The moderating effect of combat sports discipline on the relationship between emotion regulation and aggression among athletes practicing Brazilian jiu-jitsu, karate, wrestling, and mixed martial arts

Abstract

Background

Findings on aggression in martial arts and combat sports are mixed, and relatively little is known about whether associations between emotion regulation strategies and aggression differ across combat sport disciplines. The present study examined whether discipline moderated the relationships of cognitive reappraisal and expressive suppression with aggression among athletes practicing Brazilian jiu-jitsu (BJJ), karate, wrestling, and mixed martial arts (MMA).

Methods

In this cross-sectional paper-and-pencil study, 277 athletes aged 16–57 years (M = 28.81, SD = 10.12) completed the Buss–Perry Aggression Questionnaire and the Emotion Regulation Questionnaire. Because the verbal aggression subscale showed poor reliability (α = 0.53), it was excluded from the moderation analyses. Moderation models were estimated for total aggression, physical aggression, anger, and hostility, with discipline as the moderator and age, sex, sport experience, and competition level entered as covariates.

Results

Most aggression dimensions were not significantly associated with emotion regulation strategies. In the adjusted moderation models, discipline did not significantly moderate the associations of cognitive reappraisal or expressive suppression with total aggression, physical aggression, or anger. However, discipline significantly moderated the association between cognitive reappraisal and hostility. A positive association was observed in wrestling and MMA, but not in BJJ or karate. Expressive suppression was positively associated with hostility, independent of discipline.

Conclusions

The relationship between emotion regulation and aggression in combat sports is nuanced. Most associations between emotion regulation strategies and aggression were small or non-significant, and the only moderated association concerned cognitive reappraisal and hostility. Rather than viewing aggression as uniformly negative, the observed associations may tentatively be interpreted as reflecting functional, discipline-specific forms of aggression; the present data, however, do not directly measure tactical aggression, sport performance, competitive outcomes, or aggressive behavior. These cross-sectional findings highlight the need for sport-specific psychological support and longitudinal research to understand how different styles shape emotional management.

Background

Martial arts (MA) are characterized as systems that integrate elements of philosophy, tradition, and physical combat with other features, setting them apart from standard physical exercises [1]. Combat sports (CS) are popular worldwide. They are featured in major international competitions such as the Olympic Games and the World Combat Games, which require a high level of physical fitness, technical skill, and tactical acumen [2]. While often used interchangeably in the broader literature, MA and CS possess distinct conceptual frameworks. Traditional MA (e.g., aikido, kendo) is frequently characterized as a comprehensive system integrating philosophical, spiritual, and physical development, often lacking formal organization or competitive structures [3, 4]. Conversely, CS are defined by their competitive nature, standardized rules, and governance by sporting federations, emphasizing sportization and physical dominance [5, 6]. Modern disciplines often blur these lines; however, distinguishing them is crucial for understanding the differing psychological and tactical demands placed on practitioners [3–6]. In the present study, the four analyzed disciplines: Brazilian jiu-jitsu (BJJ), karate, wrestling, and mixed martial arts (MMA), are examined as competitive combat sport disciplines. To avoid conceptual ambiguity, we use the broader term martial arts and combat sports (MA and CS) only when referring to the wider field, and combat sport disciplines when referring to the four groups compared in the analyses.

Wrestling and BJJ emphasize techniques that involve gripping and tossing opponents [1, 2, 7], whereas karate focuses on striking opponents using different body parts, including the fists, elbows, knees, and feet [2]. MMA is a versatile CS that integrates techniques from both grappling and striking disciplines, including boxing, BJJ, and wrestling [7, 8]. Wrestling, Brazilian Jiu-Jitsu, karate, and MMA involve physical contact through sparring and competitive matches, which increases the risk of injury, especially to the head and limbs [9, 10]. Systematic training forms the basis for mastering sport skills, as it generally teaches individuals to control aggressive urges, develop self-defense skills, improve physical performance, and reduce fear reactions [11].

Extensive research has been conducted on the practice of MA and CS to evaluate its effects on aggression and emotion regulation. However, research on aggression in MA and CS has produced mixed findings. Some studies have reported no clear differences in aggression among practitioners [12, 13], whereas others have shown that aggression may vary across disciplines and training contexts [14–16]. Aggression in these settings is also multidimensional: physical aggression, anger, hostility, and verbal aggression do not necessarily show the same pattern of associations. This heterogeneity makes it difficult to infer a single effect of MA and CS participation on aggression, suggesting that both discipline-specific context and individual self-regulatory characteristics may be relevant.

Engaging in traditional MA has been found to reduce anger and aggression levels, especially among adults and young people with behavioral issues [11]. Kids who engage in judo show greater empathy than those who do not, suggesting that martial arts may improve emotional self-regulation and self-awareness [17]. A meta-analysis revealed that engaging in MA or CS training may help decrease externalizing behaviors like aggression, anger, and violence among young people [18]. Moreover, teenagers who engage in martial arts exhibit greater self-discipline and reduced tendencies toward bullying compared to their peers who do not participate in such activities [19–23].

A recent review highlighted a wide range of psychological, social, emotional, physical, cognitive, and academic advantages of Ma and CS interventions for young people [24]. Sport training cultivates self-regulation, allowing young individuals to manage their emotions and conduct themselves effectively. High school students engaged in Ma and CS exhibited improved self-discipline, strategic planning, and emotional awareness [23]. Studies have shown that MA training can enhance cognitive abilities, such as attention and processing speed, and offer psychological benefits, including boosting self-esteem and reducing aggression [25]. Furthermore, such training can foster positive social behaviors and diminish negative ones, leading to improved social interactions and better emotional regulation [19, 26–28].

It is important to note that evidence regarding the possible psychological benefits of MA and CS is mixed. Some studies suggest that training may be associated with better self-control, empathy, or lower externalizing behavior. In contrast, other findings are more equivocal and depend on age, training format, and outcome definition [29–31]. A recent review, therefore, supports a cautious interpretation, emphasizing the heterogeneity of interventions, samples, and methodological quality rather than a uniform protective effect of combat-sport participation [32].

Emotion regulation may be one factor that helps explain these inconsistencies. Gross and John [33] distinguish two primary emotion regulation strategies: cognitive reappraisal and expressive suppression. Cognitive reappraisal involves changing thoughts about emotion-eliciting situations to alter their impact, occurring before emotional responses are generated [34, 35]. Expressive suppression inhibits outward emotional signs after response generation, focusing on controlling expression rather than altering the experience [34, 35]. These strategies differently affect aggression levels, with reappraisal reducing aggression while suppression increases it [36, 37]. In non-sport settings, cognitive reappraisal is usually considered more adaptive than suppression, but whether these associations generalize to competitive combat-sport contexts remains unclear. From a sport-psychology perspective, this question is important because athletes across different combat sports face varying technical demands, rules, and competitive situations. These differences may shape how emotion regulation strategies co-occur with specific forms of aggression. Still, they do not justify assuming that any discipline inherently causes a given emotional or behavioral pattern. The available literature, therefore, supports examining discipline as a potential moderator while maintaining a cautious, non-causal interpretation of observed associations [32].

Research revealed a dual nature of Ma and CS training, both as a means of self-control and as a potential source of aggression, necessitating further exploration of its effects across various populations [38, 39]. Individual differences in emotional regulation can be crucial in understanding aggressive behavior among CS athletes. However, the impact of the CS discipline on the relationships between emotion regulation strategies and aggression is unknown. The present study addressed this gap by testing whether combat-sport discipline moderated the relationships between cognitive reappraisal and expressive suppression and aggression dimensions among athletes practicing BJJ, karate, wrestling, and MMA. Studies indicate that CS training is associated with improved emotional self-control, which contributes to better emotional management and mitigates aggressive behaviors [11, 20, 39]. We assume that differences across techniques and objectives in selected CS styles will yield distinct effects of emotional regulation strategies on specific aggression dimensions.

To our best knowledge, the present study will examine for the first time the moderating effect of CS discipline (including karate, BJJ, wrestling, and MMA) on the association between emotional regulation strategies (i.e., cognitive reappraisal and expressive suppression strategies) and aggression (including the general aggression index, and its dimensions: physical and verbal aggression, anger, and hostility dimensions). By focusing on improving emotion regulation skills and understanding the sports contexts of aggression, interventions can be specifically tailored to reduce aggressive tendencies in individuals.

Based on previous research, we hypothesize that: (H1) cognitive reappraisal is related to lower aggression levels while expressive suppression correlates with high aggression among CS athletes; (H2) the strength of these associations differs depending on aggression dimension (physical and verbal aggression, anger, and hostility) and CS discipline (karate, BJJ, wrestling, and MMA). Accordingly, the study tested whether discipline moderated the relationships between the two emotion regulation strategies and the aggression outcomes examined here.

Methods

Study design and procedure

Cross-sectional studies were conducted among CS athletes from the Opole and Podkarpackie provinces, between February and September 2024. It is posited that various patterns of sports training may affect athletic performance and aggression management. All procedures followed the Declaration of Helsinki, and the research protocol was approved by the Research Ethics Committee at the Opole University of Technology (Decision No. 2/2022). Written informed consent was obtained from all participants. To ensure that only CS athletes participated in the study, the following inclusion criteria were established: (1) membership in a sports club of one of four disciplines: BJJ, Karate Kyokushin, wrestling, or MMA; (2) active participation in training for a minimum of six months; (3) minimum age of 16 years. The following clubs were involved in the study: Forca Opole (BJJ, MMA, and Wrestling), Forca Nysa (BJJ), Heel Hoo Namysłow (BJJ and Wrestling), Next Level (BJJ and MMA), Orzeł Namysłów (Wrestling), Club Karate Kyokushin Sandomierz, and Club “Wisłok” Rzeszów (Boxing and MMA).

The exclusion criterion was the lack of informed consent to the study. The reason these combat sports disciplines were chosen is that they share similarities and differences in fighting technique, which determines how emotions are managed and the risks to physical health. Wrestling and BJJ primarily focus on grappling techniques that involve gripping, controlling, and throwing opponents, emphasizing ground control and submission holds. In contrast, karate centers on striking techniques, using various body parts such as fists, elbows, knees, and feet to attack opponents. MMA is a comprehensive combat sport that combines elements of grappling and striking, integrating techniques from boxing, BJJ, wrestling, and other martial arts to create a versatile fighting style. All four CSs involve close physical contact through sparring and competitive matches, which significantly increases the risk of injuries, particularly to the head and limbs. The high-contact nature of these sports demands systematic training, which is essential not only for skill acquisition but also for teaching athletes to manage aggressive impulses, develop effective self-defense techniques, enhance physical conditioning, and reduce fear responses during competition. This specificity highlights the distinct technical emphases and injury risks inherent in each sport while underscoring the shared necessity of structured training for performance and safety.

The sample size was determined a priori using G*Power software [40]. Considering a moderate effect size for linear regression (ƒ² = 0.15), the minimal value of p = 0.05, and power 0.95, the expected sample size was N = 107. For Pearson’s correlation, assuming a moderate effect size (ρ = 0.30), a p-value of 0.05, and a power of 0.95, the expected sample size was N = 111. The invitation to the study was accepted by 285 CS athletes. However, three people refused to participate, and five people were rejected from statistical tests because they did not meet the inclusion criteria (age under 16). The final sample included 277 participants, which indicates more than adequate power.

Demographic characteristics of the sample

The total sample consisted of 277 participants aged 16 to 57 years (M = 28.81, SD = 10.12). The majority of respondents were men with higher levels of education and good self-reported economic status (Table 1). Among CS competitors, the majority practiced BJJ (n = 107; 38.63% of the total sample), followed by wrestling (n = 78; 28.16%), MMA (n = 50; 18.05%), and karate (n = 42; 15.16%). Participants’ sports experience ranged from 0.5 to 41 years, with an average of 8 years (M = 7.65, SD = 5.68). On average, athletes had 7 years of experience in the sport, with the majority competing at the national level.

Table 1.

Sample characteristic (N = 277)

Variable	Categories	Range/ coding	Total sample (n = 277)		BJJ (n = 107)		Karate (n = 42)		Wrestling (n = 78)		MMA (n = 50)
			n	%	n	%	n	%	n	%	n	%
Age	(M/SD)	16–57	28.81	10.12	30.30	9.41	29.33	14.60	28.49	8.92	25.66	8.14
Gender	Women	0	71	25.63	44	41.12	17	40.48	4	5.13	6	12.00
	Men	1	206	74.37	63	58.88	25	59.52	74	94.87	44	88.00
Education	Primary	0	54	19.50	12	11.22	14	33.33	20	25.64	8	16.00
	Vocational	1	16	5.78	4	3.74	1	2.38	7	8.97	4	8.00
	Secondary	2	99	35.74	35	32.71	10	23.81	31	39.74	23	46.00
	High	3	108	38.99	56	52.34	17	40.48	20	25.64	15	30.00
Subjective assessment of economic status	Disappointing	0	11	3.97	5	4.67	3	7.14	1	1.28	2	4.00
	Average	1	26	9.39	8	7.48	12	28.57	2	2.56	4	8.00
	Good	2	136	49.10	49	45.79	12	28.57	47	60.26	28	56.00
	Very good	3	82	29.60	35	32.71	10	23.81	22	28.21	15	30.00
	Excellent	4	22	7.94	10	9.35	5	11.91	6	7.69	1	2.00
Competition range	Regional	1	87	31.41	26	24.30	24	57.14	9	11.54	28	56.00
	National	2	146	52.71	58	54.21	15	35.71	53	67.95	20	40.00
	European	3	34	12.27	20	18.69	2	4.76	10	12.82	2	4.00
	World	4	10	3.61	3	2.80	1	2.38	6	7.69	0	0.00
Sports experience (years)	(M/SD)	0.5–41	7.65	5.68	7.44	5.26	9.08	7.92	8.93	4.95	4.92	4.34

Measures

The Buss and Perry Aggression Questionnaire (BPAQ) is a self-assessment tool comprising 29 items designed to evaluate facets of aggression (Buss & Perry, 1992). It includes four distinct elements: verbal aggression (VA, 5 items), physical aggression (PA, 9 items), hostility (H, 8 items), and anger (A, 7 items). Participants rate how accurately a statement reflects their characteristics on a 5-point Likert scale, where 1 indicates “Extremely Uncharacteristic,” and 5 indicates “Extremely Characteristic.” Scores between 29 and 149 suggest stronger agreement with aggressive statements, indicating a higher likelihood of aggressive thoughts, emotions, and actions. The Polish validation [41] showed adequate reliability. In this study, the overall BPAQ score demonstrated acceptable reliability (Cronbach’s α = 0.89). The Cronbach’s α for the subscales PA, VA, A, and H were 0.73, 0.53, 0.75, and 0.79, respectively. Due to an unacceptable level of reliability, the verbal aggression scale was entirely excluded from further statistical analyses to prevent interpretative errors.

The Emotion Regulation Questionnaire (ERQ), developed by Gross and John [33], is a widely used tool for assessing two primary emotion regulation strategies: cognitive reappraisal (CR; 6 items) and expressive suppression (ES; 4 items). Participants respond to each of 10 questions using a 7-point Likert scale, with options ranging from 1 (strongly disagree) to 7 (strongly agree). The ERQ has been validated across various populations and languages, including the Polish version [42], providing insights into its psychometric properties and applicability. Reliability in the present study was Cronbach’s α = 0.87 for the CR scale and α = 0.69 for the ES scale.

The socio-demographic inquiries focused on age (in completed years), gender (Men, Women, Other), educational attainment (Primary, Vocational, Secondary, Undergraduate studies, Graduate studies, Master’s degree or higher), and a self-evaluated perception of economic status (Disappointing, Average, Good, Very good, Excellent). For CS athletes, the questions pertained to their sports discipline (BJJ, Karate, Wrestling, MMA), the highest level of championship achieved (None, Regional, National, European, World), the average number of minutes spent on physical activity each week, and their years of experience in CS sports.

Statistical analysis

The distributional properties of the BPAQ and ERQ scales were described using means, standard deviations, skewness, and kurtosis. Given the sample size (N = 277) and acceptable skewness and kurtosis values, we used parametric analyses. Pearson correlations were calculated to examine bivariate associations between aggression dimensions and emotion regulation strategies. Verbal aggression was excluded from the moderation analyses because of poor internal consistency (α = 0.53). Before hypothesis testing, assumptions of normality, homoscedasticity, and multicollinearity were assessed. Kolmogorov-Smirnov tests, visual inspection of residual plots, Variance Inflation Factor (VIF < 3), and tolerance values (> 0.2) indicated that assumptions for regression analyses were met. Both predictors (cognitive reappraisal and expressive suppression) were mean-centered before computing interaction terms to facilitate substantive interpretation. The moderating effect of combat sports (BJJ, karate, wrestling, and MMA) on the relationship between emotion regulation strategies and aggression outcomes was tested using Hayes’s PROCESS macro (Model 1) in SPSS. To ensure robust inferential criteria, 95% bootstrap confidence intervals based on 5,000 resamples were used as the primary criterion for statistical significance, rather than relying solely on p-values. Age, sex, sport experience, and competition level were included as covariates in each regression model. Statistical significance was evaluated at p < 0.05, and bootstrap intervals are reported as an additional estimate of precision.

Results

Associations between emotion regulation and aggression dimensions

Initially, the parametric properties of the variables were examined, including total aggression score (BPAQ), physical aggression, anger, and hostility subscales, and two emotion regulation strategies: cognitive reappraisal and expressive suppression (Table 2). Because the verbal aggression subscale showed poor reliability, it was not entered into correlation and regression analyses. Since the sample was quite large (N > 200) and skewness and kurtosis were within ± 1, we assumed that deviations from normality were not a serious problem and proceeded with parametric tests. Pearson’s correlation analysis showed that the BPAQ scales correlated positively and moderately with one another. The aggression dimensions were positively intercorrelated. Regarding emotion regulation, the bivariate analyses revealed only two statistically significant associations: cognitive reappraisal was weakly and positively associated with physical aggression, and expressive suppression was positively associated with hostility. Therefore, we conclude that the correlation analysis did not support the hypothesis H1.

Table 2.

Descriptive statistics (N = 277)

Variable	1.	2.	3.	4.	5.	6.
1. Aggression
2. Physical aggression	0.78***
3. Anger	0.86***	0.52***
4. Hostility	0.83***	0.45***	0.65***
5. Cognitive reappraisal	0.10	0.14*	0.07	0.05
6. Expressive suppression	0.11	0.08	0.03	0.20***	0.14
Minimum	30	9	7	8	1	1
Maximum	126	39	33	40	7	7
Mean	72.83	21.73	16.53	19.96	4.90	3.83
Standard Deviation	17.35	5.95	5.55	6.60	1.25	1.29
Skewness	0.41	0.58	0.29	0.29	-0.75	-0.04
Kurtosis	-0.16	0.04	-0.58	-0.49	0.90	-0.31

* p < 0.05, *** p < 0.001

Moderation analyses

To test the hypothesis that the CS discipline (coded as BJJ = 1, Karate = 2, Wrestling = 3, MMA = 4, with BJJ as reference) moderates the relationship between both emotion regulations strategies (cognitive reappraisal and expressive suppression) and aggression scales (total score, physical aggression, anger and hostility), a moderation analysis was conducted using Hayes’s PROCESS macro (Model 1), with 5,000 bootstrap samples for 95% percentile confidence intervals. Age, sex, sports experience, and competition level were included as covariates in the model (N = 277).

Moderating effect of cognitive reappraisal on aggression across combat sports disciplines

Assumption checks for the linear regression indicated that linearity was supported by the overall model fit, independence of residuals by the Durbin–Watson test (DW = 1.971, p = 0.720), homoscedasticity by the Breusch–Pagan (χ² = 8.20, p = 0.609), and normality of residuals by the Kolmogorov–Smirnov test (D = 0.069, p = 0.144). Multicollinearity was not a concern with all VIF values below 1.17 and tolerance values above 0.85. The overall regression model, testing the moderating effect of sport discipline on the relationship between cognitive reappraisal and total aggression score, accounted for a significant 13% of the variance in aggression, R² = 0.13, F(11, 265) = 3.43, p < 0.001. However, among covariates, only younger age was a significant predictor of higher aggression levels, b = − 0.22, SE = 0.11, t(265) = − 2.05, p = 0.041, 95% CI [− 0.43, − 0.01]. Sex, sports experience, and competition level were not significant. The omnibus test of the highest-order interaction (X × W) was not statistically significant, ΔR² = 0.023, F(3, 265) = 2.29, p = 0.079, indicating that the conditional effect of cognitive reappraisal on aggression did not differ significantly across sports disciplines at the conventional α = 0.05 threshold.

A moderation analysis was conducted to examine whether the CS discipline moderated the relationship between cognitive reappraisal and physical aggression. Before interpretation, the assumptions of linear regression were examined. Linearity was confirmed by inspection of residual plots, independence of residuals was supported by the Durbin–Watson test (DW = 1.963, p = 0.684), homoscedasticity was confirmed by the Breusch–Pagan test (BP = 5.222, p = 0.876), residuals deviated slightly from normality according to the Kolmogorov–Smirnov test (D = 0.087, p = 0.031) though this is acceptable given the sample size (N = 277) and the robustness of OLS to mild non-normality, and multicollinearity was not a concern as all predictors showed VIF values below 1.17 and tolerance values above 0.86. The overall model was statistically significant, accounting for 13% of the variance in physical aggression, R² = 0.13, F(11, 265) = 3.66, p < 0.001. Among the covariates, only sex was a significant predictor of physical aggression, with males reporting higher physical aggression scores (b = 3.08, SE = 0.87, t(265) = 3.55, p < 0.001, 95% CI [1.37, 4.79]). The main effect of cognitive reappraisal on physical aggression was non-significant, b = 0.08, SE = 0.08, t(265) = 1.04, p = 0.300, 95% CI [-0.07, 0.23]. The omnibus test of the highest-order interaction showed that sport discipline did not significantly moderate the relationship between cognitive reappraisal and physical aggression, ΔR² = 0.007, F(3, 265) = 0.75, p = 0.525.

A moderation analysis was performed to examine whether the CS discipline moderated the relationship between cognitive reappraisal and anger. Prior to interpretation, regression assumptions were verified and satisfied: linearity was confirmed by inspection of partial regression plots; residuals were independent based on the Durbin–Watson statistic (DW = 2.047, p = 0.824); homoscedasticity was supported by the Breusch–Pagan test (χ² = 13.355, p = 0.204); residuals were normally distributed according to the Kolmogorov–Smirnov test (D = 0.038, p = 0.823); and multicollinearity was not a concern, as all VIF values were below 1.20 and Tolerance values exceeded 0.85. The overall model was statistically significant, R² = 0.12, F(11, 265) = 3.41, p < 0.001. The predictors accounted for 12% of the variance in anger. The main effect of cognitive reappraisal on anger was non-significant, b = 0.03, SE = 0.07, t(265) = 0.48, p = 0.630, 95% CI [− 0.11, 0.17]. Among the covariates, sex was the only significant predictor, with females reporting higher anger scores than males (b = − 2.29, SE = 0.81, t(265) = − 2.82, p = 0.005, 95% CI [− 3.89, − 0.69]). None of the interaction terms reached significance level. The omnibus test of the highest-order interaction confirmed that sport discipline did not significantly moderate the association between cognitive reappraisal and anger, ΔR² = 0.014, F(3, 265) = 1.44, p = 0.230.

A moderation analysis examined the moderating effect of the CS discipline on the relationship between cognitive reappraisal and hostility (Table 3). The assumptions of linear regression were verified and met, as suggested inspection of residual plots, no autocorrelation (DW = 1.928, p = 0.448), homoscedasticity by Breusch–Pagan test (χ² = 11.976, p = 0.287), normality assumption by the Kolmogorov–Smirnov test (D = 0.057, p = 0.321), and collinearity statistics, including all VIF values below 1.17 and tolerance values above 0.85. The overall model was statistically significant and accounted for 16% of the hostility variance, R² = 0.16, F(11, 265) = 4.71, p < 0.001. However, the main effect of cognitive reappraisal on hostility was non-significant. Among the covariates, younger age and female sex were significant predictors of hostility, whereas sports experience and competition level were not. Among the interaction terms of sport discipline with cognitive reappraisal, only wrestling and MMA were statistically significant. The omnibus test of the highest-order interaction confirmed that sport discipline significantly moderated the relationship between cognitive reappraisal and hostility, ΔR² = 0.038, F(3, 265) = 4.06, p = 0.008. Examination of conditional effects revealed that cognitive reappraisal was significantly and positively associated with hostility in wrestling and MMA, but not in BJJ or karate (Table 3). Figure 1 presents the regression effect of cognitive reappraisal on hostility across combat sports disciplines.

Table 3.

Linear multiple regression model of cognitive reappraisal effect on hostility within each combat sport discipline (N = 277)

Variable	B	SE	t	p	95% CI
					LL	UL
constant	24.88	1.73	14.36	0.000	21.47	28.29
Cognitive reappraisal (CR)	-0.06	0.08	-0.75	0.456	-0.22	0.10
Age	-0.13	0.04	-3.35	0.001	-0.21	-0.05
Sex	-2.32	0.95	-2.46	0.015	-4.18	-0.46
Sport experience	0.08	0.08	1.08	0.281	-0.07	0.23
Competition level	-0.68	0.56	-1.22	0.223	-1.79	0.42
W1 Karate (BJJ ref.)	3.10	1.18	2.64	0.009	0.79	5.42
W2 Wrestling (BJJ ref.)	0.58	0.99	0.59	0.559	-1.37	2.53
W3 MMA (BJJ ref.)	2.96	1.15	2.57	0.011	0.69	5.22
CR x W1	-0.14	0.16	-0.83	0.405	-0.46	0.19
CR x W2	0.27	0.12	2.24	0.026	0.03	0.51
CR x W3	0.36	0.14	2.47	0.014	0.07	0.64
Conditional effects
Brazilian jiu-jitsu (BJJ)	-0.06	0.08	-0.75	0.456	-0.22	0.10
Karate	-0.20	0.14	-1.39	0.164	-0.48	0.08
Wrestling	0.21	0.09	2.33	0.021	0.03	0.39
Mixed martial arts (MMA)	0.30	0.12	2.50	0.013	0.06	0.53

Discipline coding was 1 = Brazilian jiu-jitsu (BJJ), 2 = karate, 3 = wrestling, and 4 = Mixed martial arts (MMA). Positive conditional effects indicate that higher cognitive reappraisal scores were associated with higher hostility within that discipline. The conditional effect was statistically significant for wrestling and MMA, but not for BJJ or karate

CI confidence interval, LL lower level, UL upper level

Fig. 1.

Scatter plot displaying the linear relationship between cognitive reappraisal and hostility across four combat sports: Brazilian jiu-jitsu (BJJ), karate, wrestling, and mixed martial arts (MMA). Note. Regression lines indicate the modeled linear trend for each group, and the shaded regions represent the 95% standard error confidence interval. The conditional unstandardized effects of cognitive reappraisal on hostility for each discipline are as follows for BJJ (b = -0.06, p = 0.455), Karate (b = -0.202, p = 0.154), wrestling (b = 0.209, p = 0.020), and MMA (b = 0.298, p = 0.012). Significant slopes (p < 0.05) are denoted by an asterisk

Moderating effect of expressive suppression on aggression across combat sports disciplines

A second series of moderation analyses examined whether sport discipline moderated the relationship between expressive suppression and aggression outcomes, controlling for age, sex, sports experience, and competition level (N = 277). In the first model, the outcome was the total aggression score. Before interpreting the linear regression model, the assumptions were verified and met. Linearity of the predictor–outcome relationships was confirmed by inspection of residual plots, no autocorrelation of residuals was indicated by the Durbin–Watson statistic (DW = 1.985, p = 0.800), homoscedasticity was supported by the Breusch–Pagan test (χ² = 11.733, p = 0.303), normality of residuals was confirmed by the Kolmogorov–Smirnov test (D = 0.070, p = 0.133), and no multicollinearity among predictors was observed, with all VIF values ≤ 1.16 and Tolerance values ≥ 0.86. The overall model was statistically significant, explaining 11% of aggression, R² = 0.11, F(11, 265) = 3.05, p = 0.001. However, the main effect of suppression on total aggression was non-significant, b = 0.46, SE = 0.32, t(265) = 1.42, p = 0.158, 95% CI [-0.18, 1.09]. Among the covariates, only younger age was a significant predictor of higher total aggression, b = -0.24, SE = 0.11, t(265) = -2.22, p = 0.027, 95% CI [-0.44, -0.03], whereas sex, sports experience, and competition level were not significant. None of the interaction terms was statistically significant. The omnibus test of the highest-order interaction confirmed that sport discipline did not significantly moderate the suppression-aggression relationship, ΔR² = 0.011, F(3, 265) = 1.43, p = 0.332.

The moderating effect of sport discipline on the relationship between expressive suppression and physical aggression was statistically significant, R² = 0.12, F(11, 265) = 3.31, p < 0.001. The assumptions for linear regression were verified and met. Linearity was supported by inspection of residual plots; residuals were independent, with no autocorrelation detected by the Durbin–Watson test (DW = 2.011, p = 0.910); the Breusch-Pagan test confirmed homoscedasticity (χ² = 9.079, p = 0.525); the Kolmogorov–Smirnov test indicated a minor departure from normality of residuals (D = 0.087, p = 0.030), which is acceptable given the sample size (N = 277) and the robustness of OLS to mild non-normality; and multicollinearity was not a concern, with all VIF values ≤ 1.16 and Tolerance values ≥ 0.86. The predictors accounted for 12% of the variance in physical aggression. The main effect of expressive suppression on physical aggression was not significant, b = 0.11, SE = 0.11, t(265) = 0.95, p = 0.341, 95% CI [-0.11, 0.32]. Among the control variables, male sex was a significant predictor of physical aggression, b = 3.24, SE = 0.87, t(265) = 3.72, p < 0.001, 95% CI [1.52, 4.96]. Age, years of sport experience, and competition level did not make a significant contribution to the model. The overall highest-order interaction test did not reach statistical significance, ΔR² = 0.006, F(3, 265) = 0.60, p = 0.619.

The overall model predicting anger was statistically significant and explained 12% of the variance in anger, R² = 0.12, F(11, 265) = 3.34, p < 0.001. Assumptions for linear regression were met, as suggested inspection of residual plots, Durbin–Watson test (DW = 2.054, p = 0.748), Breusch–Pagan homoscedastic statistic (χ² = 15.081, p = 0.129), the Kolmogorov–Smirnov test for normality distribution (D = 0.047, p = 0.577), and no multicollinearity statistics (all VIF ≤ 1.16, Tolerance ≥ 0.86). Expressive suppression was not a significant predictor of anger, b = 0.05, SE = 0.10, t(265) = 0.49, p = 0.627, 95% CI [− 0.15, 0.25]. Among the covariates, only female sex was significantly related to anger, b = − 2.26, SE = 0.81, t(265) = − 2.78, p = 0.006, 95% CI [− 3.86, − 0.66], whereas age, sport experience, and competition level did not reach significance. Contrary to expectations, the highest-order interaction term between expressive suppression and sport discipline did not significantly improve the prediction of anger, ΔR² = 0.020, F(3, 265) = 2.01, p = 0.113. Thus, sport discipline did not significantly moderate the association between expressive suppression and anger.

The overall regression model predicting hostility was also statistically significant, accounting for 17% of the hostility variance (R² = 0.17; F(11, 265) = 4.93, p < 0.001). Assumptions for the multiple linear regression were examined and largely met, since linearity was confirmed by inspection of residual scatterplots; residuals were independent, as indicated by the Durbin–Watson statistic (DW = 1.902, p = 0.334); homoscedasticity was borderline, with the Breusch–Pagan test marginally significant (BP = 18.747, p = 0.044) while the Goldfeld–Quandt (p = 0.404) and Harrison–McCabe (p = 0.375) tests supported equal error variances; residuals were approximately normally distributed according to the Kolmogorov–Smirnov test (D = 0.047, p = 0.577); and no multicollinearity was detected, with all VIF values ≤ 1.16 and Tolerance ≥ 0.86. Expressive suppression emerged as a significant positive predictor of hostility, b = 0.39, SE = 0.12, t(265) = 3.24, p = 0.001, 95% CI [0.15, 0.62], indicating that athletes who more frequently used expressive suppression reported higher hostility. Among the covariates, both younger age, b = − 0.15, SE = 0.04, t(265) = − 3.94, p < 0.001, 95% CI [− 0.23, − 0.08], and female sex, b = − 2.25, SE = 0.94, t(265) = − 2.40, p = 0.017, 95% CI [− 4.10, − 0.41], were significantly related to hostility, whereas sport experience and competition level were unrelated. The interaction between expressive suppression and sport discipline did not significantly contribute to predicting hostility, ΔR² = 0.013, F(3, 265) = 1.38, p = 0.248. Inspection of the individual product terms revealed no interaction between sport discipline and expressive suppression. Sport discipline, therefore, did not significantly moderate the association between expressive suppression and hostility.

Discussion

The present study aimed to examine associations between emotion regulation strategies and aggression and the moderating effect of combat sports discipline on this relationship. After adjusting for critical covariates, the key findings indicate that most aggression dimensions were not significantly associated with emotion regulation strategies. However, cognitive reappraisal was found to positively moderate hostility specifically within wrestling and MMA groups. Furthermore, expressive suppression emerged as a significant positive predictor of hostility across all disciplines, though the specific combat sport did not moderate this relationship. These results highlight that evidence for moderation is limited and must be interpreted cautiously, without drawing causal links, especially given the cross-sectional nature of this study.

The scientific literature shows that participation in CS can be associated with mental health benefits, such as improved self-control and reduced feelings of hostility [43]. This seemingly paradoxical outcome, where a combat-focused sport leads to decreased aggression, underscores the socialization processes in martial arts training that emphasize discipline and respect [44]. However, these benefits may differ between individuals who engage in CS recreationally and those who compete. Additionally, specific CS disciplines, such as BJJ, tend to exhibit lower aggression levels than more aggressive styles, such as MMA. This study examined the associations between emotional regulation strategies and aggression among athletes representing four CS styles: BJJ, karate, wrestling, and MMA. In line with previous studies [36, 37], we assumed that higher cognitive reappraisal scores would be associated with lower aggression levels. In contrast, expressive suppression will be positively associated with aggression among CS athletes.

Although some previous research has linked participation in martial arts and combat sports with better self-control and, in some contexts, lower hostility or aggression, the evidence is mixed across disciplines, competitive settings, and outcome definitions [11–16, 43–46]. Against this background, the present study examined whether combat-sport discipline moderated the associations between cognitive reappraisal and expressive suppression and aggression among athletes practicing BJJ, karate, wrestling, and MMA. The results indicate a more limited pattern than hypothesized. Most associations between emotion regulation strategies and aggression outcomes were small or non-significant, H1 was not supported overall, and H2 was supported only partially. CS discipline moderated the association between cognitive reappraisal and hostility, whereas no moderation was found for total aggression, physical aggression, or anger. In addition, expressive suppression showed a positive association with hostility across disciplines. The moderating effect of CS discipline on the relationship between the expressive suppression and aggression scales was not supported.

Our findings align with the broader literature that emphasizes heterogeneity in combat sports outcomes. Lee et al. [32], in their systematic review of Olympic combat sports and mental health, highlighted significant methodological limitations and heterogeneity in how combat sports affect emotional regulation and aggression. Their work underscores the need for cautious interpretation. While combat sports can foster self-regulation, the cross-sectional associations observed here do not imply that disciplines intentionally ‘promote’ distinct strategies or that athletes ‘strategically increase’ hostility. Rather, the psychological demands of different rulesets may attract or cultivate varying emotional profiles.

Emotion regulation and aggression in combat sport contexts

Contrary to hypothesis H1, findings suggest that cognitive reappraisal is weakly and positively related to physical aggression, while expressive suppression is positively related only to hostility. The links between emotion regulation and aggression in CS athletes are negligible and not uniform across aggression dimensions. This interpretation is consistent with work showing that aggression in sport is multidimensional and that not all aggressive tendencies have the same meaning or function in competitive settings [15, 29, 30, 37]. Emotion regulation strategies in martial arts and combat sports are crucial for athletes to manage stress and enhance performance. Emotion regulation involves cognitive strategies, training effects, and psychological interventions. Gross’s [47] emotion regulation theory and process model provides a framework for understanding how individuals manage emotions, emphasizing regulation as a temporal process in which goals influence emotion trajectories [33, 47–49]. The extended model identifies mechanisms behind emotion generation and regulation through hierarchical valuation systems, helping distinguish between generation and regulation while integrating cognitive reappraisal [50, 51].

Adaptive strategies such as cognitive reappraisal and problem-solving lead to better psychological outcomes, whereas maladaptive strategies, such as suppression, increase psychological distress [52, 53]. People often use multiple strategies rather than a single method [54], with effectiveness varying by context and individual differences [52]. Maladaptive emotion regulation negatively impacts resilience and self-compassion, while adaptive strategies help mitigate negative emotional responses in MA and CS practitioners [55]. Positive emotions help manage pre-competition anxiety and correlate with control and stress coping ability [56]. Research showed that martial artists with a history of concussions have different emotion regulation strategies, employing more perspective-taking and perceiving stimuli more positively [57]. MMA athletes demonstrate better emotional regulation than non-athletes [58], and CS training generally improves control over negative emotions [11, 45].

The lack of a strong association between emotional regulation strategies and aggression among CS athletes can be explained by the ambiguous effect of MA and CS training on aggression [11–13]. Some studies have indicated that training in MA and CS enhances self-control and reduces aggression [11, 18, 43], whereas others showed no significant relationship between martial arts practice and aggression levels [11, 45]. Some findings indicate that such training, in particular in MMA, may potentially promote violence [14, 19]. The training hypothesis suggests that traditional martial arts’ focus on self-control and self-esteem may reduce aggression [59]. Studies show that MA and CS training improve cognitive functions and self-regulation, thereby decreasing aggression [17, 19–23, 25]. MA and CS enhance prosocial behaviors and emotional management, as suggested by previous studies [19, 26–28].

To contextualize the positive association between cognitive reappraisal and hostility in certain disciplines, it is essential to distinguish between functional (competitive) aggression and hostile aggression. As noted by Abrams [29], aggression in sports can serve a tactical purpose when appropriately channeled. As Abrams [29] argues, anger, aggression, and violence should not be treated as interchangeable constructs in sport. Yazici and Bostanci [30] similarly highlighted the fine line between functional aggression and dysfunction, demonstrating that competitive aggression can be a regulated state rather than merely an uncontrollable outburst. For the present study, this distinction is important because the BPAQ hostility scale reflects a cognitive-affective disposition marked by mistrust, resentment, and negative evaluations of others. In contrast, physical aggression in combat sports may partly overlap with rule-governed competitive assertiveness. Accordingly, the present findings should not be read as evidence that all forms of aggression are either maladaptive or functional in the same way. Thus, practitioners in high-contact sports like MMA and wrestling might associate cognitive reappraisal with higher hostility scores not because they lack emotional control, but because the sport demands a regulated, functional assertiveness that overlaps with hostility metrics on self-report scales.

Several factors were considered in previous studies to explain the inconsistency, including the type of MA and CS training style, and individual characteristics such as age and gender, as well as pre-existing aggression levels [13, 19]. The present study, however, showed that cognitive reappraisal can also interact with wrestling and MMA to increase hostility among athletes. Since emotion regulation is a dynamic process, future studies should examine the effects of using emotion regulation strategies during competition to test their mutual effects on aggression and sport achievement in particular disciplines of CS.

Cognitive reappraisal and aggression across sport disciplines

Contrary to previous studies [36, 37], the positive bivariate association between cognitive reappraisal and physical aggression was weak and did not remain significant in the adjusted moderation model. The theoretical inverse association between reappraisal and physical aggression was not supported, even after accounting for discipline and covariates such as age, sex, sport experience, and competitive level. Rather than indicating that reappraisal increases aggressive behavior, this pattern may reflect the specific meaning of physical aggression in combat sports, where forceful but rule-constrained actions are integral to performance [29, 60, 61]. Athletes who report greater use of reappraisal may be cognitively reframing competitive stress, threat, or physical confrontation without necessarily becoming less behaviorally forceful in sanctioned sport situations [33, 47–51]. At the same time, because the adjusted effect was non-significant, this interpretation must remain cautious and should not be overstated.

Cognitive reappraisal (CR) involves changing interpretations of situations to alter emotional impact [34, 35, 37]. In CS, higher physical aggression has been associated with competitive success in prior research; the present cross-sectional data, however, do not indicate that athletes use cognitive reappraisal to enhance physical aggression, and such an interpretation would require longitudinal or behavioral evidence. Wei [45] showed that martial arts enhance conflict-resolution skills and emotional control, though not empathy or anger-cause identification. Kruszewski and Walczak [62] noted that MA and CS training increases athletes’ aggressiveness, emphasizing aggression as both physical and psychological in practice. Martial arts training improves cognitive functions like attention and processing speed [25]. It remains an open question, not testable with the present cross-sectional data, whether better cognitive control is associated with competitive aggression during CS fights.

The only significant moderated association concerned hostility. Higher cognitive reappraisal was associated with higher hostility in wrestling and MMA, but not in BJJ or karate. This result contradicts the common assumption that reappraisal is uniformly protective against aggression [36, 37, 48, 49, 63, 64]. A cautious explanation is that, in high-contact disciplines such as wrestling and MMA, reappraisal may sometimes be used to reinterpret provocation, pressure, or antagonistic cues in ways that help athletes stay engaged in competition. Yet, this process may co-occur with a more hostile cognitive stance rather than reduce it. Importantly, the present data do not show that athletes strategically use hostility, nor do they justify causal claims that reappraisal increases hostility. The finding indicates only that these variables were positively associated in this cross-sectional sample of wrestlers and MMA fighters. Future research should test whether this pattern reflects discipline-specific competitive norms, threat appraisal, or unmeasured characteristics such as trait competitiveness, impulsivity, or coaching climate [29, 30, 65, 66].

Research indicates that social context and individual traits affect how serotonin dysfunction relates to aggression [65]. Aggressive individuals often struggle with impulse control, emotional management, and social connections. Individual differences, including gender and emotional awareness, influence aggression [66]. Due to the limited number of female participants, gender effects were not examined directly in this study. However, younger age and female sex were significant predictors of hostility in the moderation model. Future research should consider impulsiveness, age, and gender using a more diverse sample of CS practitioners.

Expressive suppression and aggression

Consistent with hypothesis H1, expressive suppression was only positively related to hostility across all disciplines. However, contrary to H1, its associations with total aggression, physical aggression, and anger were non-significant. This result is more consistent with the broader emotion-regulation literature, which shows that suppression often limits emotional expression without reducing the underlying emotional experience and may therefore co-occur with greater negative affect, distress, or maladaptive interpersonal processing [33, 36, 37, 47, 48, 63, 67–70]. Research shows that negative affect mediates between expressive suppression and aggression, as expressive suppression may increase negative feelings and aggression [36]. Poor emotional control, especially negative urgency, correlates with aggression forms [66], while negative emotional states can motivate aggressive behavior as a mood-improvement strategy [71].

In combat-sport environments, suppressing visible emotional reactions may be socially reinforced, as athletes are expected to maintain composure and avoid showing vulnerability. However, the present findings suggest that frequent reliance on suppression may be linked specifically to the cognitive-affective component of aggression, namely hostility, rather than to overt aggression more broadly.

Previous studies suggested that expressive suppression (ES) can lead to higher aggression, particularly regarding hostility [33, 47, 48]. Hostility, an aggression dimension in BPAQ, is a cognitive component characterized by feelings of ill will and injustice, leading to aggressive thoughts. Hostility involves negative beliefs towards others and correlates with anger and physical aggression [72, 73]. ES involves inhibiting emotional expression without reducing the emotional experience itself [63, 64]. ES acts late in emotion-generative processes, focusing on inhibiting outward signs rather than altering emotions [47, 67]. Unlike cognitive reappraisal, suppression only reduces behavioral expression while potentially impairing memory and increasing physiological responses [47, 63, 67]. ES is common in social contexts when avoiding conflict [68]. MA&CS athletes must suppress emotions during sports fights to prevent opponents from exploiting weaknesses. However, frequent expressive suppression has health consequences, associated with increased negative emotions and distress symptoms [69, 70, 74], particularly affecting adolescents’ emotional, social, and physiological outcomes [75].

Controlled aggression may enhance athletic performance, but excessive aggression can cause injuries through uncontrolled violence [60]. Some accounts suggest that CS athletes may rely less on expressive suppression and more on cognitive reappraisal when managing physical aggression, although the present cross-sectional data do not permit conclusions about sporting success. While aggression is inherent in combat sports, it can be managed through training and mental conditioning [61]. Martial arts training can reduce hostility, as practitioners learn to channel aggression into assertiveness [60]. Sparring provides an outlet for aggressive impulses in a controlled environment [76], though this effect may stem from the discipline emphasized in martial arts. Studies show aggression varies across MA and CS styles [14–16, 19, 46]. Martial arts apprentices showed lower levels of hostility and aggression than combat sports athletes [15]. MMA fighters displayed higher anger, physical aggression, hostility, and verbal aggression than karate practitioners [14]. Traditional MA training tends to reduce aggression, while modern styles may not. BJJ decreases aggression, while MMA may increase it, particularly in those with pre-existing high aggression [18, 19].

Training methods emphasizing discipline and emotional regulation are crucial for practitioners to harness aggression positively [77, 78]. Karate training reduces aggression and improves emotional self-regulation in adolescents [16, 77, 79, 80]. Martial artists in Kyokushin and Shotokan karate show greater self-regulation, mitigating aggressive impulses [77]. This stems from moral reasoning and self-discipline in traditional martial arts [11, 18, 59, 81]. Programs combining martial arts with therapeutic techniques reduce aggression in youth [82].

Studies show that beginners’ hostility decreases with continued martial arts training [43], unlike in non-combative or aggressive sports [61, 83]. Traditional martial arts, including meditation and philosophy, are more effective than modern martial arts at reducing aggression due to their focus on self-control [11, 59]. Internal techniques, such as breathing exercises, reduce hostility, highlighting the importance of mental factors in managing aggression [84]. While MA&CS training fosters emotional regulation [43, 85], the path to managing these skills varies across fight styles.

A correlation exists between aggression and injury severity in martial arts contexts. Studies show varying injury rates across disciplines, with Taekwondo having the highest (59%) and Tai Chi the lowest (14%) [86]. Age and training frequency were associated with injury risk, with experienced practitioners more prone to injury (83). Head impacts were the primary cause of stoppages in mixed martial arts matches [87]. MA&CS athletes with moderate to severe injuries showed increased aggression levels [60, 88, 89]. Research indicates a positive correlation between injury severity and aggressive behaviors [90].

Discipline differences and practical implications

Another important result is that discipline differences were limited rather than pervasive. Contrary to the initial expectations, combat sport discipline did not moderate the associations of either emotion regulation strategy with total aggression, physical aggression, or anger, and no interaction emerged for expressive suppression. Thus, the data do not support a broad claim that the emotion regulation-aggression relationship differs systematically across all four combat sport disciplines. Instead, the evidence points to one narrower sport-specific pattern involving cognitive reappraisal and hostility in wrestling and MMA. This restrained interpretation is more consistent with previous reviews emphasizing heterogeneity across martial arts and combat sports rather than a single psychological profile for all disciplines [11, 31, 32, 46, 91]. It is also compatible with prior work suggesting that traditional and more competition-oriented combat systems may differ in how self-control, discipline, and aggression are socialized. However, such differences should not be inferred mechanically from the present data alone [11, 15, 16, 43, 44, 59, 77, 79–85].

The covariate pattern also deserves note. Younger age was associated with higher total aggression and hostility, male sex with higher physical aggression, and female sex with higher anger and hostility in some models. These findings suggest that individual characteristics remain relevant when interpreting aggression scores in combat sport samples. In practice, the present results support psychologically informed training that distinguishes between reducing maladaptive hostility and managing rule-bound competitive intensity. Interventions should therefore focus less on the assumption that all aggression is uniformly problematic and more on helping athletes identify when cognitive or emotional strategies are linked to resentment, mistrust, or dysregulated anger rather than to controlled competitive engagement [29, 30, 38, 60].

Limitations of the study

The present study has several limitations. First, its cross-sectional design does not permit causal inference; therefore, the observed associations should not be interpreted as evidence that emotion regulation strategies affect aggression in a directional sense. Second, all key variables were measured using self-report questionnaires, which increases the risk of shared-method bias and fails to capture actual aggressive behavior during training or competition. Third, although age, sex, sport experience, and competition level were included as covariates in all moderation models, other potentially important control variables were not assessed, including training frequency and volume, weight category, bout exposure, injury history, coaching climate, trait impulsivity, and competitive success. Residual confounding, therefore, remains possible. Furthermore, the discipline groups differed in their demographic composition (e.g., gender distribution, age, and competition level). Although we adjusted for these covariates in our revised models, unmeasured confounding variables may still influence the results. Fourth, the study did not directly assess tactical or functional forms of competitive aggression, nor did it measure performance outcomes or emotional states in specific sporting situations; this limits the ability to distinguish rule-governed competitive intensity from hostile aggression. Fifth, the verbal aggression subscale showed poor internal consistency and was therefore excluded from the inferential analyses, which narrows conclusions about the full aggression profile. Finally, although the total sample size was adequate, some discipline-specific interaction effects may still have been estimated with insufficient precision due to unequal group sizes. Future studies would benefit from longitudinal or competition-based designs, broader behavioral measures, and more fine-grained sport-specific indicators of aggression and emotion regulation.

Conclusions

In conclusion, the relationship between emotion regulation and aggression in combat sports appears more selective and discipline-specific than initially assumed. Most associations between cognitive reappraisal, expressive suppression, and aggression outcomes were small or non-significant. The main consistent finding was that expressive suppression was positively associated with hostility across disciplines, whereas discipline moderated only the association between cognitive reappraisal and hostility, with positive conditional effects in wrestling and MMA but not in BJJ or karate. These results do not support broad causal claims about emotion regulation or the strategic use of hostility. Instead, they suggest that hostility should be distinguished from rule-bound competitive aggression and interpreted cautiously in light of the psychological demands of specific combat sports. Longitudinal and competition-focused studies are needed to clarify the direction of these associations and to identify which emotion regulation patterns are adaptive for performance without reinforcing maladaptive hostility. These cross-sectional findings highlight the complex psychological landscape of martial arts and combat sports. Tailored psychological support and mental conditioning may help athletes effectively manage negative emotions, optimize competitive performance, and minimize hostile tendencies.

Abbreviations

A

Anger

ANOVA

Analysis of variance

BJJ

Brazilian Jiu-Jitsu

BPAQ

Buss and Perry Aggression Questionnaire

CR

Cognitive reappraisal

CS

Combat sports

ERQ

Emotion Regulation Questionnaire

ES

Expressive suppression

H

Hostility

MA

Martial arts

MA&CS

Martial arts and combat sports

MMA

Mixed Martial Arts

PA

Physical aggression

VA

Verbal aggression

Authors’ contributions

CK and WC conceptualized the study, conducted the investigation, curated the data, supervised the project, secured funding, and participated in the review and editing of the manuscript. AMR reviewed the literature and managed resources, performed the formal analysis, validated and interpreted the data, created data visualization, drafted the original manuscript, and contributed to its review and editing. All authors read and approved the final version of the manuscript and agree with the authorship.

Funding

The research did not have any financial support.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and research protocol was approved by the Research Ethics Committee at the Opole University of Technology (Decision No. 2/2022, 9 July, 2022). Written informed consent was obtained from all participants before starting the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.