Effects of a diabetes-themed cartoon-based education on disease knowledge and physical activity among Japanese children: a preliminary randomized controlled trial

Abstract

We examined whether cartoon-based diabetes education could enhance diabetes knowledge and physical activity among children. During the first visit, participants completed a diabetes knowledge test; accelerometers, which were worn, were distributed. During the second visit, participants were randomly assigned to the cartoon distribution or lecture attendance group. At 6 mo post-intervention, the children received the diabetes knowledge test and accelerometers via mail, completed them at home, and returned them. Thirty children participated, with a 100% follow-up rate. Changes in diabetes knowledge test scores or physical activity levels showed no significant between-group differences; however, both groups exhibited significant within-group improvements. In the diabetes knowledge test, the cartoon distribution and lecture attendance groups improved by 11 (19 [17, 22] to 30 [23, 35], p = 0.003) and 10 (17 [14, 23] to 28 [26, 32], p = 0.007) points, respectively. Physical activity significantly increased, with the cartoon distribution and lecture attendance groups gaining 1,427 ± 2,980 (6190 ± 2529 to 7617 ± 3315, p = 0.047) and 1,371 ± 2,659 (6640 ± 3056 to 8011 ± 3046, p = 0.037) steps/d, respectively. This pilot intervention revealed that diabetes-themed cartoons and lecture attendance were similarly effectiveness in improving diabetes knowledge and physical activity.

Highlights

● Cartoon based educational intervention is as effective as lecture-based educational intervention.

● Diabetes education via cartoons and educational lectures increase physical activity.

● Cartoons aimed at disease awareness are associated with psychological stress.

Introduction

Type 1 diabetes mellitus (T1DM) is a common chronic disease in children resulting from the autoimmune destruction of pancreatic β-cells and causing absolute insulin deficiency. Global data indicate that T1DM accounts for over 85% of all diabetes cases among individuals aged < 20 yr (1). T1DM incidence exhibits an increasing pattern, with reports showing a trend toward onset at a younger age over the past two decades (2, 3). Similarly, recent studies have highlighted a substantial increase in the prevalence of type 2 diabetes mellitus (T2DM) among children and adolescents, with incidence rates increasing several-fold in the past 20 yr in the United States (4) and globally (5, 6). Thus, T1DM and T2DM in children remain significant societal concerns. Addressing misconceptions and fostering a comprehensive understanding of these diseases, including their pathophysiology and treatment, represent crucial public health challenges.

To improve the quality of life (QOL) of children with T1DM and T2DM, the patients and their families should acquire accurate knowledge and maintain strong motivation for self-management. Additionally, receiving appropriate support and understanding from schools, peers, and the broader society may enhance environmental QOL (7, 8). However, public awareness of pediatric diabetes remains limited. Common misconceptions include confusion between T1DM and T2DM, insufficient understanding of self-care behaviors such as insulin administration and carbohydrate management, and a lack of awareness regarding atypical diabetes forms such as maturity-onset diabetes of the young. Even children with diabetes usually experience uncertainty about their condition.

Given these challenges, educational initiatives targeting a broad audience, including the general pediatric population, are essential rather than restricting efforts for children with diabetes. In Japan, Amano et al. conducted a stroke awareness campaign (The Stroke Heroes Act FAST Campaign, hereafter referred to as “Act FAST”) targeting middle school students through cartoon-based flyers and booklets. Their findings demonstrated significant retention of disease-related knowledge among participants (9). Moreover, knowledge improvements were observed among parents, suggesting that targeted educational materials for children facilitate information dissemination at the household level (9). These results indicate that awareness can be achieved through the distribution of cartoon-based educational materials among children, comparable to that achieved through mass media campaigns. Furthermore, we previously conducted an intervention study by distributing newly developed diabetes cartoon materials to children with T1DM in a single-group study. While no significant changes were observed in disease knowledge, findings suggested a potential for improvements in exercise habits (10). However, because the previous study was limited to children with diabetes, whether these materials effectively fulfill their primary purpose remains unclear, raising awareness at the societal level.

In this study, we aimed to examine the effects of distributing the previously developed cartoon materials for diabetes education to the general pediatric population in the community. We compared their impact with a lecture-based disease education approach by assessing improvements in diabetes-related knowledge and modifications in physical activity behaviors. Additionally, to evaluate the potential psychological impact of the cartoon materials, we examined psychological stress experienced by participants, explored whether psychological stress is associated with changes in disease knowledge and physical activity, and analyzed the relationship between changes in disease knowledge and physical activity resulting from the intervention.

Patients and Methods

Participants

The inclusion criteria for children in this study were (1) age 8–15 yr, (2) residence in Ibaraki Prefecture, and (3) ability to continuously wear an accelerometer throughout the study period. The exclusion criteria included (1) severe diseases or injuries and (2) restriction from physical activities by a physician. Standard deviation (SD) scores of the study participants for height and body mass index (BMI) were assessed. Both indices were calculated using data from the 2000 School Health Statistics Annual Report of the Ministry of Education, Culture, Sports, Science and Technology. The recruitment process is described below. Participants were recruited through flyers detailing the study and announcements in local newspapers. The recruitment materials explicitly stated that participation or non-participation had no disadvantages.

The study’s purpose and procedures were explained verbally and in writing to the applicants and their parents. It was emphasized that participation was entirely voluntary; participants could withdraw their participation willingly without unfavorable consequences, and consent could be revoked at any stage of the study. In addition, if a child strongly wished to participate, enrollment was not permitted unless parental consent was obtained. No financial burden was placed on participants. As compensation, participants received a QUO card worth 4,000 yen, provided upon obtaining written consent. Participants who withdrew were not required to return the compensation. The primary outcome measure in this study was the change in disease knowledge assessed using a pediatric diabetes knowledge test. However, no studies have reported the use of this specific test. Therefore, the sample size was not determined based on effect size or statistical power and was rather set at 30 participants, which was considered feasible within the 1-mo study period.

Ethical approval and informed consent

All procedures followed were in accordance with the ethical standards of the responsible institutional committee and the Helsinki Declaration of 1964 and later versions. This study was approved by the Institute of Systems and Information Engineering, University of Tsukuba (Approval No.: 2024R861, Approval Date: May 9, 2024).

Informed consent was obtained from all participants (or their legal guardians) prior to study participation.

Research design

A parallel-group comparison trial design was adopted for this study. All participants, accompanied by their parents, visited the University of Tsukuba twice. During the first visit, the diabetes knowledge test was conducted and accelerometers were distributed (between July and August 2024). Participants started wearing the accelerometers immediately after the first visit and wore them until just before the next visit. The second visit occurred 2 wk later, during which participants were randomly assigned to one of two groups. The Cartoon Distribution (CD) group received cartoon-based educational materials on diabetes, which they read on-site. The Lecture Attendance (LA) group participated in an educational lecture on diabetes. A psychological stress questionnaire was distributed before and after the CD and LA. After 6 mo, the diabetes knowledge test materials and accelerometers were mailed to participants’ homes (between February and March 2025). The diabetes knowledge test was conducted at home, and the accelerometer was worn for 2 wk upon arrival. Subsequently, participants were instructed to return the completed diabetes knowledge test materials and the accelerometers. Therefore, all studies were of 28 wk duration. The deadline for responses was 1 wk after mailing; participants who failed to return the materials within this timeframe were considered dropouts. Additionally, the cartoon materials were gifted to the participants of the LA group at the end of the study period.

The study has been registered in a clinical trial database (UMIN000054761).

Diabetes knowledge test

The diabetes knowledge test used in this study was developed based on existing validated assessments (Supplementary data 1). Questions were selected and translated into Japanese with some corrections from the “A Brief Diabetes Knowledge Test” by J. T. Fitzgerald et al., which evaluates the knowledge of type 2 diabetes (11), and the “KAT-1 Diabetes Knowledge Scales” by A. O’Neill et al., designed for children with type 1 diabetes (12). The test comprised 45 multiple-choice questions categorized into nine sections: General Questions, Nutrition, Activity, Hypoglycemia, Sick Days and Ketones, Basic Management, Complications, Diabetes Pathophysiology, and Diabetes at School. Each question had three to five answer choices, and the estimated completion time was 15 min (45 points maximum). The test comprehensively assessed diabetes-related knowledge and included topics not covered in the manga and lecture materials. Fifteen out of the 45 test questions (33%) were based on the manga materials and 24 out of 45 questions (53%) were based on the content of the lecture materials. The difficulty level was set to evaluate basic diabetes knowledge. The development of this test was supervised and reviewed by a panel of three experts, including a specialist in metabolic medicine, a certified physical therapist specializing in metabolism, and a registered dietitian.

Psychological stress questionnaire

The psychological stress questionnaire was administered to evaluate the psychological stress and emotional fluctuations experienced by participants after receiving the cartoon materials or attending the lecture. The questionnaire comprised five items assessing psychological stress:

“How much do you feel happy right now?” (Interestingness)

“How much do you feel satisfied right now?” (Satisfaction)

“How much do you feel sad right now?” (Sadness)

“How much do you feel anxious right now?” (Anxiety)

“How much do you feel scared right now?” (Fear)

Each item was assessed using a visual analog scale (VAS) ranging from 0 to 100%. The questionnaire was administered before and after the CD and LA using a paper-based survey. The difference in scores before and after the intervention was used as a representative measure, defined as psychological stress. This questionnaire was originally developed for this study by a university Institute member who specializes in psychology.

Physical activity

Physical activity during the intervention was measured using an accelerometer (Mediwalk, Terumo Co., Tokyo, Japan). The accelerometer was distributed to all participants at the beginning of the study. Participants were instructed to wear the device on their waist (stored in a trouser pocket) for 2 wk, except during sleep, changing of clothes, bathing, swimming, and contact sports. For data aggregation, participants with valid recordings of at least 10 h/d on at least 2 weekdays and 1 weekend day were included in the analysis (13). Days with less than 10 h of valid data were classified as non-wear days (non-wear rate = number of non-wear days/14 d). The mean values were calculated without considering the specific day of the week. Data on the average number of steps, average moderate-to-vigorous physical activity (MVPA) time (three metabolic equivalents or more), and energy expenditure through exercise were collected, as previously described (14). Additionally, the principal investigator had confirmed the participants’ schools through their parents. The study plan and its significance were explained to the schools, and individual approval was obtained from each institution.

Diabetes-themed cartoons

The cartoon materials were jointly developed by the Institute of Medicine and the Institute of Art and Design at the University of Tsukuba. The overall storyline was designed by a physical therapist, and based on this draft, artist Naho Horiuchi was commissioned to illustrate and write the content. The materials contained two chapters, each focusing on type 1 and 2 diabetes. They provided a clear and accessible explanation of disease onset, symptoms, treatment methods, and significant considerations for daily life. The material had 56 pages, comprising 23 pages for Chapter 1 (Type 1 Diabetes), 20 pages for Chapter 2 (Type 2 Diabetes), and 13 pages for additional content, including the table of contents, character introductions, and commentary (geijutsu.tsukuba.ac.jp/~fumiaki/gluccie/gluccie_manga2.pdf.) The content was structured to be readable and understandable within approximately 20 min. The cover design, table of contents, and character introduction pages were created by the staff of the Institute of Art and Design using professional artistic techniques. Additionally, the materials were supervised by pediatric and metabolic medicine specialists to ensure medical accuracy. The development and production of these materials required approximately 300 h of work by two staff members, with costs allocated as 300,000 and 150,000 yen for labor and materials, respectively.

Lecture-based education

The lecture materials comprised 20 slides, structured in alignment with the content and objectives of the cartoon materials (Supplementary data 2). Actual case photos of complications (e.g., foot lesions) and symptoms (e.g., hypoglycemia) were used as a part of the lecture content, and medically supervised edited recordings of television programs were used as reference materials. The lecture was conducted by a physical therapist for approximately 20 min in a small group format, with seven to eight participants per session. Additionally, the handouts distributed during the lecture were given to the participants after the session.

Statistical analysis

For all evaluated factors, the normality of the data was assessed using the Shapiro–Wilk test. Appropriate statistical methods were selected based on each variable distribution. Variables that followed a normal distribution were presented as mean ± standard deviation, whereas those with a non-normal distribution were presented as medians with interquartile ranges (25th and 75th percentiles).

Comparisons between groups for participant characteristics before the intervention, including age, sex, height, weight, BMI, diabetes knowledge test scores, psychological stress questionnaire results, and physical activity during the observation period (step count and MVPA time), were conducted using independent t-tests for normally distributed continuous variables and Mann–Whitney U tests for non-normally distributed continuous variables. Categorical variables were compared using chi-square tests. To evaluate interventional changes in diabetes knowledge test scores and physical activity, as well as differences in psychological stress between the two groups, independent t-tests were used for normally distributed variables, and Mann–Whitney U tests were applied for non-normally distributed variables. Given the preliminary nature of this study, a sub-analysis was conducted for each of the nine diabetes knowledge test domains, applying the same statistical approach as the total diabetes knowledge test score analysis. For variables that followed a normal distribution, a two-way analysis of variance (ANOVA) was conducted to examine the main effects of group (CD group vs. LA group) and time (pre-intervention vs. post-intervention), as well as their interaction. Post hoc comparisons following ANOVA were adjusted for multiple testing using the Bonferroni correction. Furthermore, for the entire diabetes knowledge test, a three-way analysis of variance (Group × Scope × Time) including factors of in-scope and out-of-scope tests (content in cartoon materials and lecture materials) was conducted, and when normality was not met, the Friedman test or Aligned Rank Transform ANOVA was used to confirm the robustness of the results. In this analysis, we calculated the percentage score by normalizing each test score with the maximum points assigned to the respective section. Specifically, for the cartoon materials, the maximum scores were 15 points for the in-scope test and 30 points for the out-of-scope test, while for the lecture materials, the maximum scores were 24 points and 21 points, respectively. To explore the associations between psychological stress (change) and diabetes knowledge test scores (change) and physical activity (change), as well as the relationship between changes in diabetes knowledge test scores and physical activity (change) following the intervention, correlation analyses (adjusted factors: age) were performed. If the data exhibited normality, a partial correlation analysis was conducted. If normality was not observed, partial correlation analysis was conducted based on Spearman rank correlation.

All statistical analyses were performed using R software (version 4.3.2; R Core Team, Vienna; mixed-effects models were fitted with the lme4 package) and SPSS version 24.0 (IBM Japan). The significance level was set at 5%, and the trend level was set at 10%.

Results

Thirty participants were recruited for this study, achieving the planned sample size with a 100% participation rate. All participants completed the entire study process, implying that the final analysis included 30 children (18 males, 12 females) with a mean age of 11 ± 2 yr, resulting in a 100% follow-up rate (Fig. 1). During the initial verbal screening by the physical therapist, no participants met the exclusion criteria.

Fig. 1.

Flowchart of selection of study participants.

At baseline, the psychological stress questionnaire scores were as follows (median (interquartile range)): Interestingness 50 (46, 61), Satisfaction 48 (43, 55), Sadness 4 (1, 37), Anxiety 27 (6, 44), and Fear 2 (1, 28). The baseline diabetes knowledge test score was 19 (16, 22) points. For physical activity, the daily step count was 7,213 ± 2,899 for males and 5,218 ± 2,546 for females, while the MVPA times were 15.7 ± 7.5 and 10.6 ± 7.6 min/d for males and females, respectively. The non-wear rate of the accelerometer was 16 ± 19% at baseline and 12 ± 18% after the intervention. No significant differences were observed between the two groups concerning basic participant characteristics, diabetes knowledge test scores, psychological stress questionnaire results, or physical activity (Table 1).

Table 1. Characteristics of participants.

Table 2 shows the between-group comparisons of change for each factor (before and after intervention), as well as the within-group comparisons (pre-post). As diabetes knowledge test scores were not normally distributed, interaction effects could not be assessed, and intragroup comparisons were performed using Bonferroni correction. In the CD group, significant increases were observed in the total score (19 [17, 22] to 30 [23, 35], 11 points, p < 0.001) and in the General Questions, Nutrition, Activity, Hypoglycemia, Basic Management, and Diabetes Pathophysiology domains. In the LA group, significant increases were noted in the total score (17 [14, 23] to 28 [26, 32], 10 points, p < 0.001), along with General Questions, Nutrition, Activity, Hypoglycemia, and Diabetes at School. No significant between-group differences were observed, except for Basic Management. All physical activity variables were normally distributed. Within-group comparisons showed significant increases in step counts (CD group: 6190 ± 2529 to 7617 ± 3315, 1,427 ± 2,980 steps/d, p = 0.047; LA group: 6640 ± 3056 to 8011 ± 3046, 1,371 ± 2,659 steps/d, p = 0.037) and MVPA times (CD group: 13.6 ± 7.0 to 18.2 ± 8.0, 4.6 ± 9.2 min/d, p = 0.040; LA group: 13.7 ± 8.1 to 19.3 ± 9.6, 5.6 ± 11.0 min/d, p = 0.039). However, no significant between-group differences were observed, and no significant interaction effects (group × time) were detected. Because both the cartoon materials and the lectures (including the teaching materials) contained sections that did not explain the test content, sub-analyses were conducted (Supplementary Table 1). Since the data (percentage score) did not follow a normal distribution, rank transformation was applied prior to analysis and, a three-way analysis of variance (Group × Scope × Time) was performed, adding the factors for in-scope and out-of-scope tests. The results showed a significant time effect (pre- vs. post-intervention) (p < 0.001), indicating a significant change in scores before and after the intervention. However, neither the group difference (CD group vs. LA group, p = 0.139) nor the scope difference (In vs. Out, p = 0.302) was significant. Furthermore, the interactions for time × scope (p = 0.811), group × time (p = 0.147), scope × time (p = 0.353), and group × scope × time (p = 0.521) were not significant.

Table 2. Comparison before and after intervention and between groups (Diabetes knowledge test scores and physical ability) of participants.

Regarding changes in the psychological stress questionnaire, satisfaction was only significantly higher in the CD group than in the LA group (VAS: 4.0 vs. –7.5, p = 0.045) (Fig. 2). The relationship between changes in diabetes knowledge test scores and physical activity was analyzed using partial correlation analysis based on Spearman rank correlation, with step count serving as the representative measure of physical activity. A significant positive correlation was observed only in the CD group (r = 0.610, p = 0.015) (Fig. 3).

Fig. 2.

Comparison of the change in psychological stress questionnaire of diabetes test between the cartoon distribution group and the lecture attendance group. CD, Cartoon distribution group; LA, Lecture attendance group.

Fig. 3.

Relationship between change in physical activity and change in diabetes test score in cartoon distribution group and the lecture attendance group. †p < 0.01.

As the psychological stress questionnaire scores (change) were not normally distributed, their relationships with diabetes knowledge test scores (change) and physical activity (change) were analyzed using partial correlation analysis based on Spearman rank correlation. In the CD group, the following significant or trend-level correlations were observed: Diabetes knowledge test scores and Fear (r = –0.443, p = 0.098), Steps and Interestingness (r = 0.633, p = 0.011), Sadness (r = –0.459, p = 0.085), Fear (r = –0.657, p = 0.008), MVPA time and Interestingness (r = 0.466, p = 0.080) and Fear (r = –0.714, p = 0.003). In contrast, in the LA group, no significant or trend-level correlations were observed for any variables (Table 3).

Table 3. Relationship between psychological stress (change) and total scores of diabetes knowledge test (change) in physical activity (change).

Discussion

In this study, we investigated whether an educational intervention using cartoon-based materials could improve diabetes-related disease knowledge and physical activity among children living in the community. There were no significant differences in the changes in diabetes knowledge test scores between the CD and LA groups during the study period period, suggesting that the cartoon-based educational approach was not more effective in improving the disease knowledge compared with the lecture-based approach. However, significant increases in overall diabetes knowledge test scores were observed in both groups following the intervention. Additionally, of the 18 section-specific analyses (nine sections per group), significant improvements were seen in 11 sections within each group. These findings suggest that cartoon- and lecture-based educational interventions contribute to behavioral modification by enhancing diabetes-related disease knowledge among children in the community.

Educational intervention involving cartoons in this study can be considered a form of gamification. Gamification refers to a behavioral modification approach where individuals engage in activities driven by hedonic motivation, leading to improvements in academic performance, income outcomes, and health outcomes (15). When applied to disease education in healthcare settings, cartoon-based materials may be a promising tool for engaging children and enhancing knowledge retention (16,17,18). As tools for education, cartoons and traditional health information materials share similarities, but differ significantly from tools such as bullet-point lists or photo-based pamphlets, offering a more interactive and engaging experience for learners. Cartoons have been used cross-culturally for decades as a tool to convey accurate and relevant information on critical and technical topics. They have been effectively used in disease education for medical and nursing students (19) and in prenatal education programs for pregnant women (20), demonstrating enhanced comprehension and potential for behavioral modification. Previous studies have demonstrated that lecture-based disease education significantly enhances disease knowledge among medical students. For example, a single-group intervention study reported a significant increase in knowledge about infectious diseases following a lecture-based educational program (21). Additionally, a study targeting students aged 12–20 yr showed that school-based education, which incorporated role-playing and workshops, on sexually transmitted infections (STIs) caused an improvement in STI-related knowledge (22). Furthermore, a previous study suggested that short presentations are effective in lecture-based education (23). As the presentation used in this study lasted approximately 20 min, this concise format may have contributed to its effectiveness. Given this body of evidence, cartoon- and lecture-based educational approaches are effective in medical and health education contexts. This may explain why, in this study, both groups showed comparable improvements in disease knowledge, and no significant difference were observed between the two intervention methods.

A unique aspect of this study is that the evaluation of educational effects was conducted after a significant time gap following the intervention. As a long-term benefit, the improvement in knowledge test scores 6 mo after the intervention is unlikely to be solely due to the temporary distribution of manga or conduction of seminars at the start of the intervention. Furthermore, the sub-analysis based on the scope of the instructional materials showed no significant differences between in-scope and out-of-scope items for either teaching modality, suggesting that the long-term effects of a single educational resource alone may be limited. In other words, this suggests participants may have undergone behavioral changes, such as voluntary review, during the observation period. On the other hand, a previous study (9) reported that distributing cartoon-based flyers followed by repeated lectures and practical training resulted in knowledge retention even 3 mo after the intervention. In other words, while previous studies demonstrated a sustained effect due to the “provision of additional educational opportunities (optional),” this study is significant in that it suggests a sustained effect due to a different mechanism: “voluntary learning behavior.”

Previous studies have reported the average physical activity levels of upper elementary school children to be 9,925 steps (7,711–10,765 steps) and 33.9 min (27.7–40.2 min) of MVPA for boys and 9,238 steps (8,204–11,645 steps) and 26.8 min (21.2–32.4 min) of MVPA for girls (24). However, the baseline step counts and MVPA time of the children in this study were lower than the previously reported values. Furthermore, the 2020 World Health Organization Guidelines on Physical Activity and Sedentary Behavior (25) recommend that children and adolescents average at least 60 min of moderate-intensity physical activity (MVPA) per day, but none of the participants in this study met this standard. These findings suggest that the study population represented children with lower physical activity levels than their peers and needed improvement in that context. Given this background, the results of this study are noteworthy. In other words, both the CD and LA groups showed significant increases in step counts and MVPA time after the intervention, indicating that both forms of education impartment may contribute to improving physical activity levels. Notably, since no significant differences were observed between the two groups, it is believed that the comic materials and lectures had similar effects. These results suggest that educational interventions can be used to promote behavioral change in inactive children. Furthermore, a previous study using the same comic materials found that distributing the materials to children with type 1 diabetes improved their exercise habits (10). The findings of this study are consistent with those of the previous studies and support the possibility that comic materials may be useful for promoting healthy behavior in children, regardless of whether they have the disease or not.

While both interventions demonstrated similar effects, a significant difference was observed in the psychological stress experienced by participants. Satisfaction with the intervention was significantly higher in the CD group than in the LA group, suggesting that children perceived the cartoon-based intervention more positively. Additionally, regarding the relationship between psychological stress and the two outcome measures, the CD group showed a positive correlation between Steps, MVPA time and Interestingness. This implies that positively engaging with the cartoons enhances physical activity. Conversely, diabetes knowledge test scores, Step, MVPA time were negatively correlated with Sadness and Fear, indicating that a negative perception of the cartoons may hinder behavioral modification in knowledge acquisition and physical activity. These findings suggest that using cartoons as an educational tool helps visualize the process leading to behavioral modification. Higher satisfaction, greater enjoyment, and reduced fear associated with the cartoon materials could contribute to greater behavioral change. In contrast, in the LA group, no significant correlations were observed between psychological stress and the outcome measures. This suggests that lecture-based education had no direct relationship between psychological stress and subsequent behavioral modification.

An analysis of the relationship between changes in diabetes knowledge test scores and step count in both intervention groups revealed a positive correlation only in the CD group. This suggests that the cartoon-based intervention caused simultaneous behavioral modification in both indicators. The finding that diabetes education improved disease knowledge and equivalently impacted physical activity is significant. This effect was not observed in the LA group, indicating that the mechanisms underlying behavioral modification may differ structurally between cartoon- and lecture-based interventions.

Limitations

This study has some limitations. First, a scientifically determined sample size was not used. In future studies, an appropriate sample size should be calculated based on the findings of this study to ensure that sufficient participants are recruited to ensure accurate verification of the intervention effects. Second, a pure control group was not established. Although both outcome measures showed significant improvements following the intervention, the absence of a control group may have introduced bias. For diabetes knowledge test scores, participants were aware of the study schedule and may have engaged in a self-directed review before the post-test. In addition, because the same tests were administered at home before and after the study, it is possible that learning effects and actions such as referring to study materials during the test may have impacted to the study results. Regarding physical activity, the pre-intervention measurement period coincided with the summer season, potentially influencing participants’ outdoor activity levels owing to heat exposure. Notably, children’s physical activity levels are easily influenced by environmental factors such as holidays and events. To accurately assess intervention effects, future studies should include a non-intervention control group alongside the cartoon-based intervention group to facilitate inter-group comparisons. Third, the average correct answer rate on the diabetes knowledge test after intervention was only 64%, raising questions about whether this level fully reflects the effectiveness of the educational intervention. A previous study reported an average correct answer rate of 76% when administering the diabetes knowledge test to patients with type 1 diabetes and their parents (12). The low accuracy rate in this study compared with that reported previously suggests that understanding the disease may be difficult for healthy children without experience of diabetes, limiting the effectiveness of the education. Future approaches may be more sustainable and effective in achieving knowledge retention by combining repetitive education, gamification, interactive teaching materials, and parent/teacher involvement.

Conclusion

A preliminary intervention involving cartoon-based diabetes education and LA among community-dwelling children led to increased diabetes knowledge test scores and physical activity levels. For cartoon-based diabetes education, these improved scores were suggested to be influenced by the psychological stress they experienced while reading cartoon. However, future studies with adequately powered research designs are required to replicate these findings and confirm their generalizability.

Conflict of interests

The authors declare that they have no conflict of interest.