The sampling method that was selected was convenience sampling. In order to be considered for inclusion in this study, participants were required to meet the following criteria: first, they had to be between 12 and 16 years of age; second, they had to have been diagnosed with type 1 diabetes mellitus (T1DM) at least six months prior to the diagnosis; and third, they had to attend the pediatric endocrinology outpatient service at a hospital on a regular basis for medical follow-up.

The following exclusion criteria were established: a) diagnosis of cerebral palsy or epilepsy; b) presence of brain tumors; c) psychological diagnosis prior to the onset of the organic disease.

The study sample was recruited through DiabetesLATAM, a non-governmental organization and foundation headquartered in Panama with outreach across Latin America. The principal mission of this organization is to provide education and support to individuals diagnosed with type 1 diabetes and their families. Participant recruitment was conducted between June and November 2024.The flowchart of the intervention protocol followed in the research is shown below (see Figure 1 ).

To calculate the sample size, we conducted an a priori power analysis using the software G*Power 3.1 (34), for a repeated-measures ANOVA (within-subjects factor: time; 3 levels: T1, T2, T3), with α = .05 and power (1–β) = .95. We specified a medium effect size (ηp² = 0.06) following Cohen’s conventions (35), as adopting a medium effect size balances realism and clinical relevance, avoiding under- or overestimation of the intervention’s potential. We further assumed a correlation among repeated measures of r = .50 (consistent with psychosocial outcomes over short-term follow-ups). Under these assumptions, the required sample size was N = 46. Considering expected attrition in adolescent sample, we initially enrolled a larger number of participants (N = 84). Finally, as stated in the Consort Flow Diagram ( Figure 1 ), the final analyzable sample was N = 44.

The intervention was meticulously designed as follows: through the DiabetesLATAM association, information about the emoTICare project was disseminated to its partners and members so that, on a voluntary basis, the legal guardians of interested parents or adolescents could contact the research team to express their interest in participating in the study.

Following the receipt of information regarding the project by the adolescents’ guardians and the provision of their informed consent, as well as that of the minors under their care, a researcher disseminated the link to the assessment to the parents/legal guardians of the participants. The assessment was conducted online via Limesurvey, a complimentary platform designed for the creation and administration of surveys and questionnaires (T1). Therefore, the responsibility falls upon the parents/legal guardians to furnish the minors with the links to the three assessments and the emoTICare application. Initially, immediately after completing the informed consent online, a first evaluation was conducted for all participants (T1), covering the main areas that constitute the psycho-emotional state of the participants. Following this preliminary evaluation, the subjects were reassessed after six weeks and subsequently initiated the intervention, which spanned a duration of six weeks. After the participants completed the emoTICare serious game, a third assessment (T3) was carried out on all participating adolescents. Assessments T1 and T2 functioned as a control measure for all participants in relation to the variables evaluated. The present study exclusively examined alterations that occurred naturally over time without external intervention. In T3, conducted after the intervention, the benefits achieved in these variables as a result of the emoTICare intervention were analyzed. The complete enrollment and evaluation schedule can be found in Figure 2 .

emoTICare is a serious game that was developed with the specific purpose of serving adolescents who have been diagnosed with type 1 diabetes mellitus (T1DM). The model is grounded in the fundamental principles of the Disease Adjustment Model from an Integrative Perspective (DAMIP) (33), which integrates the main theoretical components of Livneh’s Integrative Model (36, 37), Antonovsky’s Salutogenic Model (38), Leventhal’s self-regulation model of illness (39), and the Health Belief Model (40, 41). This approach enables a multidimensional examination of the adaptation process to the disease, encompassing cognitive, emotional, behavioral, and contextual factors. The DAMIP model identifies five key areas when addressing proper adaptation to chronic disease in adolescence. The aforementioned domains encompass physical well-being, emotional well-being, cognitive coping, identity and social relationships, and support.

emoTICare is characterized by a graphic adventure narrative (for a more detailed exposition of the structure and content of the serious game, consult the protocol article by Martín-Ávila et al. (42), in which the user is tasked with completing six missions that are associated with a variety of scenarios and themes. These missions are designed to reinforce the key areas identified by the DAMIP model. Each mission corresponds to a pivotal domain encompassed by the theoretical framework and the intervention, namely: (a) psychoeducation concerning T1DM; (b) emotional awareness; (c) emotional regulation; (d) cognitive coping; (e) identity and self-concept; and (f) social skills and communication. The activities within the game combine interactive dynamics, psychoeducational techniques, and challenges linked to the everyday context of adolescents with T1DM. This combination fosters meaningful learning and the transfer of skills to their daily lives.

The implementation of the “emoTICare” intervention program in this pilot sample of Panamanian nationals followed a quasi-experimental, single group, pre-post design, that uses descriptive, correlational, and comparative methodology. This study adheres to the ethical guidelines of the 2013 Declaration of Helsinki of the World Medical Association and has been approved by the Human Research Ethics Committee of the University of Valencia (Reference 2023-PSILOG-3178945, UVINV_ETICA-3178945). Appropriate measures have been taken to ensure the complete confidentiality of participants’ data, in accordance with Organic Law 3/2018, of December 5, on the Protection of Personal Data (LOPD). This study and its protocol have been approved and registered as a clinical trial in the ClinicalTrials.gov PRS (Protocol Registration and Results System) (Reference - ID: emoTICare NCT06331429). In addition, copyright: © emoTICare registration number: UV-SW-202481R © University of Valencia & Polytechnic University of Valencia, 2024. All rights reserved.

In order to evaluate the different variables related to the performance of emoTICare in adolescents with T1DM, the following battery of computerized questionnaires was developed:

Statistical analyses were conducted using the SPSS program, version 28.0 for Windows. The statistical procedures employed to address the hypotheses were as follows:

As illustrated in Table 1 , a comparison of means among the varying measurement times (T1-T2-T3) was conducted to analyze the alterations in the aforementioned psychological variables. This analysis was conducted prior to and following the subjects’ receipt of benefits associated with the utilization of emoTICare. Following the loss of 29 subjects due to dropouts (a common occurrence in longitudinal studies, as previously outlined in preceding sections), the final sample size for this longitudinal analysis was 44.

Prior to conducting the analysis, it was imperative to ascertain that all variables satisfied the assumption of normality. However, upon examination, it was discovered that certain variables did not meet this criterion. Consequently, a decision was made to undertake a mixed analysis contingent upon the fulfillment of this criterion. Consequently, the table presents both the “F” statistic of the ANOVA test and the X2 statistic of Friedman’s nonparametric test. In both cases, the Bonferroni correction method has been employed for the post-hoc pairwise test, the objective of which is to ascertain the presence of significant changes between specific time points. The partial eta squared statistic (ηp2) was employed to calculate the effect size in the context of repeated measures ANOVA, while Kendall’s W was utilized to calculate the effect size of the Friedman test. As in the previous contrast, the non-normal variables are identified by the presence of the data of the average ranges, identified with the symbol//Ar//.

Among the results found, we highlight for its significance the statistically significant decrease in the perception of threat of disease between the three-time measurements. A subsequent examination of the post-hoc checks reveals that the observed change is statistically significant in both the time intervals between T1 and T3, as well as between T2 and T3. However, no statistically significant change was detected in the interval between T1 and T2, prior to the adolescents’ engagement with emoTICare. The effect size observed in this case can be defined as moderate/high. A positive trend in the results is evident when examining the changes presented in the other variables analyzed, suggesting that the benefits received after the application of emoTICare are significant.

A notable increase in scores was observed in factors such as resilience, which improved between T1 and T2, and between T2 and T3, with a bigger variation in this last time period. Concurrently, a decline in emotional symptomatology and hyperactivity (measured by SDQ) between T2 and T3 was documented, exhibiting moderate and large effect sizes respectively. This decline occurred concurrently with the gaming experience, indicating a positive effect of the gaming experience on the emotional state of adolescents.

With respect to self-concept, the findings suggest a positive trend across all subdimensions of the scale. The T3 scores of all subdimensions of the Garley self-concept scale, as well as the overall self-concept score, have been demonstrated to exceed the initial values. Some of these changes are bigger between T1 and T2 (physical self-concept, social self-concept) while others are bigger in the time period between T2 and T3 (Total Self-Concept, Control Capacity, Intellectual Self-Concept).

A similar upward trend was observed in adolescents’ global emotional awareness (measured by EAQ) following the emoTICare play period. This same positive increase occurred to a greater degree between T2 and T3 in the subdimensions of emotion distinction and non-concealment of emotions and in the total emotional awareness score. Furthermore, it is imperative to acknowledge the significant increase between T1 and T3 in the subdimension of “verbal exchange of emotions,” which was accompanied by a small effect size.

In the context of problem-solving, an enhancement in the positive orientation toward problems between T2 and T3, as well as the adaptive factor toward problems (i.e., the aggregate of positive orientation and rational coping with the occurrence of a problem) also between T2 and T3. Finally, it should be noted that there was a decrease in the degree to which adolescents exhibited a negative orientation in coping with their problems between T1 and T3.

Regarding the adolescents’ personal evaluation of their emotional state (0-I feel very bad; 10-I feel very good) during the emoTICare experience, it is evident that their emotional state has shown significant improvement over time, particularly following the experience itself ( Table 2 ). Consequently, the mean of the emotional state increased from 6.89 at T1 to 7.16 at T3, and the difficulties (i.e., difficulties present in their daily life; 0-I have no difficulties or problems; 10-I have many difficulties, my situation is very complicated) decreased from 4.8 to 3.3.

Additionally, the adolescents with T1DM rated emoTICare with an average of 7 and 6.7, respectively, on a scale ranging from 0 to 10, indicating the degree to which it has helped them cope with daily life challenges and improve their emotional state.

A thorough examination of the data reveals that adolescents perceive the sixth and fifth areas as the most beneficial, with the sixth area receiving the highest average score of approximately 7 out of 10. Notably, the eighth area also emerges as a frequent choice, with an average score of 8, suggesting a consistent preference among adolescents. In the subsequent section, the remaining points of results will be described and analyzed, with a focus on the profile of 73 participants who were assessed prior to the commencement of therapeutic work with emoTICare. This analysis will facilitate the identification of the challenges faced by this demographic, thereby substantiating the necessity for the implementation of programs such as emoTICare, in addition to the benefits previously outlined in this section of the results.

As illustrated in Table 3 , the initial sample comprised 73 adolescents diagnosed with T1DM, of whom 64.4% were female. A significant proportion of the sample had a history of psychological consultation, with 58.9% reporting at least one such visit. Furthermore, 86.3% had been hospitalized at least once, highlighting the substantial clinical burden associated with the disease. With respect to their physical comorbidities, asthma, allergies, and epilepsy accounted for 5.5% each, while 79.5% of the subjects did not suffer from any other chronic disease in addition to T1DM. With regard to the presence of psychological comorbidities, anxiety (9.6%) and ADHD (8.2%) were the most prevalent, although the majority of the sample (80.8%) did not exhibit any psychological disorder. With regard to the most prevalent treatment for T1DM, our observations indicated that injectable insulin was the predominant approach (91.8%), followed by dietary management (39.8%). The physical consequences derived from this disease that were most frequently reported by the participants are visual difficulties (38.4%) and pain in the extremities (20.5%).

As we can see in Table 4 , the mean age of the sample was 14.49 years (SD = 1.69), with a minimum of 12 years and a maximum of 17 years. The adolescents reported a mean time since diagnosis of the disease of 61.68 months (SD = 44.83), with a wide range between 6 and 180 months. The mean number of hospitalizations due to the disease was 2.18 (SD = 2.04), with a substantial range between the minimum and maximum, 1 and 15, respectively. The frequency of visits to the endocrinologist due to T1DM averaged 3.16 months (SD = 1.76), with a minimum of two weeks between visits and a maximum of 12 months.

Finally, on a scale ranging from 0 to 10, the adolescents reported a mean of 5.10 (SD = 2.77) for the degree of difficulty posed by their disease and a mean of 6.79 (SD = 2.76) for their overall emotional state.

In Table 5 , we can observe the variables strongly related to physical well-being and the perception of threat from their illness. Regarding the perceived threat of the disease, the participants presented a mean of 30.9 (SD = 9.69) on the BIPQ scale. A comparison of these results with those from other studies employing this instrument in a similar population (patients aged 9–16 years with T1DM) reveals that the former are considerably higher.

The mean observed in this sample exceeds the 80th percentile (29.40) of the score obtained in studies with similar patients, suggesting that adolescents with T1DM in this sample have a high perception of disease. The percentiles further elucidate that a mere 25% of the sample has a score that is equal to or lower than 24, a point that can be identified as the “average” scores of the scale.

In the context of physical well-being, the mean direct scores of the sample were 14.33 (SD = 4.38). Given the scale of the test, it can be concluded that these scores are intermediate/low. A consultation of the scale of this instrument reveals that the mean is approximately the 32nd percentile, thereby confirming that the scores obtained by the adolescents in the physical well-being variable are deficient to a certain extent.

Table 6 presents the profile of the relevant variables related to the contents of this area. With respect to psychological well-being, the mean of the direct scores obtained is 23.63 (SD = 6.64). When evaluated on the standardized scale, the score obtained falls close to the mean.

With respect to the emotional awareness of the sample, it was observed that the sum of the subdimensions had a mean of 62.33 (SD = 6.45), which indicated that adolescents with T1DM exhibited an intermediate emotional awareness. Two of the most salient subdimensions were “attention to others’ emotions” and “analysis of emotions,” with 50% of the sample scoring above 12, approaching the upper limit of the dimension. Conversely, the sample scored low on “not hiding emotions,” with 50% of the sample scoring below 9.

A thorough examination of the subdimensions of the SDQ questionnaire reveals that the scale on which they achieve the highest scores is “hyperactivity,” with a mean of 4.57 (SD = 2.04). Conversely, “behavioral problems” emerges as the subdimension where they attain the lowest scores, with a mean of 3.27 (SD = 2.08). An examination of Table 7 , which delineates the number of participants falling within each range specified by the instrument, reveals that the majority of participants are situated within the “normal” range across the subdimensions. However, when the overall scale score is considered, it becomes evident that 50.7% of the sample falls within the “borderline” and “clinically significant” categories with respect to their psychopathology and overall emotional adjustment. Concurrently, 41.1% of the participants exhibited “borderline” or “clinically significant” scores in the “behavioral problems” dimension, 37% in the “emotional symptomatology” dimension, and 30.2% in the “hyperactivity” subdimension.

This domain encompasses variables associated with decision-making and problem-solving capacity, as well as the ability to overcome the challenges that adolescents with T1DM may encounter. Table 8 presents the results obtained for both the problem-solving variable, measured with the SPSI, and the resilience variable, measured with the BIPQ questionnaire, of adolescents with T1DM.

The subdimensions of the SPSI scale indicate that adolescents have mean scores in both the “positive towards problems” (M = 10.06; SD = 4.76) and “negative towards problems” (M = 10.32; SD = 5.09) coping types, as well as the “rational towards problems” coping type (M = 10.06; SD = 4.56). Conversely, adolescents exhibited somewhat lower scores than the mean of the questionnaire in the types of coping characterized by avoidance (M = 8.42; SD = 4.14) and impulsivity (M = 7.98; SD = 4.26) when confronted with the occurrence of a problem.

This finding suggests that, while adolescents may engage in meticulous examination of the pros and cons prior to making significant decisions or attempting to resolve problems, a discrepancy exists in their approach to decision-making, as evidenced by the comparable utilization of both negative and positive orientations.

Regarding resilience, the adolescents obtained a mean of 22.20 (SD = 8.21), which could be identified as an intermediate score given the range of possible scores of the instrument, although it is also observed that 25% of the sample has scores below 18.50, a fact that reveals that certain participants have difficulties in adapting to adverse situations in an adaptive manner.

This section encompasses the variables that pertain to the self-concept, both physical and psychological, of the subjects. That is to say, the variables that indicate the perceptions of adolescents with T1DM regarding themselves and their own abilities. Table 9 below presents the profile of the adolescents in the self-concept variable.

The Garley self-concept questionnaire yields a set of scaled scores that facilitate the comparison of our sample’s scores with a reference score (see Table 9 ). As demonstrated in the results, the subdimension in which the adolescents demonstrate the highest mean scores is family self-concept (M = 28.90; SD = 5.50). A total of 67.1% of the sample exceeds the 50th percentile of the standardized scale in this domain. The subdimension in which they present more problems is personal self-concept (M = 27.42, SD = 4.32), where 75.3% of the sample is below the 50th percentile of the scale. Other dimensions that merit attention for their substandard scores include control capacity (M = 25.10, SD = 5.17) and social self-concept (M = 26.88, SD = 6.42), with 64.4% and 63% of the sample, respectively, falling below the 50th percentile of the standardized scale.

Finally, this section groups together the variables referring to the social skills and prosocial behavior of adolescents with T1DM, as well as their close relational framework, in which we can include their family relationship and school performance ( Table 10 ).

In consideration of the dimensionality of the CHASO questionnaire, it was determined that adolescents exhibit the lowest scores in the interaction domain, encompassing interactions with both acquaintances and individuals of a romantic or sexual interest. This observation is further substantiated by the observation that both categories exhibit comparable means, with a mean of 9.73 (SD = 4.87) in the context of peers and a mean of 9.49 (SD = 3.70) in the context of acquaintances of a romantic or sexual interest. The subjects demonstrated particularly strong aptitude in apologizing and declining requests, with a mean score of 14.84 and a standard deviation of 4.78. This was followed by the expression of positive emotions, which exhibited a mean score of 14.27 and a standard deviation of 4.78.

The mean (M) and standard deviation (SD) of the subdimensions of the social skills scale, as well as their overall score (M = 122.15; SD = 26.85), could be considered as somewhat below average when compared with other applications of the questionnaire, especially with regard to their ability to interact with people they find attractive, in which 25% of the sample presents low scores (scores below 6). The scores on the dimensions of the Kidscreen-27 questionnaire were found to be lower than the mean of the official scales of the instrument. Specifically, the mean score pertaining to the relationship with parents and relatives would be at the 30th percentile on the standardized scales of the instrument. This suggests that these adolescents may encounter difficulties in establishing boundaries and may not perceive their parents as providing sufficient autonomy (M = 24.15; SD = 5.82).

Table 11 presents the percentage and number of subjects who fall into each of the ranges established by the SDQ instrument for the variables of prosocial behavior and problems in relationships with peers. As demonstrated in the data, a significant proportion of the sample falls within the category of borderline or clinically significant scores for both variables (11% and 35.7%, respectively). However, the majority of these scores are within the “normal” range in these subdimensions.

Although we have distributed its different subdimensions throughout the areas, we consider that the health-related quality of life variable deserves to be considered globally, given that it is one of the main factors in assessing adaptation to the disease. In Table 12 we can observe the mean T scores (the previous ones were the direct scores of the instrument, without transformation) of each of the subscales and their comparison with the 50th percentile of the European reference scale, which will allow us to make a more precise interpretation of these.

A subsequent analysis of the scores obtained in the different subdimensions of quality of life evaluated with the Kidscreen questionnaire revealed that the majority of the sample is below the 50th percentile in all subdimensions. This finding was then compared with the standardized scales corresponding to their age group and sex. Specifically, the dimensions of “Psychological Well-being” and “Family Relationship” demonstrate the most significant impact, with 79.5% and 84.8% of the sample falling below the 50th percentile, respectively.

As illustrated in Table 13 , the primary correlations of the study were determined using Spearman’s Rho statistic, given that certain variables did not satisfy the assumption of normality.

The results of the correlation matrix in reveal a negative correlation between the presence of psychopathology and negative emotional adjustment (SDQ) and total self-concept (ρ = -.453, p <.01), resilience (ρ = -.358, p <.01), and psychological well-being (ρ = -.611, p <.01). The correlation is also negative with the variable health-related quality of life, which is the sum of all subdimensions of the Kidscreen-27 questionnaire (ρ = -.469, p <.01). This finding suggests that adolescents grappling with significant emotional challenges may also exhibit diminished self-confidence, weaker coping mechanisms, and a diminished perception of overall quality of life when confronted with life’s challenges.

Conversely, heightened perceived illness threat (BIP-Q) exhibited a strong correlation with the presence of psychopathology (ρ= .619, p <.01) and a diminished capacity for resilience (ρ= -.461, p <.01). These findings imply that the negative perception of T1DM exerts a direct influence on emotional well-being and the individual’s capacity to effectively manage its challenges. Individuals with a higher perception of illness also tend to adopt maladaptive coping styles in response to problems (ρ = .366, p <.01).

The findings revealed a significant positive correlation between total self-concept score and psychological well-being (rho = .631, p <.01), health-related quality of life (rho = .700, p <.01), and adaptive coping style toward problems (rho = .436, p <.01). Conversely, lower self-concept demonstrated a negative correlation with maladaptive coping style (rho = -.333, p <.01) and elevated perceived threat to illness (rho = -.422, p <.01).

No significant correlations were observed between participants’ age and their perception of illness threat, resilience, or social skills. However, a low, but not significant correlation was identified between age and health-related quality of life (ρ = -.103).

A comparative analysis was conducted of the scores on the primary psychological variables of the study, categorized according to the gender of the adolescents. The results of this analysis are presented in Table 14 . Due to the fact that certain variables did not satisfy the assumption of normality necessary for the Student’s t-test to be employed, a mixed analysis was deemed the optimal approach. Consequently, the variables that did not meet this assumption were subjected to analysis using the Mann-Whitney U test. The identification of these non-normal variables is facilitated by the data of average ranges, denoted by the symbol//Rp///.

A comparison of male and female subjects reveals that males exhibit higher scores in the following subdimensions related to quality of life: physical well-being (p <.01); school well-being (p <.05); and psychological well-being (p <.05). However, no significant differences were observed in the dimensions of family relationships and social support. The high effect size (d=.859) for the physical well-being variable and moderate effect size for the psychological well-being variables (d=.517) should be highlighted. With respect to the social skills subdimensions, significant differences were observed in the subdimensions of “public speaking” (p <.01) and “interacting with people they are attracted to” (p <.05), with high (d = .717) and moderate (r = .255) effect sizes, respectively.

With regard to emotional awareness, substantial disparities are evident in the subdimension of emotion analysis, wherein male subjects demonstrate significantly higher scores compared to female subjects (p <.05), exhibiting an effect size approaching moderate (d = -.248). A significant disparity emerges in the context of their approach to problem-solving, with male subjects exhibiting a pronounced inclination toward a positive orientation in problem-solving (p <.05). Conversely, boys exhibited a higher adaptive attitude toward problem-solving than girls (p <.05). The effect sizes for these differences are relatively moderate in the case of adaptive style (r=-.250) and positive problem orientation (r=-.246).

After thorough examination, it was determined that there were no substantial disparities observed in the total self-esteem score, the emotional symptomatology exhibited by adolescents, or the potential presence of psychopathologies.