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 Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on https://formative.jmir.org, as well as this copyright and license information must be included.
The prevalence of neurodevelopmental and psychiatric diagnoses (NPDs) in youth is increasing, and unhealthy physical activity (PA), diet, screen time, and sleep habits contribute to the chronic disease disparities and behavioral challenges this population experiences.
This pilot study aims to adapt a proven exergaming and telehealth PA coaching intervention for typically developing youth with overweight or obesity; expand it to address diet, screen, and sleep behaviors; and then test its feasibility and acceptability, including PA engagement, among youth with NPDs.
Participants (N=23; mean age 15.1 years, SD 1.5; 17 males, 9 people of color) recruited in person from clinic and special education settings were randomized to the Adaptive GameSquad (AGS) intervention or wait-list control. The 10-week adapted intervention included 3 exergaming sessions per week and 6 real-time telehealth coaching sessions. The primary outcomes included feasibility (adherence to planned sessions), engagement (uptake and acceptability as reported on process questionnaires), and PA level (combined light, moderate, and vigorous as measured by accelerometer). Descriptive statistics summarized feasibility and engagement data, whereas paired, two-tailed t tests assessed group differences in pre-post PA.
Of the 6 coaching sessions, AGS participants (n=11; mean age 15.3 years, SD 1.2; 7 males, 4 people of color) completed an average of 5 (83%), averaging 81.2 minutes per week of exergaming. Of 9 participants who completed the exit questionnaire, 6 (67%) reported intention to continue, and 8 (89%) reported feeling that the coaching sessions were helpful. PA and sleep appeared to increase during the course of the intervention over baseline, video game use appeared to decrease, and pre-post intervention PA per day significantly decreased for the control (−58.8 min;
AGS shows promise in delivering a health behavior intervention remotely to youth with NPDs, but a full-scale efficacy trial with a larger sample size is needed to confirm this finding. On the basis of feedback from beta testers and intervention participants, the next steps should include reduced technology burden and increased exergame choice before efficacy testing.
ClinicalTrials.gov NCT03665415; https://clinicaltrials.gov/ct2/show/NCT03665415.
Mental health and the resultant adverse chronic disease consequences among youth are growing concerns in the United States. Recent estimates suggest that pediatric psychiatric disorders occur in more than one-fourth of people aged <18 years [
The health disparities faced by children and youth with heterogeneous neurodevelopmental and psychiatric diagnoses (NPDs) are considerable and include a higher risk of obesity and type 2 diabetes [
Unhealthy lifestyles among youth with NPDs are doubly unfortunate; they not only confer increased health risks across the life course but also exacerbate the challenges to cognitive and behavioral functioning experienced by this population. More than 25 published studies have documented associations between light-, moderate-, and vigorous-intensity PA and improvements to mood and executive functioning, such as the ability to focus and self-regulate, and meta-cognitive processes among children with NPDs [
Exergames have shown promise in promoting cost-effective engagement in light- to moderate-intensity PA in youth [
GameSquad is an intervention originally designed to improve PA among a socioeconomically and racially diverse population of typically developing children who meet the criteria for overweight or obesity [
Inclusion team science is an intervention development and testing framework that brings together disability researchers and intervention scientists to fast-track intervention adaptation for this underserved population [
All study procedures were approved by the Merrimack College, Pennington Biomedical Research Center, and Boston Medical Center Institutional Review Boards, and the study was registered as a clinical trial and is available on ClinicalTrials.gov (NCT03665415). As we adapted GameSquad to a population of youth with NPDs to create Adaptive GameSquad (AGS), we expanded the theoretical framework to include the Reserve Capacity Model [
Adaptation was undertaken by an advisory team including a developmental psychologist, school psychologist, registered dietitian, clinical social worker, fitness coaches, and parents of a child with NPD. As a result, we expanded caregiver support components, such as scheduling reminders; modified the original challenge booklet to increase MVPA goals more slowly; trained telehealth coaches in positive behavioral reinforcement techniques; and targeted exergames that were particularly engaging for the demographic profile of our target population. We also expanded coaching scripts to not only address PA but also to include health education and goal setting for dietary intake, sleep, and screen time habits. Coaches were also trained to work with youth with NPDs, particularly in positive reinforcement and de-escalation methods [
Finally, 3 beta testers and their caregivers were purposively recruited from the school study site for diversity of diagnosis, gender, and grade level (demographic information is not included here to protect them from inferred identification, given the small sample size). After written parental consent and child assent, the youth participated in a 4-week trial of the initial AGS intervention. We then used their feedback to make additional modifications to the telehealth coaching script and gaming menu, including increased agency in game selection and allowing for solitary game play. These modifications resulted in the final AGS intervention deployed for the pilot feasibility and engagement study.
Participants for the 10-week pilot study were recruited in person from October 15, 2018, to February 15, 2019, from either the Boston Medical Center Developmental or Behavioral Pediatrics Clinic from therapeutic programs at a large public middle and high school (
The study was powered as a pilot feasibility and acceptability trial [
Study recruitment and enrollment flow diagram.
Interested parents who met the criteria during the initial web-based screening then participated in an additional phone screening with research assistants and provided web-based written consent for their child to participate. Eligible children whose parents had consented were then assented in person during the initial data collection visit, during which height, weight, and BP, and psychosocial questionnaire data (see Measures section) were collected. Participants were given and instructed on the use of a hip-worn accelerometer (wGT3X-BT, ActiGraph). These accelerometers do not have digital readouts and thus minimize the effects of measurement on participants’ normal PA patterns. Participants were provided with the accelerometer to wear for 7 days before randomization and again the week after the 10-week intervention, with a follow-up visit scheduled to collect the accelerometer. After the initial wear, the participants were randomized to the condition using the randomization algorithm in REDCap. Participants assigned to the wait-list control condition were asked to maintain their normal level of PA for 10 weeks but did not receive additional information or a PA tracker during the 10 weeks. They then received the same gaming equipment as the intervention participants and 4-week of telehealth coaching after the final data collection. All data assessors and primary investigators were blinded to participant condition. Telehealth coaches and research assistants who installed equipment and training participants were not blinded.
All participants randomized to the intervention group received an Xbox One gaming console with a Kinect motion sensor (Microsoft), a 12-week Xbox Live subscription, and 3 exergames (Just Dance 3, Shape Up 3, and Kinect Sports Season 2). Two research assistants visited intervention participants in their homes to install the equipment and to train participants and caregivers on the use of the games, the Skype portal for coaching sessions, and the Fitbit for tracking of steps during the intervention. The AGS intervention asked participants to play exergames 3 days per week with a family member or friend, if possible (not required). On nonexergaming days, participants were asked to meet tailored and incrementally increasing minutes of PA (starting with 10 min in week 1 and increasing to as much as 40 min per day by week 4). The activities prescribed were laid out week by week in the intervention
Participants and caregivers were asked to meet with a telehealth coach over a video chat using the Xbox console. Meetings were held in weeks 1, 2, 4, 6, 8, and 10 and were rescheduled as necessary. Participants were asked to wear a Fitbit Charge throughout the 10-week intervention period. They and their caregivers received charging and sync reminders 3 times per week. Steps per day could then be wirelessly and automatically uploaded and reviewed by the telehealth coach. Coaches followed a script for meetings that focused on reviewing the week’s PA, praising progress, troubleshooting barriers, and then discussing a new healthy habit to try each week.
Feasibility and engagement measures were tracked by telehealth coaches in REDCap, including telehealth coaching session attendance (coach report), number of exergame sessions completed per week (participant log), minutes of exergaming per week (participant log), and steps per week (Fitbit Charge, Fitabase). We also administered exit surveys with participants and conducted semistructured interviews with beta testers and their caregivers to assess acceptability and elicit suggestions for improvement. Exit surveys were administered after all other follow-up data were collected. Two participants were unavailable for exit surveys; thus, there was a reduced sample size for acceptability measures only.
Participant weight was measured by trained research assistants in light clothing without shoes on a digital scale (Model 813, Seca), and height was measured using a clinical stadiometer (Model 217, Seca). Two measurements were taken for both weight and height, with the average of each used to calculate the BMI (kg/m2). After height and weight were recorded, appropriate cuff size was selected, and diastolic and systolic BP readings were taken, following the 2017 TRUE (International Consortium for Quality Research) Consortium guidelines [
Exercise stage of change (youth report) was measured using a pen and paper version of the Change of Stages of Exercise–University of Rhode Island Change Assessment, the third generation [
The sample size was determined on the basis of the aims and design (feasibility and engagement pilot study), with a target of 50 participants but a minimum of 20, based on the current guidelines [
Participant demographic information is presented in
Participant characteristics at baseline.
Characteristics | All participants (n=23) | Control group (n=12) | Intervention group (n=11) | Test statistic | ||||||||
|
|
|
|
z test | Chi-square ( |
|
||||||
Age (years), mean (SD) | 15.1 | 14.9 | 15.3 | −0.57 (20) | N/Ab | N/A | .57 | |||||
Sex (male), n (%) | 17 (74) | 10 (83) | 7 (64) | N/A | −1.58 | N/A | .12 | |||||
|
|
|
|
N/A | N/A | 1.2 (5) | .88 | |||||
|
Asian | 2 (9) | 1 (8) | 1 (9) |
|
|
|
|
||||
|
Black | 2 (9) | 1 (8) | 1 (9) |
|
|
|
|
||||
|
Hispanic | 3 (13) | 2 (17) | 1 (9) |
|
|
|
|
||||
|
White | 14 (60) | 7 (58) | 7 (64) |
|
|
|
|
||||
|
Declined | 2 (9) | 1 (8) | 1 (9) |
|
|
|
|
||||
Qualifies for free or reduced-price lunch, n | 8 (35) | 3 (25) | 5 (45) | N/A | N/A | 1.5 (1) | .23 | |||||
|
N/A | N/A | N/A | N/A | ||||||||
|
Autism spectrum disorder | 12 (52) | 7 (58) | 5 (45) |
|
|
|
|
||||
|
Attention-deficit/hyperactivity disorder | 13 (56) | 8 (67) | 5 (45) |
|
|
|
|
||||
|
Anxiety | 6 (26) | 5 (42) | 1 (9) |
|
|
|
|
||||
|
Depression | 6 (26) | 3 (25) | 3 (27) |
|
|
|
|
||||
|
Other | 4 (17) | 4 (33) | 0 (0) |
|
|
|
|
||||
Taking medication that causes weight gain, n (%) | 7 (30) | 5 (42) | 2 (18) | N/A | N/A | 5.5 (2) | .06 |
aTwo-tailed
bN/A: not applicable.
cMore than 50% of participants had multiple diagnoses.
Equipment was successfully installed and used by all intervention participants (n=11), who completed an average of 5 out of 6 possible coaching sessions (range 0-6). None of the participants were familiar with the 3 exergames used during the study before the start of the intervention. Including 1 participant who never exergamed or attended coaching sessions after the initial home visit and training session, the intervention group executed an average of 1.7 out of 3 planned exergame sessions per week, averaging 81.2 minutes per week (SD 18.9) of self-reported exergaming and averaging 3559 steps per day over the course of the intervention.
Planned session adherence trends.
Exercise engagement trends.
Considering the 11 participants, adherence to the coaching sessions exceeded 60% each week—from a high of 10 (91%) attending in weeks 1 and 2 to a low of 7 (64%) attending in week 8 (
The average daily MVPA measured by accelerometry declined between baseline and 1 week postintervention for both the control group (−9.3 min, SD 21.5;
Secondary outcomes included pre-post changes in self-reported exercise stage of change, sleep duration, video game use, mealtime behaviors, BMI, and BP. The exercise stage of change improved significantly for the intervention group (
The average daily minutes of sleep decreased by 12 minutes for the control group (
Youth with NPDs are at high risk of unhealthy lifestyle behaviors, including low PA levels [
The aim of this pilot study is to adapt and expand an existing, evidence-based exergaming and telehealth coaching intervention [
Although not as high as the original GameSquad intervention that targeted a younger population (10-12 years vs 12-17 years), compliance with planned exergaming was good (participants completed an average of 57% of planned exergame sessions), and attendance at coaching sessions was excellent (participants attended an average of 5 out of 6 coaching sessions), particularly given the unique barriers faced by participants with NPDs and the technological challenges noted by some caregivers.
Despite the short duration and small sample size that limited our power to detect intervention effects, the results indicated potential improvements to PA during the intervention over baseline and smaller declines in MVPA relative to control participants after the intervention had ended. Engagement in both exergame-based and nonscreen-based PA was good; however, it declined in weeks 9 and 10 of the pilot. When combined with a decline in MVPA in both the control and intervention groups after the intervention was completed, this may reflect a seasonality effect. Such an effect could have been because of the school year ending, discontinuation of physical education classes and school sports, or other external factors. However, the decline in duration of exergame sessions after week 7, combined with qualitative feedback from beta-testing participants, may suggest that the adolescent population in this study became bored with the limited menu of exergames available more quickly than the preadolescent population in the original GameSquad study. Alternatively, it may reflect frustration with or dislike of the increasing intensity demands of the exergaming menu as the intervention progressed. Further exploration of these factors is required to optimize long-term engagement.
Long-term implementation of AGS might reorient coaching to leverage the early weeks of exergame engagement into a greater emphasis on nonscreen-based PA modalities later in the intervention. Additional game choices and technologies should also be evaluated and included to maintain participants’ interest and better meet the needs of this diverse population. The coaching sessions appeared to fill a psychosocial need for participants, independent of exergaming. Compliance with and acceptability of coaching sessions were higher than those of exergaming, with several participants repeatedly rescheduling sessions to enable them to meet with their coaches around significant clinical events such as inpatient hospital stays.
Although the study was underpowered to evaluate secondary outcomes, participants reported increased sleep duration and decreased video game use time, despite the introduction of exergaming sessions. There were no improvements in problematic meal behaviors as a result of participation. Intervention participation also appeared to positively affect the exercise stage of change; although no treatment participants were in the active stage of exercise before the intervention, nearly 50% were in the active stage after the intervention conclusion. It is important to note, however, that although the exercise stage of change improved, the maintenance stage was not evaluated through long-term follow-up.
This pilot study has several limitations. The small sample size limits generalizability and decreased power to detect intervention effects; it also precludes stratified examination of differences in outcomes by subgroup. Although participants were randomized to condition, the sample size was small; thus, as models were not adjusted for potential covariates and confounders, readers need to take caution in interpreting the results. In addition, although the strength of this study was that pre- and post-PA were objectively measured using accelerometers, several other measures were self-reported, including duration of exergaming sessions, sleep duration, and video game use. The 10-week intervention design and lack of long-term follow-up prevented the evaluation of sustained engagement and effects of the intervention; this must be the primary aim of any full-scale efficacy study.
This study has several notable strengths. We were able to include youth with a variety of mental health and neurodevelopmental disorders, recruiting participants from both clinical and therapeutic school settings. We believe this heterogeneity improves external generalizability and eventual translation to a variety of clinical and community venues. Next steps should include the development of a mobile health (mHealth) app to seamlessly deliver a wider variety of exergames, telehealth coaching sessions, and parental and participant reminders and integrate mood, health habit, and PA tracking, while eliminating the technical barriers associated with the Xbox-based approach. This will also help minimize disruptions to upscaling caused by changes in commercially available gaming technologies. Additional research should also be conducted with youth with NPDs to evaluate personal mediators of health behaviors, exercise preferences, and barriers to engagement that may inform mHealth app design and intervention optimization.
This pilot study is an innovative example of inclusion team science, a term coined by Rimmer and Vanderbom [
attention-deficit/hyperactivity disorder
Adaptive GameSquad
autism spectrum disorder
behavioral inhibition bias
blood pressure
mobile health
Meals In Our Household Questionnaire
moderate- to vigorous-intensity physical activity
neurodevelopmental and psychiatric diagnosis
physical activity
Research Electronic Data Capture
Primary support for this study was provided by the Healthy Weight Research Network of the Maternal and Child Health Bureau, Health Resources and Services Administration (UA3MC25735). This study was partially supported by a Nutrition Obesity Research Center grant P30DK072476 entitled
None declared.