This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mHealth and uHealth, is properly cited. The complete bibliographic information, a link to the original publication on https://mhealth.jmir.org/, as well as this copyright and license information must be included.
Adolescent health promotion is important in preventing risk behaviors and improving mental health. Health promotion during adolescence has been shown to contribute to the prevention of late onset of the mental health disease. However, scalable interventions have not been established yet.
This study was designed to test the efficacy of two adolescent health promotion interventions: a well-care visit (WCV) with a risk assessment interview and counseling and self-monitoring with a smartphone cognitive behavioral therapy (CBT) app. Our hypothesis was that participants who had received both WCV and the CBT app would have better outcomes than those who had received only WCV or those who had not received any intervention. We conducted a prospective multi-institutional randomized controlled trial.
Participants were 217 adolescents aged 13-18 years. They were randomly divided into two intervention groups (WCV group and WCV with CBT app group) and a nonintervention group. WCV comprised a standardized physical examination along with a structured interview and counseling for youth risk assessment, which was designed with reference to the Guideline for Health Supervision of Adolescents of Bright Futures. A smartphone-based CBT program was developed based on the CBT approach. The CBT app comprised a 1-week psychoeducation component and a 1-week self-monitoring component. During the CBT program, participants created several self-monitoring sheets based on the CBT model with five window panels: event, thoughts, feelings, body response, and actions. The primary outcome was the change in scores for depressive symptoms. Secondary outcomes included changes in scores for self-esteem, quality of life, self-monitoring, and an adolescent health promotion scale. These outcomes were evaluated at baseline and at 1, 2, and 4 months after baseline. The exploratory outcome was the presence of suicidal ideation during the observation period. Intervention effects were estimated using mixed effect models.
In total, 94% (204/217) of the participants completed the 4-month evaluation. Both intervention groups showed a significant effect in the form of reduced scores for depressive symptoms at 1 month in high school students; however, these effects were not observed at 2 and 4 months. The intervention effect was significantly more predominant in those scoring above cutoff for depressive symptoms. There was significantly less suicidal ideation in the intervention groups. As for secondary outcomes, there was significant increase in health promotion scale scores at the 4-month follow-up among junior high school students in the WCV group. Moreover, the CBT app was significantly effective in terms of obtaining self-monitoring skills and reducing depressive symptoms.
Although adolescent health promotion interventions may have short-term benefits, the frequency of WCV and further revision of the CBT app should be considered to evaluate long-term effectiveness.
University Hospital Medical Information Network Clinical Trials Registry UMIN 000036343; https://center6.umin.ac.jp/cgi-open-bin/ctr/ctr_view.cgi?recptno=R000041246
Care settings for young children and adolescents are changing from treating acute diseases to managing chronic diseases as vaccinations are developed and deployed and as medical treatment advances. With reductions in serious illnesses, physicians can place greater importance on lifestyle-related diseases, mental health disorders, and developmental behavioral disorders [
The American Academy of Pediatrics recommended the delivery of preventive services and anticipatory guidance for adolescents aged ≤21 years through annual primary care visits [
A universal intervention program based on cognitive behavioral therapy (CBT) using an internet-based setting has shown potential to prevent depressive symptoms among adolescents [
Both primary care visits and internet-based CBT programs may be beneficial in promoting adolescent health. Interestingly, these intervention procedures are delivered in completely opposite ways; the former is characterized by a face-to-face encounter, whereas the latter is based on self-help therapy without an interview. Face-to-face encounters have the advantage of providing health education securely and allow providers to respond to individual requirements, but have disadvantages in terms of cost and time. Although internet-based programs have various advantages such as accessibility to large groups, cost-effectiveness, and less labor, a major disadvantage is that they depend on user motivation. In this study, we conduct a randomized controlled trial with a well-care visit (WCV) combined with a risk assessment interview and counseling and self-monitoring using a smartphone CBT app to promote adolescent health. The primary outcome of the interventions is improvement in depressive symptoms, and the secondary outcome is increase in health promotion score. Our hypothesis is that participants who receive both WCV and the CBT app would have better outcomes than those who receive only WCV or those who did not receive any intervention.
We conducted a prospective multi-institutional randomized controlled trial involving 217 adolescents (aged 13-18 years) from Fukuoka, Saitama, and Okayama prefectures and Tokyo. The trial was registered in the University Hospital Medical Information Network Clinical Trials Registry (UMIN 000036343). Participants were enrolled in the trial and randomized into two intervention groups (WCV only or WCV with CBT app) or a nonintervention (control) group. Outcome data were collected at baseline, after the intervention (4 weeks), and at 2- and 4-month follow-ups.
Participant flow chart. CBT: cognitive behavioral therapy; WCV: well-care visit.
The design of this study and procedures for obtaining informed consent were approved by the Medical Ethics Committee of the Kurume University School of Medicine (#18138).
The principal investigator and coinvestigators in each prefecture explained the purpose of this study and the study design to each regional educational committee and school principals’ association. School principals who were interested in the study informed students in their school about the content of this study using a leaflet. The leaflet was developed by the principal investigator and covered the purpose of the study, study design, participant recruitment, and the URL for the study’s home page. Students who were interested in this study could talk with their parents about enrollment. To receive a detailed explanation of the study design, the student or their parents applied for an appointment with an investigator at the relevant research facility (hospital outpatient clinic) through the study’s home page. During these appointments, students and their parents received detailed information about the study and provided informed consent. A total of 240 students from 23 junior high schools and 25 high schools received appointments, and 217 (90.4%) of them agreed to participate in this study. The inclusion criteria were the following: (1) aged 13-18 years, (2) able to visit a research facility with their parent or caregiver twice to receive a WCV or for installation of the CBT program, and (3) had access to a smartphone or Wi-Fi network (smartphones were available to rent if a participant had no smartphone). The exclusion criterion was the presence of severe depressive symptoms or suicidal ideation. After the participants and parents signed the informed consent form, screening for severe depressive symptoms or suicidal ideation was performed using the 9-item Patient Health Questionnaire (PHQ-9) [
This study had two intervention groups (WCV group and WCV with CBT app group) and a nonintervention group. Participants in all groups were asked to complete a questionnaire that included several outcome measures at four time points: at baseline and at 1, 2, and 4 months after baseline. The participant flow chart is shown in
The WCV was designed with reference to the Guideline for Health Supervision of Adolescents of Bright Futures. The purpose of the WCV was to address the individual’s concerns or stressors, check social determinants of health, and provide anticipatory guidance through an interview. We used the Home, Education, Eating, Activities, Drugs, Sexuality, Suicide, Safety (HEEADSSS) framework to help structure the WCV interviews [
A smartphone-based CBT program for iPhones, named
Screenshots of the smartphone cognitive behavioral therapy app.
The primary outcome of this intervention was improvement in depressive symptoms. The Depression Self-Rating Scale for Children (DSRS-C), an 18-item self-report questionnaire that measures depressive symptoms, was used to measure depressive symptoms in this study [
The Adolescent Health Promotion Short Form (AHP-SF) is a self-administered instrument that was designed by Chen et al [
The Rosenberg Self-Esteem Scale (RSES) is the most recognized and widely used measure to quantify global positive and negative attitudes toward the self [
The Pediatric Quality of Life Inventory (PedsQL) is a brief measure of adolescents’ health-related quality of life [
We counted the number of participants presenting suicidal ideation on the PHQ-9 during the observation period in each group. We defined participants as having suicidal ideation if they scored 2 or 3 (“more than half” or “almost every day,” respectively) on item 9 of the PHQ-9 [
The Trait Emotional Intelligence Questionnaire–Adolescent Short Form (TEIQue-ASF) assesses how adolescents perceive their ability to deal with their emotions while communicating with others [
Participants’ use of the CBT app was confirmed using server data on the number of days they browsed the CBT app (Mugimaru) and the number of self-monitoring sheets they created during the CBT app intervention period.
The sample size for this study was calculated based on the results from previous studies that set depressive symptoms as the primary outcome [
To investigate the effect of the intervention on the primary outcome measure (depressive symptoms), data analysis was performed using 2 strategies. First, we assessed changes in the mean depressive scores from baseline to the 1-month evaluation and from baseline to the 4-month evaluation as the immediate effect and maintenance effect of the intervention, respectively. Changes were statistically compared among the WCV, WCV with CBT app, and nonintervention groups using mixed effect models, accounting for the within-participant serial correlation of repeated measures. We also investigated the intervention effect separately for junior high school students and high school students. Second, we examined the associations between baseline depressive symptoms and the intervention. Therefore, participants in both intervention groups were classified into a group with baseline DSRS-C score >16 and another group with baseline DSRS-C score ≤16. An analysis similar to that described above was used to examine the immediate and maintenance effects of the intervention. The nonintervention group was excluded from this assessment. For the secondary outcome measures (including AHP-SF, RSES, and PedsQL scores), the changes in each score from baseline to the 1-, 2-, and 4-month evaluations were compared among the 3 groups using mixed effect models.
Furthermore, to assess the effect of the CBT app on depressive scores, the association between the number of self-monitoring sheets created by participants on their smartphone and the changes in depressive scores at the 1-, 2-, and 4-month evaluations were investigated using correlation coefficients. Similarly, to clarify the self-monitoring efficacy of the CBT app, we assessed the changes in TEIQue-ASF scores from baseline to the 1-, 2-, and 4-month evaluations.
At every measurement point, the presence of suicidal ideation in participants was assessed using PHQ-9. The prevalence of suicidal ideation between the intervention groups and nonintervention group was compared using chi-square test.
As this study was an exploratory investigation of the proposed intervention, no adjustment was used in multiple comparisons. All data analyses were performed using SAS (version 9.4; SAS Institute Inc).
A total of 217 participants from 48 schools (23 junior high schools and 25 high schools) were enrolled in this study and randomized into the 3 groups. From the 217 participants, 6 (2.7%) participants were excluded owing to the presence of suicidal ideation. Thus, 97.2% (211/217) of the participants were included in our analyses (WCV group: 68/211, 32.2%; WCV with CBT app group: 71/211, 33.6%; and nonintervention group: 72/211, 34.1%). There were 37.9% (80/211) male participants and 62.1% (131/211) female participants, with 38.9% (82/211) of the participants from junior high school and 61.1% (129/211) of the participants from high school. During the follow-up period, 1.4% (3/211) of the participants (3/3, 100% women; 2/3, 67% from the WCV group; and 1/3, 33% from the WCV with CBT app group) canceled their study attendance, and 1.9% (4/211) of the participants (1/4, 25% men and 3/4, 75% women; 3/4, 75% from the WCV with CBT app group and 1/4, 25% from the nonintervention group) dropped out of the study without giving any reason. Consequently, the follow-up rate was 96.7% (204/211). The flow of participants is shown in
For all participants, the mean changes in DSRS-C scores from baseline to 1, 2, and 4 months did not significantly differ among the WCV group, WCV with CBT app group, and nonintervention group (
Continuous outcome scores from baseline to the follow-up period for each group.
Outcome measure and follow-up (months) | WCVa group (n=68) | WCV with CBTb app group (n=71) | Nonintervention group (n=72) | |||||||||
|
Participants, n (%) | Score, mean (SD) | Change in score, mean (SD) | Participants, n (%) | Score, mean (SD) | Change in score, mean (SD) | Participants, n (%) | Score, mean (SD) | Change in score, mean (SD) | |||
|
||||||||||||
|
0 | 67 (99) | 8.43 (5.51) | N/Ad | 70 (99) | 9.26 (6.46) | N/A | 72 (100) | 11.21 (5.97) | N/A | ||
|
1 | 66 (97) | 7.18 (5.34) | −1.18 (3.18) | 69 (97) | 8.12 (5.60) | −1.10 (4.11) | 72 (100) | 11.20 (6.03) | −0.01 (4.49) | ||
|
2 | 66 (97) | 8.12 (6.04) | −0.29 (4.05) | 68 (96) | 9.46 (6.85) | 0.25 (4.58) | 72 (100) | 10.54 (6.80) | −0.46 (3.98) | ||
|
4 | 66 (97) | 7.40 (6.10) | −1.08 (4.11) | 66 (93) | 9.14 (6.68) | −0.02 (4.37) | 70 (97) | 10.76 (6.81) | −0.17 (4.81) | ||
|
||||||||||||
|
0 | 67 (99) | 71.11 (13.38) | N/A | 70 (99) | 72.24 (12.43) | N/A | 72 (100) | 68.30 (15.30) | N/A | ||
|
1 | 66 (97) | 71.05 (14.73) | 0.34 (9.79) | 69 (97) | 73.14 (13.97) | 1.14 (8.72) | 72 (100) | 69.30 (14.59) | 0.94 (7.76) | ||
|
2 | 66 (97) | 74.02 (14.39) | 3.28 (9.85) | 68 (96) | 74.64 (13.54) | 1.97 (9.75) | 72 (100) | 69.55 (15.79) | 0.88 (8.32) | ||
|
4 | 66 (97) | 75.13 (15.61) | 3.93 (11.97) | 66 (93) | 75.94 (14.08) | 2.98 (9.63) | 70 (97) | 70.60 (16.71) | 2.61 (9.32) | ||
|
||||||||||||
|
0 | 67 (99) | 28.62 (5.90) | N/A | 70 (99) | 27.44 (5.56) | N/A | 72 (100) | 26.39 (6.48) | N/A | ||
|
1 | 66 (97) | 29.75 (5.58) | 1.03 (3.17) | 68 (96) | 28.54 (4.88) | 1.21 (2.70) | 72 (100) | 26.77 (6.39) | 0.32 (3.31) | ||
|
2 | 66 (97) | 29.29 (6.48) | 0.58 (3.22) | 68 (96) | 28.71 (5.95) | 1.17 (3.39) | 72 (100) | 26.89 (6.25) | 0.47 (3.34) | ||
|
4 | 66 (97) | 29.44 (6.22) | 0.77 (2.78) | 66 (93) | 28.64 (6.43) | 0.92 (3.64) | 70 (97) | 27.28 (5.90) | 0.91 (3.91) | ||
|
||||||||||||
|
0 | 67 (99) | 90.88 (11.99) | N/A | 70 (99) | 89.65 (11.40) | N/A | 72 (100) | 85.30 (13.13) | N/A | ||
|
1 | 66 (97) | 93.48 (9.18) | 2.60 (8.75) | 69 (97) | 90.61 (9.49) | 0.94 (7.19) | 72 (100) | 85.89 (13.27) | 0.54 (8.32) | ||
|
2 | 66 (97) | 91.17 (14.92) | 0.37 (10.17) | 68 (96) | 90.82 (10.46) | 0.98 (10.85) | 72 (100) | 86.82 (14.99) | 0.96 (12.15) | ||
|
4 | 66 (97) | 92.43 (14.08) | 1.37 (11.10) | 66 (93) | 90.41 (12.42) | 0.69 (12.59) | 70 (97) | 87.63 (13.83) | 2.31 (9.66) |
aWCV: well-care visit.
bCBT: cognitive behavioral therapy.
cDSRS-C: Depression Self-Rating Scale for Children.
dN/A: not applicable.
eAHP-SF: Adolescent Health Promotion Short Form.
fRSES: Rosenberg Self-Esteem Scale.
gPedsQL: Pediatric Quality of Life Inventory.
No immediate or maintenance effects by intervention was observed. However, in high school students, there were significant differences in the changes in DSRS-C scores from baseline to 1 month between the WCV group (mean −0.88, SD 3.16) and the nonintervention group (mean 0.90, SD 4.49) and between the WCV with CBT app group (mean −1.67, SD 3.80) and the nonintervention group (
Changes in Depression Self-Rating Scale for Children (DSRS-C) scores in the intervention and nonintervention groups during the follow-up period; a: There were significant differences in the changes in DSRS-C scores from baseline to 1 month between the WCV group and the nonintervention group (
Furthermore, an intervention effect was observed in the classification of participants who scored above the DSRS-C cutoff score (16 points). The mean DSRS-C score for participants scoring >16 in the intervention groups was significantly decreased at 1 month (mean 15.56, SD 5.68) and at 4 months (mean 17.63, SD 5.88) compared with the mean score at their first visit (mean 20.53, SD 3.79). However, no such differences were observed in participants with DSRS-C score ≤16 in the intervention groups (
Changes in Depression Self-Rating Scale for Children (DSRS-C) scores for participants in the intervention groups by the DSRS-C cutoff score; a: the mean DSRS-C score for participants scoring >16 in the intervention groups was significantly decreased at 1 month compared with the mean score at their first visit (
Regarding the effect of the CBT app on changes in the DSRS-C score, participants who created more self-monitoring sheets had better DSRS-C scores at the 4-month follow-up. The relationship between the number of self-monitoring sheets and the changes in DSRS-C score showed a significant negative correlation at 4 months (
Correlation between the number of self-monitoring sheets created by participants and the changes in Depression Self-Rating Scale for Children (DSRS-C) scores in the well-care visit with cognitive behavioral therapy app group. Significant negative correlation was observed between the changes in DSRS-C scores and number of self-monitoring sheets created by participants at 4 months.
For all participants, the mean changes in AHP-SF scores from baseline to 1, 2, and 4 months did not significantly differ among the 3 groups (
Changes in Adolescent Health Promotion Short Form (AHP-SF) scores in the intervention and nonintervention groups during the follow-up period; a: the changes in AHP-SF scores from baseline to 4 months were significantly different between the WCV group and the nonintervention group (
For all participants, the mean changes in RSES scores from baseline to 1, 2, and 4 months did not significantly differ among the 3 groups (
For all participants, there was no significant difference in the mean changes in PedsQL scores from baseline to 1, 2, and 4 months among the 3 groups (
A total of 5.5% (12/217) of the participants presented with suicidal ideation during the observation period. Of these 12 participants, 6 (50%) participants were identified before entry, 5 (42%) participants were identified in the nonintervention group, and 1 (8%) participant was identified in the WCV with CBT app group. This showed a significant intervention effect for the prevention of suicidal ideation (
Regarding the effect of the CBT app on changes in participants’ self-monitoring scores, the more participants created self-monitoring sheets, the better their self-monitoring scores were at the 1-month follow-up. The relationship between the number of self-monitoring sheets created by the participants and the changes in self-monitoring scores showed significant positive correlation (
Correlation between the number of self-monitoring sheets created by participants and the changes in self-monitoring scores in the well-care visit with cognitive behavioral therapy app group. Significant positive correlation was observed between changes in self-monitoring scores and number of self-monitoring sheets created by participants at the 1-month visit.
In this randomized controlled trial, we were unable to demonstrate an effect of universal intervention with either the WCV intervention or the WCV with CBT app intervention in terms of changes in adolescents’ depressive symptoms; however, both interventions showed a temporary effect in improving depressive symptoms in high school students. Furthermore, the effect was significant for individuals who scored above the cutoff point for depressive symptoms. In addition, the interventions significantly reduced suicidal ideation during the observation period. However, our hypothesis of obtaining better outcomes in the WCV with CBT app group was not supported.
Initially, the interventions showed a significant temporary effect of improving depressive symptoms in high school students irrespective of intervention type (WCV only or WCV with CBT app). However, the second WCV session showed no effect on depressive symptoms in either intervention group after 1 month. This result may reflect volunteer bias, whereby participants’ responses met the expectations of the researchers. As volunteers who participate in research studies are generally high-functioning individuals with higher willingness [
This study revealed significant effect in terms of reduced suicidal ideation in adolescents during the 4-month observation period in both the WCV group and the WCV with CBT app group. Although 7% (5/72) of the participants were identified as having suicidal ideation in the nonintervention group during the study period, only 0.8% (1/132) of the participants in the intervention groups was identified as having suicidal ideation. This result suggested that the intervention may potentially be effective, and the WCV with or WCV without the CBT app may be an effective means to prevent children from committing suicide. As adolescent suicide is a global mental health concern [
We developed a smartphone CBT app for adolescents, which contained psychoeducation and self-monitoring and was expected to improve their depressive symptoms. App users were coached to observe their own thoughts, feelings, body response, actions, and relationships relating to daily events by repeatedly creating monitoring sheets, and they could monitor their own mind and develop solutions for changing their cognitive processes throughout the sessions. The CBT app was significantly effective in terms of obtaining self-monitoring skills and reducing depressive symptoms, which was confirmed by the association between the number of self-monitoring sheets created and the changes in self-monitoring and depressive symptoms scores. This effect may have contributed to the suppression of the adolescents’ suicidal ideation similar to the WCV. An increasing number of mobile apps are available for adolescents with mental health problems, many of which are equipped with CBT programs. However, currently, there is insufficient research evidence to support the effectiveness of these apps for adolescents [
Another important finding in this study was the significant increase in health promotion scale scores at the 4-month follow-up in junior high school students in the WCV group. This indicated that they may have become interested in health promotion activities, such as nutrition, exercise, and stress management. Participants in the WCV group participated in risk assessment interviews and received counseling (HEEADSSS) twice (at baseline and 1 month after the first visit). Although several school-based interventions to promote adolescent health revealed both significant and nonsignificant effects in reducing health problems [
This study had some limitations that need to be addressed. First, the follow-up period in this study (4 months) was a relatively short observation period to draw conclusions about a universal intervention effect. Although there were significant differences in the prevalence of suicidal ideation between the intervention and nonintervention groups during the observation period, participants may have developed suicidal ideation after the observation period, even in the intervention groups. However, many adolescent intervention studies have used relatively short assessment durations (eg, 4-12 weeks) [
In conclusion, this study contributes by informing research directions to promote adolescent health. A standard interview framework for adolescent health promotion (ie, HEEADSSS) may be applied in primary care settings in Japan to improve adolescents’ mental health, as there are no screening and intervention systems in the regular school-based health checkup. To minimize the time required for this screening, development of either a short form of the HEEADSSS or electronic HEEADSSS screening may be required. Furthermore, our CBT app, which uses a mobile device, may emerge as a new health promotion tool for adolescents if more CBT modules are added. Integrating direct and indirect interventions (HEEADSSS and CBT apps, respectively) may further promote adolescent health.
CONSORT e-HEALTH checklist (V 1.6.1).
Adolescent Health Promotion Short Form
cognitive behavioral therapy
Depression Self-Rating Scale for Children
Home, Education, Eating, Activities, Drugs, Sexuality, Suicide, Safety
major depression disorder
Pediatric Quality of Life Inventory
9-item Patient Health Questionnaire
Rosenberg Self-Esteem Scale
Trait Emotional Intelligence Questionnaire–Adolescent Short Form
well-care visit
This study was supported by grants from the Japan Agency for Medical Research and Development (20gk0110028h0003 and 21gk0110055h0001) and Ministry of Health, Labour and Welfare–Health Research on Children, Youth and Families Program (grant number 21DA1001). The authors would like to thank all those who helped them, especially the Board of Education, principals of the participating schools, and the contributing students. The authors also thank Edanz for editing a draft of this manuscript.
SN, MH, KS, RS, AO, and KM participated in the design of this study, and SN compiled the manuscript. SN, KS, RS, and AO were the representatives in each research facility and recruited the participants. MH, A Kanie, MI, A Katayanagi, and T Katayama designed the cognitive behavioral therapy app. T Kakuma conducted the statistical analyses. All authors except MH, A Kanie, MI, A Katayanagi, T Katayama, and T Kakuma examined and interviewed the participants using the Home, Education, Eating, Activities, Drugs, Sexuality, Suicide, Safety method. All authors read and approved the manuscript. YY supervised the preparation of the manuscript.
MI received royalties from several publishing companies for books related to cognitive behavioral therapy. He also received honorariums for workshops and supervisions regarding cognitive behavioral therapy.