Effectiveness of a cognitive behavioural intervention on mental health in young footballers: A quasi-experimental study

Abstract

BACKGROUND

Adolescence and early adulthood are critical periods for mental health in footballers, during which competitive demands such as performance expectations, training load, and competitive pressure may negatively affect psychological well-being. In football, these demands can lead to emotional deterioration throughout the season, underscoring the need for cognitive behavioural interventions that promote mental health and performance.

AIM

To evaluate the effectiveness of a psychological intervention programme based on cognitive-behavioural techniques on the mental health of young footballers.

METHODS

A quasi-experimental longitudinal design was applied with non-equivalent groups (experimental and control) and pre-test and post-test measurements. The sample consisted of 56 young footballers (53.6% men and 46.4% women; mean age 19.45 years) competing in the men’s Juvenil División de Honor and the women’s Third Division of the Royal Spanish Football Federation. The intervention was delivered throughout the competitive season and included eleven group sessions and four individual sessions of 50 minutes each, focusing on motivation, concentration, visualisation, activation, confidence, and group cohesion. Psychological variables were assessed using validated self-report instruments measuring trait anxiety, emotional distress, mood states, general psychological distress, and subjective sleep quality.

RESULTS

The experimental group showed more favourable progress in several mental health indicators compared to the control group. Specifically, the intervention contributed to the stabilisation or reduction of stress and depressive symptoms, a decrease in fatigue and tension, and an increase in vigour, as well as improvements in overall well-being and reductions in psychological distress. More stable dimensions, such as trait anxiety and subjective sleep quality, remained unchanged in both groups.

CONCLUSION

The results suggest that the cognitive behavioural therapy intervention acted as a protective factor against the psychological burden inherent in competitive football seasons, contributing to the preservation of players’ well-being. These findings support the systematic integration of cognitive behavioural programmes in youth and women’s football as a preventive and educational strategy in highly demanding sporting contexts. Future research should examine the long-term effects of such interventions and explore their applicability across different competitive levels and age groups.

Key Words: Psychology; Football; Psychological intervention; Mental health; Youth athletes

Core Tip: We evaluated whether a cognitive-behavioural intervention could improve mental health outcomes in young footballers using a quasi-experimental design. We compared participants who received the intervention with a comparison group and assessed mental health indicators before and after the programme. Overall, the intervention was associated with improvements in key mental health measures, suggesting that brief, structured cognitive-behavioural strategies may be useful to support psychological wellbeing in youth sport settings. These findings highlight the value of integrating evidence-based mental health interventions within youth football programmes.

INTRODUCTION

Adolescence is a stage of physical, psychological and social change, sensitive to environmental stimuli, which can increase vulnerability to anxiety, stress, depression and other mental health problems[1].

During adolescence, mental health is a public health priority, with a prevalence of approximately 13%, and almost half of mental disorders begin before the age of 18[2]. Sport can promote subjective well-being, emotional self-management and social skills[3,4]; however, in high-performance youth contexts, factors such as early specialisation, excessive workload and pressure to perform can increase the risk of emotional difficulties and reduced enjoyment[5,6].

Furthermore, at the individual level, factors such as recent injuries, ADHD, fear of failure, maladaptive perfectionism, avoidant coping, and sleep problems increase the risk of anxiety, depression, and burnout in young athletes, as well as emotional vulnerability and susceptibility to injury[7-11]. Therefore, it is recommended to monitor the mental health of young athletes throughout the season and to carry out psychological interventions that go beyond merely improving athletic performance[12-16].

Mental health problems in young footballers, such as anxiety, depression, burnout, and maladaptive perfectionism, can significantly reduce enjoyment of the sport, motivation, confidence, impair concentration, and negatively affect performance both in training and matches[8,15,17]. These difficulties are often amplified in high-pressure environments, where constant criticism from coaches or parents and limited social support can increase emotional distress and the risk of dropout. In the absence of structured cognitive-behavioural interventions, young players may struggle to cope with mistakes, injuries, or competitive pressure and to regulate their emotional responses effectively[18]. Implementing systematic cognitive-behavioural programmes is therefore essential, not only to prevent these negative outcomes but also to foster psychological well-being and support long-term athletic development.

In this context, cognitive-behavioural psychological techniques have been shown to be effective in improving athletic performance and caring for athletes’ mental health by training factors such as motivation, self-confidence, concentration, emotional self-management[19-21], and group cohesion[22-26]. In youth football, these interventions are especially relevant because they provide young players with practical tools to regulate emotions, manage competitive pressure, cope with errors and injuries, and develop adaptive cognitive and emotional regulation strategies during a critical period of psychological development, thereby supporting both psychological well-being and long-term athletic development[18]. These issues have also been confirmed in youth football, both in individual and group interventions[27-31].

However, there is a need to strengthen studies on the systematic evaluation of psychological intervention programmes in young footballers, which allow their effect to be measured throughout the competitive season, considering not only athletic performance but also indicators of mental health and psychological well-being.

The aim of this study is to evaluate the effectiveness of a psychological intervention programme based on cognitive-behavioural techniques on indicators of mental health and psychological well-being in young footballers during a competitive season. The hypothesis is that footballers participating in the intervention programme will show improvements in indicators of mental health and psychological well-being compared to those who continue with their usual training routine without psychological training.

MATERIALS AND METHODS

Design

A quantitative, descriptive, quasi-experimental, longitudinal study was conducted with non-equivalent groups and repeated measures at pre-test and post-test[32]. Quasi-experimental designs are used when random assignment is not possible, and comparison groups are naturally formed. In this case, it was impossible to randomly assign the players, who were already organized into teams with fixed schedules and routines. Specifically, non-equivalent pre-test-posttest designs allow for the comparison of changes in outcome variables before and after an intervention in existing groups, measuring changes in variables over time[33]. Although this design does not allow for definitive causal conclusions, it is suitable in applied sport psychology contexts, as it balances methodological rigor with practical feasibility.

The reporting of this quasi-experimental study was guided by the Transparent Reporting of Evaluations with Nonrandomized Designs statement, which provides criteria to ensure transparent and complete reporting of nonrandomized intervention research, including study design, participants, interventions, outcome measures, and analytic methods[34].

Participants

The sample consisted of 56 footballers from five teams in the Region of Murcia (Spain): (1) 30 players from the Men’s Youth Division of Honour; and (2) 26 players from the Women’s Third Division of the Royal Spanish Football Federation.

A non-probabilistic convenience sampling method was used. Teams were selected based on their availability and feasibility to participate in the psychological intervention, considering the need to dedicate time and provide suitable facilities for the sessions. In line with the quasi-experimental design, group allocation was determined by the teams’ willingness to participate; no randomization was performed[35].

Teams were initially contacted through the club management, who expressed strong interest in having the intervention implemented. Subsequently, the management contacted the coaches, who also agreed to participate, and finally the footballers were approached to invite them to participate voluntarily in the study.

The mean age of the male group was 17.48 years (SD = 0.75), while that of the female group was 21.50 years (SD = 3.29), with an overall sample mean of 19.45 years (SD = 3.09). The mean sporting experience was 12.43 years (SD = 2.02) for men and 11.46 years (SD = 5.39) for women, with an overall mean of 11.98 years (SD = 3.95). Participants trained 3-4 days per week in 1.5-hour sessions and competed in league matches on weekends.

Inclusion criteria were being a member of the selected teams participating in the intervention, completing both the pre-test and posttest assessments, and attending the workshops included in the intervention program. The number of potential participants contacted was 65 players (31 in the experimental group and 34 in the control group). Of these, 7 players from the experimental group did not complete the study due to missing some of the workshop sessions, while 2 players from the control group were excluded for not completing some of the pre-assessments and post-assessments.

The experimental group (n = 24) included players from one male team (n = 14) and one female team (n = 10), while the control group (n = 32) included players from two male teams (n = 17) and one female team (n = 15). The unequal number of participants is due to logistical and availability factors, as some players did not complete either the pre-test or post-test, although initially, the aim was to have more balanced group sizes.

The sample size was determined by the availability of teams and players who met the inclusion criteria and agreed to participate during the study period, which is consistent with applied quasi-experimental research conducted in real-world sporting contexts.

The use of convenience sampling and non-equivalent groups is consistent with quasi-experimental designs applied in real-world sporting contexts, where participant assignment is determined by the existing team structure rather than randomization[36].

The program evaluation and intervention were conducted by a licensed psychologist with a master’s degree in sports psychology and over five years of professional experience, supervised by a senior sports psychologist with more than 25 years of professional experience.

Data collection instruments

Trait anxiety: This was assessed using the revised Spanish version of the State-Trait Anxiety Inventory (STAI-T), developed by Buela-Casal et al[36] based on the original instrument by Spielberger et al[37]. Only the Trait Anxiety subscale was administered, consisting of 20 items with a Likert format from 0 to 3 (range 0-60), where higher scores indicate greater anxiety. The Spanish adaptation has shown high internal consistency (α = 0.86-0.90) and adequate convergent validity.

Emotional distress: This was assessed using the Depression Anxiety Stress Scales-21 (DASS-21)[38] and its Spanish adaptation[39]. This questionnaire consists of 21 items distributed across three subscales (depression, anxiety, and stress), with seven items each, and responses on a Likert scale from 0 to 3. The instrument has high reliability (α = 0.82-0.90) and is suitable for detecting emotional symptoms in athletes and monitoring their evolution. Higher scores indicate a greater level of emotional distress.

Mood states: These were assessed using the 29-item abbreviated version of the Profile of Mood States (POMS)[40,41], adapted into Spanish, which measures tension, depression, anger, vigour and fatigue on a Likert scale from 0 to 4. This version, widely used in sports research, has shown acceptable internal consistency (α = 0.70-0.92) and allows the identification of emotionally relevant patterns, such as the so-called ’iceberg profile’ described by Cancro[42] and Morgan[43]. In each subscale, higher scores indicate a higher level of the corresponding factor; in tension, depression, anger and fatigue, they reflect greater distress or negative affect, while in vigor they indicate greater energy and positive affect.

General psychological distress: This was assessed using the 12-item General Health Questionnaire (GHQ-12), a brief, internationally validated instrument for detecting psychological distress[44-46]. Each item is scored on a Likert scale from 0 to 3, yielding a total score that reflects the level of general distress. The GHQ-12 has good internal consistency (α = 0.85) and adequate sensitivity for detecting emotional disturbances in both athletic and general populations. Higher scores indicate greater psychological distress.

Subjective sleep quality: This was assessed using an item derived from the Pittsburgh Sleep Quality Index (PSQI)[47] and its Spanish adaptation[48]. The question used was: ’During the last month, how would you rate your overall sleep quality?’, with four Likert-type response options: (1) Very good; (2) Good; (3) Poor; and (4) Very poor. This item provides an overall estimate of sleep perception during the previous month and is a brief and practical indicator of subjective sleep quality. The full version of the PSQI has shown satisfactory internal reliability (α = 0.81).

Data collection procedure

Contact with participants was made through the sports management and coaches of the clubs, coordinating with them the schedules of the sessions and ensuring their willingness to collaborate in the research. Before the start of the study, informed consent was requested from all participants; in the case of subjects under 16 years of age, this consent was signed by their parents or legal guardians. After obtaining consent, the pre-test evaluation was administered in October 2023 using a battery of standardised self-reports (DASS-21, POMS, GHQ-12, PSQI and STAI-T). In May 2024, coinciding with the end of the season, the same procedure was repeated to carry out the post-test evaluation.

The psychological intervention for the experimental group included 11 sessions of 50 minutes each, held every two weeks. The assessment and psychoeducational workshops were conducted in groups and in person in a room provided by the club, located in the same sports complex where the training sessions were held, which met the appropriate conditions for use (temperature, noise, lighting, space, and privacy).

During the workshops, psychological strategies were applied to improve motivation, attention, visualisation, activation and confidence, together with dynamics aimed at developing cohesion. Although the intervention was the same, it was applied separately to the male and female teams. Thus, all core intervention sessions were delivered in a group format within each team separately, without mixing male and female participants.

On a complementary and voluntary basis, four individual sessions of 50 minutes each, distributed at a rate of one session per month, were offered to players who requested them (n = 14), of whom nine belonged to the men’s Juvenil División de Honor team and five to the women’s Third Division of the Royal Spanish Football Federation. These individual sessions were offered as an optional, complementary support for players in the experimental group who wished to receive extra personalised guidance. This option was not available to the control group, which only completed the pre-test and post-test assessments and did not receive any intervention. Only a subset of players in the experimental group chose to participate in these individual sessions. These sessions addressed the specific needs of each participant (personal and sporting) and were held in the same room at the sports facilities, with a variable schedule depending on the availability of each player.

The psychological strategies used during the intervention were established based on Labrador’s Behaviour Modification Manual[49]. Each of the variables trained included a first phase of psychoeducation, aimed at explaining the concepts and their relevance to both athletic performance and psychological well-being, and a second phase of practical workshops, focused on learning and applying the techniques.

Specifically, the following variables were worked on: (1) Motivation, using the goal-setting technique; (2) Attention, through goal setting, thought stopping and self-instruction, as well as changing the focus of attention (internal-external and broad-narrow); (3) Level of activation, through breathing techniques and positive self-talk; (4) Visualisation, using breathing techniques and guided visualisation aimed at performance; and (5) Confidence, trained from sources of self-efficacy, goal setting and positive self-talk. Specific activities were also developed to strengthen group cohesion, through the establishment of common goals and the promotion of positive interpersonal relationships. The content and tasks of the intervention programme are summarised in Table 1.

Table 1 Contents and tasks of the intervention programme.

Session	Content	Between-session tasks
Session 1	Programme information, psychological assessment	None required
Session 2	Goal setting	Weekly recording of performance goals for training sessions and matches
Session 3	Motivation workshop	Motivation diary
Session 4	Activity: Goal-setting work	None required
Session 5	Attention workshop	Practise the mini routine selected during the workshop in training sessions and subsequently apply it in matches. Record perceived effectiveness
Session 6	Activity: Attention work	None required
Session 7	Visualisation workshop	Carry out one visualisation rehearsal before training/competition, following the script developed
Session 8	Activity: Group cohesion work	None required
Session 9	Activation workshop	Design a personal plan: Two strategies to increase activation and two to decrease it before/during matches; brief in-room rehearsal
Session 10	Confidence workshop	Functional analysis: ABC
Session 11	Programme closure, psychological assessment	None required

Statistical analysis

Descriptive statistics (mean ± SD) were calculated for all psychological variables to describe the evolution of the control and experimental groups from the pre-test to the post-test. Given the small sample size and the non-normal distribution observed in several variables, non-parametric tests were used. Intragroup comparisons were performed using the Wilcoxon signed-rank test, using Rosenthal’s effect size r (|0.10| = small, |0.30| = medium, |0.50| = large). Intergroup comparisons were performed using the Mann-Whitney U test, estimating the effect size using Hedges’ g index (0.20 = small, 0.50 = medium, 0.80 = large).

The analysis combined statistical significance (Z and P values) with effect size estimation, which allowed for the evaluation of the practical and clinical relevance of the results, even in those cases where statistical significance may have been limited by the small sample size.

Due to the quasi-experimental nature of the study, unequal group sizes, a small sample, and non-normal distributions in several variables, a non-parametric approach was adopted. Intragroup changes were analysed using the Wilcoxon signed-rank test, and intergroup comparisons were performed with the Mann-Whitney U test, complemented by effect sizes (Rosenthal’s r and Hedges’ g) to improve the interpretability of the results. This approach is consistent with applied quasi-experimental research in real-world contexts, where practical limitations and non-equivalent groups often require flexible methods that consider both statistical significance and practical relevance[33,50-52].

Ethical considerations

The study was conducted in accordance with the ethical principles established in the Declaration of Helsinki (World Medical Association)[53] and the ethical considerations for studies in sports science[54]. It was also approved by the Research Ethics Committee of the University of Murcia (No. 4734/2023) for studies involving human subjects. Participants were informed about the study and provided their voluntary consent to participate. For participants under 16 years of age, consent was signed by their parents or legal guardians. Confidentiality of the information was strictly maintained, and all data were treated anonymously.

Rigour

Several strategies were implemented to ensure methodological rigour in this quasi-experimental study. Validated and reliable instruments (STAI-T, DASS-21, 29-item abbreviated version of the POMS, GHQ-12, and PSQI) were used, and standardized procedures were applied across all intervention sessions, conducted by a licensed psychologist and supervised by a senior expert to ensure consistency and fidelity.

Inclusion and exclusion criteria were clearly defined, and only participants who completed the pre-test, post-test, and intervention sessions were analyzed. Data collection was performed under controlled conditions with informed consent, including parental consent for participants under 16.

Appropriate statistical analyses were conducted, combining significance testing with effect size estimation to evaluate both statistical and practical relevance. Confidentiality and anonymity of participants were strictly maintained.

These measures ensured that the study met high standards of reliability, transparency, and validity, despite the inherent limitations of a quasi-experimental design.

RESULTS

The results of the psychological variables assessed before and after the intervention are presented below, differentiating between the experimental group and the control group. For each instrument, the intragroup analyses (Wilcoxon test and Rosenthal’s effect size r) and intergroup analyses (Mann-Whitney U test and Hedges’ effect size g) are summarised. With regard to the STAI-T, no statistically significant differences in anxiety traits were observed in the control and experimental groups, indicating that levels remained stable throughout the season (Table 2). Furthermore, when comparing both groups, no statistically significant differences were observed (Table 3). In relation to the DASS-21, the control group showed an increase in depression from the pre-test (mean = 4.13; SD = 3.66) to the post-test (mean = 5.78; SD = 4.88), with a medium effect size (r = -0.30) and a trend towards statistical significance (P = 0.089), suggesting a slight worsening over the season. In this group, anxiety and stress remained stable, with no significant changes. In contrast, in the experimental group, no statistically significant changes were observed in depression, anxiety, or stress, indicating greater emotional stability in the athletes who received the intervention (Table 4). When comparing both groups, no significant differences were observed in depression levels in either the pre-test or post-test. In anxiety, the control group had significantly higher means than the experimental group in the pre-test (P = 0.010; g = 0.639) and showed a tendency to remain higher in the post-test (P = 0.068; g = 0.519). With regard to stress, although no significant differences were recorded at the beginning of the season, in the post-test the control group exhibited significantly higher levels than the experimental group (P = 0.039; g = 0.564; Table 5). In relation to the POMS, significant increases in tension and depression were observed in both groups from pre-test to post-test, with large effect sizes (control group: R = -0.68 to -0.80; experimental group: R = -0.67 to -0.73). Anger did not show statistically significant changes, while vigour decreased significantly in both groups (r = -0.81 to -0.82). In terms of fatigue, a slight decrease was observed in the control group, with a trend towards significance (r = -0.32;P = 0.067), and a significant reduction in the experimental group (r = -0.52; P = 0.011; Table 6). When comparing both groups, no statistically significant differences were observed in most of the subscales. However, in the post-test, tension (P = 0.051; g = 0.623) and fatigue (P = 0.048; g = 0.498) were significantly lower in the experimental group (Table 7). In relation to the GHQ-12, the control group did not show statistically significant changes in psychological distress levels between the pre-test and posttest. In contrast, the experimental group showed a tendency toward decreased psychological distress from the pre-test (mean = 2.13; SD = 1.90) to the posttest (mean = 1.29; SD = 1.60), with an effect close to statistical significance and a medium effect size (r = -0.34; Table 8). The comparison between groups showed that, at the end of the season, participants who received the intervention (mean = 1.29; SD = 1.60) had lower levels of psychological distress than the control group (mean = 3.03; SD = 3.25), with a trend toward statistical significance and a considerable effect size (P = 0.080; g = 0.642; Table 9). Regarding the PSQI, no statistically significant differences were observed in either the control group or the experimental group between the pre-test and post-test (Table 10). Likewise, no significant differences were found between groups at either of the two assessment points (Table 11). Figure 1 shows the evolution of average scores on the different instruments. In general terms, some emotional dimensions tend to deteriorate throughout the competitive season, as reflected in the increase in tension and depression in both groups. However, the experimental group showed less deterioration or even improvements in specific variables, such as fatigue, stress, and general psychological distress, suggesting a possible protective effect of the intervention.

Figure 1 Pre-test and post-test evolution of psychological variables in the control and experimental groups. GHQ: General Health Questionnaire; PSQI: Pittsburgh Sleep Quality Index.

Table 2 Pre-post changes by group in trait anxiety (Wilcoxon and Rosenthal’s r), mean ± SD.

Group	Pre-test	Post-test	Z (Wilcoxon)	P value	r (Rosenthal)
Control	22.38 ± 5.71	22.97 ± 6.74	-1.000	0.317	0.18
Experimental	21.71 ± 5.43	20.92 ± 4.87	-1.099	0.272	-0.22

Z: Test statistics of the Wilcoxon signed-rank test (normal approximation); P: Probability value associated with the within-group comparison; Rosenthal’s r: Effect size calculated from Z and n (|10| = small, |30| = medium, |50| = large).

Table 3 Comparison between groups in trait anxiety (Mann-Whitney and Hedges’ g).

Moment	Group	Sample size	mean ± SD	P value	g (Hedges)
Pre-test	Control	32	22.38 ± 5.71	0.914	0.118
	Experimental	24	21.71 ± 5.43
Post-test	Control	32	22.97 ± 6.74	0.327	0.336
	Experimental	24	20.92 ± 4.87

P: Probability associated with the contrast statistic (Mann-Whitney U test in intergroup comparisons); Hedges’ g: Effect size corrected for independent samples (0.20 = small, 0.50 = medium, 0.80 = large).

Table 4 Pre-post changes by group in the Depression Anxiety Stress Scales-21 (Wilcoxon test and Rosenthal’s r), mean ± SD.

Subscale	Group	Pre-test	Post-test	Z (Wilcoxon)	P value	r (Rosenthal)
Depression	Control	4.13 ± 3.66	5.78 ± 4.88	-1.701	0.089	-0.30
	Experimental	3.00 ± 2.98	3.46 ± 2.99	0.639	0.523	0.13
Anxiety	Control	5.75 ± 5.05	5.25 ± 4.28	-0.388	0.698	-0.07
	Experimental	2.92 ± 3.24	3.25 ± 3.04	-0.196	0.844	-0.04
Stress	Control	6.50 ± 5.08	6.38 ± 4.93	-0.226	0.821	-0.04
	Experimental	4.58 ± 3.12	3.79 ± 3.89	-1.388	0.165	-0.28

Z: Test statistics of the Wilcoxon signed-rank test (normal approximation); P: Probability value associated with the within-group comparison; Rosenthal’s r: Effect size calculated from Z and n (|10| = small, |30| = medium, |50| = large).

Table 5 Comparison between groups in Depression Anxiety Stress Scales-21 (Mann-Whitney and Hedges’ g).

Subscales	Moment	Group	Sample size	mean ± SD	P value	g (Hedges)
Depression	Pre-test	Control	32	4.13 ± 3.66	0.249	0.328
		Experimental	24	3.00 ± 2.98
	Post-test	Control	32	5.78 ± 4.88	0.103	0.548
		Experimental	24	3.46 ± 2.99
Anxiety	Pre-test	Control	32	5.75 ± 5.05	0.010b	0.639
		Experimental	24	2.92 ± 3.24
	Post-test	Control	32	5.25 ± 4.28	0.068	0.519
		Experimental	24	3.25 ± 3.04
Stress	Pre-test	Control	32	6.50 ± 5.08	0.244	0.434
		Experimental	24	4.58 ± 3.12
	Post-test	Control	32	6.38 ± 4.93	0.039a	0.564
		Experimental	24	3.79 ± 3.89

^aP < 0.05.

^bP < 0.01.

P: Probability associated with the contrast statistic (Mann-Whitney U test in intergroup comparisons); Hedges’ g: Effect size corrected for independent samples (0.20 = small, 0.50 = medium, 0.80 = large).

Table 6 Pre-post changes by group in Profile of Mood States (Wilcoxon and Rosenthal’s r), mean ± SD.

Subscales	Group	Pre-test	Post-test	Z (Wilcoxon)	P value	r (Rosenthal)
Tension	Control	8.22 ± 4.52	12.47 ± 2.55	-3.820	< 0.001c	-0.68
	Experimental	6.83 ± 4.36	10.92 ± 2.32	-3.260	0.001c	-0.67
Depression	Control	4.00 ± 3.48	9.03 ± 3.48	4.504	< 0.001c	0.80
	Experimental	2.88 ± 3.18	7.75 ± 2.69	-3.564	< 0.001c	-0.73
Anger	Control	11.22 ± 6.33	9.88 ± 4.82	0.891	0.373	0.16
	Experimental	9.46 ± 4.95	7.88 ± 3.64	-1.631	0.103	-0.33
Vigour	Control	12.88 ± 3.20	6.19 ± 2.48	-4.595	< 0.001c	-0.81
	Experimental	12.54 ± 2.52	6.13 ± 3.38	-4.021	< 0.001c	-0.82
Fatigue	Control	6.28 ± 4.03	4.88 ± 4.02	-1.832	0.067	-0.32
	Experimental	5.33 ± 4.44	3.00 ± 3.24	-2.553	0.011a	-0.52

^aP < 0.05.

^cP < 0.001.

Z: Test statistics of the Wilcoxon signed-rank test (normal approximation); P: Probability value associated with the within-group comparison; Rosenthal’s r: Effect size calculated from Z and n (|10| = small, |30| = medium, |50| = large).

Table 7 Comparison between groups in Profile of Mood States (Mann-Whitney and Hedges’ g).

Subscale	Moment	Group	Sample size	mean ± SD	P value	g (Hedges)
Tension	Pre-test	Control	32	8.22 ± 4.52	0.212	0.307
		Experimental	24	6.83 ± 4.36
	Post-test	Control	32	12.47 ± 2.55	0.051	0.623
		Experimental	24	10.92 ± 2.32
Depression	Pre-test	Control	32	4.00 ± 3.48	0.186	0.330
		Experimental	24	2.88 ± 3.18
	Post-test	Control	32	9.03 ± 3.48	0.194	0.399
		Experimental	24	7.75 ± 2.69
Anger	Pre-test	Control	32	11.22 ± 6.33	0.397	0.300
		Experimental	24	9.46 ± 4.95
	Post-test	Control	32	9.88 ± 4.82	0.137	0.452
		Experimental	24	7.88 ± 3.64
Vigour	Pre-test	Control	32	12.88 ± 3.20	0.526	0.112
		Experimental	24	12.54 ± 2.52
	Post-test	Control	32	6.19 ± 2.48	0.591	0.021
		Experimental	24	6.13 ± 3.38
Fatigue	Pre-test	Control	32	6.28 ± 4.03	0.303	0.222
		Experimental	24	5.33 ± 4.44
	Post-test	Control	32	4.88 ± 4.02	0.048a	0.498
		Experimental	24	3.00 ± 3.24

^aP < 0.05.

P: Probability associated with the contrast statistic (Mann-Whitney U test in intergroup comparisons); Hedges’ g: Effect size corrected for independent samples (0.20 = small, 0.50 = medium, 0.80 = large).

Table 8 Pre-post changes by group in 12-item General Health Questionnaire (Wilcoxon and Rosenthal’s r), mean ± SD.

Group	Pre-test	Post-test	Z (Wilcoxon)	P value	r (Rosenthal)
Control	3.28 ± 3.05	3.03 ± 3.25	-0.122	0.903	-0.02
Experimental	2.13 ± 1.90	1.29 ± 1.60	-1.652	0.098	-0.34

Z: Test statistics of the Wilcoxon signed-rank test (normal approximation); P: Probability value associated with the within-group comparison; Rosenthal’s r: Effect size calculated from Z and n (|10| = small, |30| = medium, |50| = large).

Table 9 Comparison between groups in 12-item General Health Questionnaire (Mann-Whitney and Hedges’ g).

Moment	Group	Sample size	mean ± SD	P value	g (Hedges)
Pre-test	Control	32	3.28 ± 3.05	0.282	0.435
	Experimental	24	2.13 ± 1.90
Post-test	Control	32	3.03 ± 3.25	0.080	0.642
	Experimental	24	1.29 ± 1.60

P: Probability associated with the contrast statistic (Mann-Whitney U test in intergroup comparisons); Hedges’ g: Effect size corrected for independent samples (0.20 = small, 0.50 = medium, 0.80 = large).

Table 10 Pre-post changes by group in Pittsburgh Sleep Quality Index (Wilcoxon and Rosenthal’s r), mean ± SD.

Group	Pre-test	Post-test	Z (Wilcoxon)	P value	r (Rosenthal)
Control	0.94 ± 0.76	1.03 ± 0.82	-0.615	0.539	-0.11
Experimental	1.08 ± 0.78	0.92 ± 0.72	-0.723	0.470	-0.10

Z: Test statistics of the Wilcoxon signed-rank test (normal approximation); P: Probability value associated with the within-group comparison; Rosenthal’s r: Effect size calculated from Z and n (|10| = small, |30| = medium, |50| = large).

Table 11 Comparison between groups in Pittsburgh Sleep Quality Index (Mann-Whitney and Hedges’ g).

Moment	Group	Sample size	mean ± SD	P value	g (Hedges)
Pre-test	Control	32	0.94 ± 0.76	0.411	-0.188
	Experimental	24	1.08 ± 0.78
Post-test	Control	32	1.03 ± 0.82	0.711	0.145
	Experimental	24	0.92 ± 0.72

P: Probability associated with the contrast statistic (Mann-Whitney U test in intergroup comparisons); Hedges’ g: Effect size corrected for independent samples (0.20 = small, 0.50 = medium, 0.80 = large).

DISCUSSION

This study evaluated the effectiveness of a psychological intervention programme based on cognitive-behavioural techniques in young footballers throughout a competitive season. Overall, the results support the hypothesis: Footballers who participated in the programme showed improvements in mental health and psychological well-being indicators compared to those who followed their usual training routine without psychological intervention. The findings obtained in the different psychological variables studied are presented and analysed below.

In the DASS-21, the control group showed increases in depression and stability in anxiety and stress, while the experimental group maintained stable levels of depression, anxiety, and stress. Furthermore, when comparing both groups, in the post-measurement, the control group had significantly higher scores in stress. From the theoretical perspective of the DASS-21[38,39], depression reflects low mood and loss of interest, anxiety indicates worry and excessive arousal, and stress represents difficulties in managing arousal and tension. The stability and slight improvement observed in the experimental group suggest that the intervention provided coping and emotional regulation resources, preventing the emotional deterioration associated with competitive demands, consistent with previous findings in athletes[55,56].

In the POMS, both groups showed increases in tension and depression and decreases in vigour, a pattern expected towards the end of the season due to the accumulation of competitive stress regardless of whether or not there is psychological intervention. According to POMS theory[40,41], stress reflects arousal and anxiety, depression indicates demotivation and sadness, vigour represents positive energy, and fatigue reflects a decrease in physical and psychological resources. The experimental group showed significant reductions in fatigue and tension, demonstrating better regulation of activation and preservation of energy in the face of the demands of the season, which reinforces the preventive efficacy of psychological strategies[57,58].

In terms of general psychological distress measured with the GHQ-12, the control group remained stable, while the experimental group showed moderate improvement, reflecting a more positive perception of overall well-being. Theoretically, the GHQ-12[44,45] allows for the assessment of the overall perception of stress and psychological distress, beyond specific symptoms, indicating that the intervention may have had an impact on the overall well-being of the athletes and not only on isolated emotional dimensions[59,60].

In terms of trait anxiety (STAI-T), no significant changes were observed throughout the season, which coincides with the relative stability of this construct described by Spielberger et al[37]. According to theory, trait anxiety reflects a relatively stable predisposition to perceive situations as threatening and to respond with anxiety, so it tends to remain constant even in the face of short-term psychological interventions. This is in line with the evidence from Niering et al[61], which indicates that psychological interventions tend to have a more pronounced impact on state anxiety, a transient and situational manifestation of anxiety, than on trait anxiety, which is more stable and long-lasting. Therefore, the lack of significant changes in the STAI-T was to be expected and does not indicate a lack of programme effectiveness, but rather the natural resistance of this construct to short-term modifications.

In terms of sleep quality (PSQI), the intervention did not generate significant changes, coinciding with Rijken et al[62], who applied general mental training to footballers. In contrast, specific interventions focused on sleep have shown improvements[63], suggesting that their optimisation requires protocols aimed at sleep hygiene and circadian regulation, beyond general psychological training. Furthermore, the use of a single item may have limited the sensitivity to detect subtle changes, an aspect to consider when interpreting the results.

CONCLUSION

The findings suggest that cognitive-behavioural intervention acted as a preventive and protective resource against psychological deterioration associated with the competitive burden and emotional demands of football in young people. In particular, the intervention contributed to maintaining stable or reducing indicators of stress and depression (DASS-21), decreasing fatigue and tension and preserving vigour (POMS), as well as improving overall perception of well-being and reducing psychological distress (GHQ-12). However, no significant changes were observed in trait anxiety (STAI-T) or subjective sleep quality (PSQI), which is consistent with the theory that posits the relative stability of these constructs and their lower sensitivity to general short-term interventions, suggesting that more specific or prolonged strategies may be required to achieve significant changes.

Furthermore, the results reinforce the effectiveness of multicomponent psychological programmes, which combine techniques for motivation, concentration, visualisation, self-confidence, and emotional regulation, as opposed to the application of a single technique, as they allow for a comprehensive approach to different aspects of well-being and performance. Furthermore, combining group intervention with individual sessions, conducted by psychologists specialising in sport and supervised by a senior psychologist, optimises the personalisation and effectiveness of the programme. Integrating these strategies into regular training promotes both improved athletic performance and the protection of young athletes’ mental health in a sustainable manner.

Future research should employ longitudinal designs with multiple assessment points, expand sample size and diversity, include objective measures and physiological biomarkers, and combine self-reports with qualitative techniques. It would also be advisable to explore other sports and competitive levels, analyse gender differences, evaluate athletic performance and additional psychological variables such as resilience or optimism, and intervene directly on specific mental health indicators such as anxiety and stress.

Strengths and limitations

This section highlights the main strengths and limitations of the study. Among the strengths, the study was conducted in a real-world training and competition context, which increases ecological validity. The psychological intervention was multicomponent, combining techniques for motivation, attention, visualisation, activation, and confidence, and included both group and individual sessions tailored to participants’ needs. The programme was delivered by experienced sports psychologists, ensuring a high level of professional supervision.

Among the limitations, a quasi-experimental design without randomisation was used, with pre-existing teams, which may have generated uncontrolled initial differences. The sample size, which was small and limited to a single geographical context, restricts the statistical power and generalisability of the results. Only two assessment points (beginning and end of the season) were included, preventing analysis of the progression of changes and capturing only the final effects. Additionally, no objective performance indicators or physiological parameters were incorporated, and the exclusive reliance on self-reports may generate perception or social desirability biases.

A limitation of the present study is the absence of a formal group × time interaction model, which would allow a more direct estimation of the intervention effects. However, given the small and unequal group sizes, participant attrition between measurements, and violations of normality assumptions, the use of non-parametric analyses combined with systematic effect size estimation was considered the most appropriate and conservative strategy.

This approach is aligned with methodological recommendations for applied quasi-experimental research in sports and health psychology, where ecological validity and feasibility often take precedence over strict statistical control, and where effect size estimation provides critical information on the practical and clinical relevance of the findings[33,51].