Prospective, unicentric, randomized, unblinded, pre- and postintervention study with a masked randomization sequence by blocks (NCT06006871) [30]. Participants were randomized by electronic case report form (eCRF) in 2 groups: group 1 (MOON), which received standard pharmacological treatment combined with personalized cognitive training through a serious video game designed for patients with ADHD, along with psychoeducational support for parents; and group 2 (control), which received standard pharmacological treatment and psychoeducational support for parents, without the video game intervention. The study followed a parallel assignment model (MOON vs control) with a 1:1 allocation ratio.

No major changes were made to the methodology before the commencement of the clinical trial. In this randomized clinical trial, the study design and reporting adhered to the CONSORT 2010 guidelines, specifically following the CONSORT checklist for reporting information in randomized trials (Multimedia Appendix 1). The research questions were formulated using the Population, Intervention, Comparison, Outcomes and Study framework. The study protocol is published [30] and registered on ClinicalTrials.gov (NCT06006871) with International Registered Report Identifier (DERR1-10.2196/53191).

This study received approval from the Research Ethics Committee of the Puerta de Hierro University Hospital on December 14, 2022 (PI 106/22). Authorization was subsequently granted by the Spanish Agency of Medicines and Health Products on February 14, 2023 (1061/22/EC-R). The study was monitored by an independent monitor. All participants signed the informed consent form. Data were anonymized through the assignment of a clinical trial-specific code. The data were treated confidentially in accordance with the Organic Law 3/2018 of December 5 on the protection of personal data and guarantee of digital rights. This study complied with the standards of good clinical practice and the declaration of Helsinki. Participants were not compensated financially.

Children and adolescents aged between 7 and 18 years (mean age 12.68, SD 2.75 years) with a clinical diagnosis of ADHD according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, were enrolled in the study. All participants had a clinical diagnosis of ADHD in any presentation and were receiving stable pharmacological treatment for ADHD. Patients were clinically stable, with a Clinical Global Impression (CGI) score between 3 and 6 before entering the clinical trial. Comorbidity was not an exclusion criterion in the study, except for those patients with suicide risk, lack of ability to follow verbal instructions, or motor difficulties that make it difficult to play a video game. Exclusion criteria also included participation in similar video games studies or the intention to initiate psychotherapeutic treatment during the 3-month study period. Medication was not altered during the study, unless adjustments were necessary for clinical reasons. A dropout rate of 15% was expected.

The randomized clinical trial had a total duration of 7 months, concluding on December 15, 2023. Recruitment took place between May and October 2023. The first evaluation was performed on May 9, 2023. The last evaluation was performed on October 31, 2023.

The intention-to-treat (ITT) population was defined as randomized participants who completed at least 10 sessions. The per-protocol (PP) population was defined as participants who completed 100% of the training. Only 32% (12/38) participants in the MOON group completed all 20 treatment sessions (PP), so we have not performed the analyses with this small sample size. In the ADHD population, adherence difficulties are common in long-term treatments; therefore, we planned to conduct the analyses on the premise of the ITT analysis. An additional analysis was also included to evaluate those participants who engaged with the intervention by completing 80% treatment (16 sessions in the MOON group or engagement for 8 weeks in the control group). Recognizing that the number of sessions could be a relevant covariate, we performed an analysis of covariance (ANCOVA), which was not initially included in the protocol. Only half of the parents (38/76, 50%; Multimedia Appendix 2) provided information on academic grades. Therefore this primary outcome (hypothesis 4) was excluded due to insufficient data.

All participants were recruited from the child and adolescent psychiatry outpatient clinics of Hospital Universitario Puerta de Hierro Majadahonda. The MOON group had 12 face-to-face visits, whereas the control group had only 2 presential visits: one at the preevaluation visit (D0) and another at the postevaluation visit (D90). The total duration of the research was 3 months for each participant. During the preinclusion phase, the principal investigator (HB-F) informed eligible patients about the procedure and research protocol. A total of 155 participants were assessed for enrollment. The study coordinator (MM-M) provided a detailed explanation of the research procedure and summoned the prerecruited participants. In total, 7.7% (12/155) of the patients did not meet the inclusion criteria (main reasons were clinical CGI, no medication, or age); 43.2% (67/155) of the patients declined to participate (the most common reason being lack of availability).

During the inclusion visit (D0), the research procedure was thoroughly explained, the informed consent form was signed, and the preevaluation was conducted (refer to the Study Outcomes section). Afterward, participants were randomly assigned to groups. Group randomization (MOON vs control) was performed using block randomization through the electronic eCRF REDCap (Research Electronic Data Capture; Vanderbilt University) with a 1:1 allocation ratio.

The intervention training (D1-D90) varied according to the assigned group. Parents in the control group received weekly telephone monitoring. For the MOON group, 20 sessions with the video game were scheduled: 10 sessions were conducted in the hospital with the researchers, and 10 sessions took place on the web at home, with participants using their computers under the researchers’ supervision (twice a week, adjusted according to the availability of the participants). Patients aged >12 years played MOON with VR, while those aged <12 years performed the video game sessions on a computer, following PlayStation guidelines. MM-M and MB-F assisted participants during their initial use of the video game, providing instructions, encouraging progress, and helping them manage any frustration caused by mistakes made while playing MOON. They also monitored for signs of motion sickness with VR, ensured proper use of the software, and addressed any potential frustration resulting from bugs in the video game. If a patient reported dizziness, even mild, they were instructed to remove the device and were given the option to decide whether or not to continue using it. All participants followed the same order pattern of the MOON games during 20 minutes of gameplay (refer to the study by Martin-Moratinos et al [30] for details).

In week 5 (D45), all participants’ parents receive a follow-up questionnaire with the main variable (SDQ). For the MOON group, both parents and children were trained on how to use the video game at home, using a flash drive. Parents and patients were additionally given an instruction manual. Monitoring was provided by telephone where the researchers checked the correct playability at home. The web-based sessions conducted were monitored by MM-M through the PlayFab data server (Microsoft Corporation). All participants had their personal username and password connected with PlayFab data server.

In the final visit (D90), a postevaluation was conducted using the same questionnaires, and participants were provided with feedback on the most relevant changes observed.

The MOON video game was designed with the Unity (version 2020.3.7; Unity Technologies) software tool by a multidisciplinary team. Autodesk 3ds Max, Autodesk Maya, Adobe Photoshop, and PreSonus Studio 4 programs were used for the modeling of characters and landscape elements. In relation to the narrative and identity features [31,32], the player is immersed in a forest with archaic chess statues of an ancient civilization [29,31]. The raccoon Movi and the fox One will help the player to solve the puzzles (game mechanics). MOON is mainly composed of a menu screen and 5 different game mechanics focused on cognitive training.

To provide manipulation and control features [32], each player has their own user profile with saved progress. Each test is set in a closed environment with distinct esthetics and game mechanics. The manipulation of the environment and interactivity are carefully controlled to avoid distractions or dizziness in the VR setting [28]. For example, in Kuburi, blocks can be rotated and moved within the space, while in Chess, the player can shift positions to view the board from different perspectives. The mobility within environments is restricted based on the individual’s cognitive capacity. In cases where mobility could cause distraction—such as in tests of sustained attention or working memory, which are more challenging for individuals with ADHD—exploration is limited.

In all game mechanics, players begin at a base level (tutorial, level 1) to ensure they understand the task. The difficulty increases in a personalized manner based on performance, ensuring the game is neither too easy (and thus boring) nor too difficult (and thus frustrating). Emotional regulation maturity, specifically the ability to reappraise, can be trained through controlled exposure to negative emotional stimuli, helping players overcome levels and obstacles within the game [31]. Regarding reward and punishment features [32], the MOON video game uses a star system based on player performance (0 stars=poor performance, must repeat the level; 1 star=acceptable; 2 stars=good; and 3 stars=excellent). Player progress in each game mechanic is communicated through level advancements, performance parameters, and the number of stars earned [30].

In terms of music, each game mechanic features different tracks [25]. The music was specifically designed to align with cognitive abilities. For example, sustained attention, which requires more effort for individuals with ADHD [24], benefits from dynamic soundtracks. To prevent monotony and boredom, more epic, fast-paced music was composed for Smasher to maintain motivation through its rhythm. Feedback is not only visual but also auditory, with distinct sounds depending on user interactions. Different sounds were created to indicate level selection and hits or misses, as well as to signal conditions of victory (successfully completing a level) or defeat (repeating the level after earning 0 stars due to poor performance) [25]. In this clinical trial, the same version of the video game was proposed in 2 platforms: VR and computer (Table 1) [30].

A PlayStation 4 Dev Kit and Test Kit were used for VR version. The Dev Kit and Test Kit are development consoles equipped with debug firmware that allow the execution of unsigned builds; access to memory and performance analysis tools; real-time error trace logging; as well as software compatibility, optimization, and development testing in a controlled environment. The PlayStation VR for PlayStation 4 is a VR headset featuring a 5.7-inch OLED display with a resolution of 1920×1080 (960×1080 per eye), a field of view of approximately 100°, and a refresh rate of 90/120 Hz. A PlayStation camera was used for motion tracking through an accelerometer, gyroscope, and 9 LEDs. The participants had to be placed at a distance of 1.5 to 2 m with a not very intense light for the correct calibration of the PlayStation camera with the LED lights. The researchers monitored what each participant experienced in VR by viewing it on a computer screen.

The PlayStation 4 VR goggles weigh approximately 600 g, so they could be uncomfortable for some participants. Regarding the added weight of sound equipment, patients were given a choice between basic stereo headphones and over-the-ear headphones. Both had 3D stereo sound and noise cancelation.

A Stealth 15M A11SDK (MSI) laptop was used for participants aged between 7 and 11 years to play MOON in the hospital. The laptop specifications were as follows: Microsoft Windows 11 Home ×64 bits; 11th Gen Intel Core i7-1185G7, 3.00 GHz, 16 GB RAM, 8 GB storage memory; GeForce GTX 1660 Ti.

The recruitment period was between May 9, 2023, and October 31, 2023. The enrollment target was not completed (n=152) [30]. The last post evaluation was performed on December 13, 2023. The clinical trial ended on December 15, 2023. A total of 76 patients with ADHD participated in the clinical trial and signed a reported consent form. No patients were excluded. The 76 participants were randomized 1:1 (MOON: n=38, 50% and control: n=38, 50%).

The total dropout rate was 9% (7/76) of the participants (n=5, 71% MOON and n=2, 29% control) and did not exceed expectations (12/76, 15%). Lack of availability was the reason mentioned by the dropout cases for not participating. In total, 3% (2/38) of the patients requested to switch to the control group after preassessment and randomization to MOON due to insufficient time to commit to the study, which was not allowed. Another MOON participant had to drop out of treatment due to lack of availability after 2 sessions. In total, 5% (4/76) of the participants (n=2, 50% MOON and n=2, 50% control) were not assessed at postevaluation and were lost to follow-up (Figure 1). The ITT population was analyzed for the primary and secondary outcomes. The ITT population included 87% (33/38) of the patients assigned to the MOON group and 95% (36/38) of the patients assigned to the control group.

Table 2 shows the sociodemographic variables for all participants who completed a preevaluation (N=76). The mean age of the groups were 12.79 (SD 2.86) years for MOON and 12.58 (SD 2.67) years for control. In total, 80% (61/76) of the sample was composed of male participants.

Regarding multimodal treatment, 30% (23/76) of the participants were receiving >1 medication per day. The remaining 70% (53/76) were taking only their ADHD pharmacological treatment. As much as possible, ADHD patients were instructed not to change their medication during the study, except for clinical reasons (eg, side effects or inadequate medication adjustment at the time of examination). For ethical reasons, the needs of the patients were prioritized, and there were 11 changes: 8 dose increases of the usual medication and 3 dose decreases of the usual medication. Of the abovementioned changes, 3 occurred in the MOON group (1 decrease and 2 dose increases) and 8 occurred in the control group (2 decreases and 6 dose increases). Currently, 41% (30/73) of the participants are receiving psychological treatment, and 30% (22/73) had received it before the study. In addition, all participants received weekly psychoeducation and counseling on ADHD.

School, social, and recreational data have been included in Table 3. Regarding difficulties with peers, 19% (14/76) of the participants experienced bullying and 31% (23/76) had difficulty making friends, as reported by parents. Additional clinical information was collected. Regarding medical history, 45% (34/76) of the mothers had difficulties during pregnancy. Regarding the developmental alterations of the participants, 26% (20/76) had delayed speech development, 15% (11/76) had slow gait development, and 24% (18/76) had delayed sphincter control. Of the participants, 35% (26/76) needed previous speech therapy support, 9% (7/76) needed psychomotor therapy, and 11% (8/76) needed early stimulation.

The results are listed in the primary and secondary outcome sections, ordered according to the hypotheses. Analysis was performed with the ITT population (69/76, 91%) and patients who completed 80% of treatment (57/76, 75%).

The main hypothesis of a 3- or 4-point drop in the primary end point of global SDQ was not achieved. We found no statistically significant differences in the primary variable (SDQ) in our pretest and posttest comparison (F_1,64=0.51; P=.47; η²=0.00; 1–β=0.11). The preassessment overall SDQ scores for the experimental group (mean 27.48, SD 5.64; 95% CI 25.45-29,51) and the control group (mean 26.61, SD 6.03; 95% CI 24.57-28.63) were higher than the post scores of the experimental group (mean 26.09, SD 5.74; 95% CI 24.06-28.12) and the post scores of the control group (mean 25.94, SD 5.93; 95% CI 23.90-27.97). Significant differences were found in the prosocial scale in favor of the control group (F_{1, 66}=3.73; P=.05; η²=0.05; 1–β=0.47). These results are shown in Table 4 and Multimedia Appendix 3.

In the most engaged subgroup (16 sessions), clinical trends of improvement on the SDQ of 2 points were observed but did not approach significance. This improvement trend is particularly notable in internalizing difficulties (F_{1, 53}=3.35; P=.07; η²=0.05; 1–β=0.43), peer problems (F_{1, 54}=2.28; P=.13; η²=0.04; 1–β=0.31) and emotional symptoms (F_{1, 54}=1.95; P=.16; η²=0.03; 1–β=0.27; Table 4; Multimedia Appendix 3). At the SDQ follow-up, parents in the MOON group reported that their children felt “a little better,” while parents in the control group reported their children felt “about the same.”

All participants completed the 10 face-to-face sessions. However, the number of sessions decreased when there was no direct face-to-face supervision by the investigators in the MOON web-based intervention. The results of the main variable (SDQ) of the 3 measures (pre-mid-term-post comparison) are shown in Table 5 and Multimedia Appendix 4. The change of modality (face-to-face and on the web) could have influenced adherence to treatment. The MOON group had an average of 16 sessions, with 3 participants having the minimum number of sessions (10) and 12 patients having the maximum number (20; Multimedia Appendix 5). Most participants in the MOON group (26/33, 79%) played the video game in VR, as they were aged >12 years, following PlayStation’s recommendations. A comparative statistical analysis was conducted within the MOON group, comparing participants who played the face-to-face VR version (26/33, 79%) with those who used the computer version for the entire treatment (7/33, 21%). We found that the VR subgroup improved most on emotional symptoms (F_{1, 31}=7.79; P=.004; η²=0.20; 1–β=0.77), internalizing (F_{1, 31}=4.18; P=.04; η²=0.11; 1–β=0.50) and, to a lesser extent, global SDQ (F_{1, 31}=2.99; P=.09; η²=0.08; 1–β=0.38).

The ANCOVA for the main hypothesis—which posited that the MOON group would exhibit improved emotional regulation, reflected by a 3- to 4-point decrease in global SDQ scores compared to the control group—was conducted with the number of sessions included as a covariate. The results approached significance (F_{1, 63}=3.25; P=.07; η²=0.04; 1–β=0.42; Figure 2; Table 6). Significant improvements were found in the MOON group in the variables of emotional symptoms (F_{1, 65}=4.47; P=.03; η²=0.06; 1–β=0.54) and internalizing problems (F_{1, 64}=6.57; P=.01; η²=0.09; 1–β=0.71).

In this prospective, single-center, randomized, unblinded pre-post evaluation study, our primary hypothesis of a 3-point decrease in the global SDQ score for emotional regulation after 3 months of MOON cognitive training in clinically stable, medicated children and adolescents with ADHD was not confirmed. However, we observed improvement trends in specific areas among patients who were more engaged with the MOON treatment. After conducting an ANCOVA with the number of sessions as a covariate, we found values approaching significance for the main hypothesis, with a 2-point drop in the global SDQ. This suggests that motivation may have played a key role in the improvement of emotional regulation in the participants. Given that motivation is a crucial factor in individuals with ADHD, it likely influenced these outcomes [21,24,31].

Considering the number of sessions as a covariate, significant differences in emotional symptoms and internalizing problems were found in favor of the experimental group. No behavioral improvement trends were observed. While externalizing difficulties are related to temperamental negative affect and anger dysregulation (bottom-up processes), internalizing difficulties suggest an overregulation of negative emotion through maladaptive cognitive strategies such as blaming and ruminating (top-down processes, more closely linked to executive functions) [48-50]. Participants with externalizing problems may have found the video game lacking in external rewards (eg, prizes), unlike those participants with greater intrinsic motivation (eg, self-improvement and sense of progress). This clinical trial had significantly more sessions than the previous one (20 vs 12) [29], increasing the need for novelty, change, and reinforcement. However, as participants noted in their feedback, the game did not seem to meet these reward-based needs effectively.

Our results also showed improvements in material organization, working memory, and inhibition in the 16-session MOON group. The improvement in working memory is in accordance with other investigations whose improvement in working memory occurred especially in those individuals who had a higher level of motivation and voluntary commitment to training [21,51] as well as the organization [52]. The ITT group also had an improvement close to significance. However, in the Corsi cubes (objective task of working memory), no significant differences were found; the MOON group had higher memory span than the control group, but this difference was very small. These results are consistent with other randomized controlled trials that showed significant improvements in ADHD symptoms and working memory but no improvements in other cognitive functions tested, such as planning or self-monitoring [53]. Regarding cognitive flexibility, significant differences were found in favor of the MOON group in set shifting on the CTMT-2 test, but it was not significant on the parent-reported flexibility subscale on the BRIEF-2 questionnaire.

Findings may point at a more powerful experience when using the VR experience. At least theoretically, VR may be more effective in tricking the brain than regular video games [28-30,47] suggesting a heightened sense of presence in the virtual environment and reduced perception of the real world, leading to fewer distractions [22]. In addition our findings may also just be a matter of age and brain maturity, perhaps related to the age of the VR users (aged >12 years years). Another putative explanation is that those who played on the computer experienced a more bugged and, consequently, more frustrating version of the game.

Furthermore, the use of MOON was well tolerated, as no clinically significant side effects were reported. There were no adverse events to report to the Spanish Agency of Medicines and Health Products. The VR intervention reported a mean dizziness score of 0.31 on the UKU scale, where 0 indicates no side effects and 1 indicates mild side effects, which did not interfere with patient performance. No significant differences were found in the GASA scale on video game addiction. In ADHD, this variable is especially important as there is a higher risk of addiction [54]. Serious video games versus commercial games allow controlling design variables about addiction [28]. Here, we found significant differences related to sleep in the subscale disorders of excessive somnolence in favor of the MOON group. As for the MOON group that played through VR, we also found that some of them remembered more of their dreams after the sessions, as assessed with the UKU scale. This phenomenon (although neither significant nor interfering) was consistent with the previous clinical trial [29]. While we do not consider this a concerning side effect, it is worth exploring the potential impact of VR on sleep.

This clinical trial has important limitations, particularly the smaller-than-planned sample size (n=152) that affected its statistical power [30]. Trends of improvement were observed in various areas, but a larger sample may have yielded statistically significant results. In addition, this investigation, such as the previous one, was not blinded [29]. The effect sizes are small, and we must interpret these results cautiously. The sessions were decreasing with the intervention with MOON on the web, and this suggests that the game was insufficiently developed, as mentioned by the users themselves. The reward system in the video game was only partially developed, which may have impacted motivation and caused the game to become repetitive. In addition, usability study would have been necessary before the clinical trial (as in the previous study [47] to refine the gameplay and correct bugs). Another limitation of the study was that hypothesis 4 could not be tested, and the corresponding analyses could not be performed, as academic grades were not collected consistently (only half of the parents provided this information).

Nonetheless, our results provide additional, albeit partial, support for incorporating serious video games into the multimodal treatment approach for ADHD. After conducting our second clinical trial, we considered that video games do not have to be for everyone. Heterogeneity in ADHD is substantial. While some patients found improvements related to cognitive training with MOON, others did not. Other training methods exist. To optimize the treatment of patients with ADHD, it is crucial to focus on individual interests and their strengths and difficulties to personalize the treatment to what the person needs. Video games can be another addition to complement multimodal treatment along with other interventions such as music training, physical exercise, regular interaction with nature, or learning a new language accompanied by a healthy peer support network and appropriate parenting. For example, to improve externalizing symptomatology, behavioral psychological therapy intervention strategies may be more effective, especially early in the preschool years [17]. Combining training with pharmacological treatment seems to enhance the benefits more than pharmacological treatment alone [23].

Serious video games combined with multimodal treatment can improve symptoms associated with ADHD. In this study, significant differences in inhibition, working memory, and material organization were observed in participants who were more engaged in the MOON treatment.