differences in participant training backgrounds, as elite athletes often exhibit distinct neural adaptations that could influence caffeine’s effects on cortical activity [1,11]. Previous EEG studies in both general and task-related contexts have shown that caffeine primarily affects alpha and beta activity [20,22], suggesting that such effects may vary depending on the cognitive and motor demands of the task, potentially differing between endurance-based and skill-based sports such as Taekwondo.

The auditory P300 ERP findings further support caffeine’s cognitive-enhancing properties, as evidenced by increased P300 amplitudes at the Cz and Pz electrodes. The lack of significant amplitude changes at Fz suggests that caffeine’s effects were more pronounced in midline central and parietal regions, which are key areas associated with attentional allocation and stimulus evaluation during oddball paradigms [47]. These amplitude enhancements indicate improved cognitive processing and attentional resource allocation during the auditory oddball task, aligning with previous research demonstrating caffeine’s role in maintaining vigilance and cognitive efficiency under fatigue-inducing conditions [8,48].

Despite these neurophysiological benefits, caffeine did not significantly enhance reaction time, accuracy, or agility-based Taekwondo performance, suggesting a dissociation between cognitive enhancement and skill-based motor execution. While endurance-based and power-focused sports often benefit from caffeine due to its role in increasing energy availability and reducing perceived exertion [48], skill-based sports such as Taekwondo require precise motor control and complex decision-making, which may be less influenced by low-dose caffeine. Additionally, the lack of changes in P300 latency suggests that, although caffeine enhanced neural responsiveness, this did not translate into measurable improvements in motor execution. Similar observations have been reported in other precision sports, where cognitive enhancements observed through EEG measures did not directly correspond to improved task-specific physical performance [49,50]. These findings indicate that even at a low dose (200 mg), caffeine may primarily enhance cognitive function by modulating neural activity, rather than directly improving motor performance. Future studies should explore whether higher caffeine doses or individualized strategies can optimize both cognitive and motor outcomes in combat sports.

4.1. Limitations and future directions

This study has several limitations. First, only male Taekwondo athletes were included, limiting the generalizability of the findings to female athletes or those in other sports. Second, a single 200 mg caffeine dose was used, which may not fully capture its dose-dependent effects. Future research should explore a range of doses, personalized dosing strategies, and the role of individual caffeine tolerance, including habitual intake and genetic variations. Lastly, longitudinal studies are needed to examine caffeine’s long-term impact on training adaptation, recovery, and competitive performance across different athletic disciplines.

5. Conclusion

This study demonstrates that, although a single 200 mg dose of caffeine did not enhance physiological markers or Taekwondo-specific performance metrics, it exerted significant neurophysiological and cognitive effects. Specifically, caffeine reduced delta wave activity, indicating decreased central fatigue and enhanced alertness, and increased P300 amplitudes, suggesting improved cognitive processing and attentional allocation. However, caffeine did not significantly affect reaction time, accuracy, or agility, reinforcing that low dose primarily benefit cognitive rather than physical performance domains.

These findings suggest that low-dose caffeine supplementation may be more beneficial for cognitive aspects of performance, particularly under fatigue-inducing conditions, rather than for enhancing physical agility or sport-specific motor execution in Taekwondo. Athletes and coaches should consider individualized caffeine strategies tailored to sport-specific demands. Future studies should investigate dose-dependent effects and long-term implications to optimize caffeine’s role in both cognitive and physical performance enhancement in competitive sports.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).