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 Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
Diabetes is a leading cause of years of life lost and accounts for approximately one-fourth of health care dollars spent in the United States. Many of these costs are related to poor medication adherence and lack of self-care behaviors and are thus preventable. Depression, which is more prevalent among people with diabetes than in the general population, predicts poorer management of one’s diabetes, whereas positive affect predicts engaging in more positive health behaviors. Consequently, interventions that improve depression and positive affect may also improve diabetes-related outcomes among people with diabetes. Although preliminary research on the impact of such interventions among people with diabetes is promising, these studies focused primarily on in-person interventions, have had small samples, and lack long-term follow-up.
This study aims to examine the short- and long-term effects of a digital therapeutic platform focused on mental health among adults with poorly managed type 2 diabetes and elevated levels of depression.
This is a randomized controlled trial in which adults with a type 2 diabetes diagnosis, elevated hemoglobin A1c (HbA1c) levels (≧7), and moderate to severe depressive symptoms will be randomly assigned to a positive emotion regulation skills intervention group or a sham digital intervention with only psychoeducational content. The study will take place over 14 months, including the 8-week intervention (or control) delivered via a digital therapeutic platform (Happify Health) and follow-up assessments at 3, 6, and 12 months postintervention. Throughout the intervention and for 1 week at each postintervention follow-up, participants will complete daily assessments of diabetes-related distress, diabetes regimen adherence, and mood. Our primary outcome, HbA1c, will be self-reported every 3 months throughout the study. Secondary and exploratory outcomes will be assessed at baseline; at 8 weeks; and at 3, 6, and 12 months postintervention.
Recruitment is expected to begin in June 2020. Participants will begin the study as they are recruited and will finish in waves. The final wave of data collection from the 8-week intervention is expected for winter 2020, with the completion of the 12-month follow-up in winter 2021.
Although previous research suggests that in-person psychological interventions have promising effects on both psychological and physical outcomes among adults with diabetes, digital interventions can be advantageous because they are easily scalable and reduce many barriers that prevent people from seeking treatment. This trial will provide important information about the effects of a digital mental health intervention among adults with type 2 diabetes, assessing both short- and long-term effects of this intervention on HbA1c, depressive symptoms, and other diabetes-specific outcomes. If successful, this may introduce a scalable intervention that would help reduce some of the preventable costs associated with diabetes.
ClinicalTrials.gov NCT04068805; https://clinicaltrials.gov/ct2/show/NCT04068805.
PRR1-10.2196/18578
Diabetes is one of the most common chronic health conditions worldwide [
The economic cost associated with diabetes is also increasing because of the increased prevalence and the increased cost per patient [
However, many of these costs are preventable, resulting from poor diabetes management. In 2014, 577,040 diabetes-related hospitalizations in the United States were preventable [
People living with a chronic illness often report that depression is a major barrier to managing their condition and, worse, often do not seek treatment for their depression due to perceived stigma [
Conversely, higher levels of positive affect are associated with reduced mortality risk [
Taken together, these findings suggest that better mental health among people with diabetes is associated with less diabetes-related distress and healthier lifestyles, which, in turn, predict better diabetes-related outcomes. Consequently, psychological interventions targeting mental health may lead people with diabetes to engage in healthier lifestyles, indirectly improving diabetes-related outcomes and reducing associated costs [
Research on the impact of mental health interventions applied specifically to people living with diabetes remains limited. Mindfulness and positive psychology (PP) interventions, which are effective in reducing depression and increasing well-being in the general population [
Such interventions appear to be particularly successful when patients have higher baseline levels of diabetes-related distress and when interventions are delivered in a group format, draw on mindfulness-based stress reduction (MBSR), and include home practice assignments [
Preliminary research also suggests that cognitive behavioral therapy (CBT) may be effective in treating depression among individuals with diabetes, although the effects on diabetes-specific outcomes are mixed, and may be limited to people with high baseline depression scores [
Although this research is promising, most of these studies have focused on the impact of in-person interventions. However, there are numerous barriers to in-person treatment that prevent people from seeking treatment [
Although previous research supports the effectiveness of internet-based interventions in physical health [
This study aims to examine the effectiveness of a digital therapeutics platform named
Activities within Happify Health were developed by identifying evidence-based tasks and interventions that were shown to be effective in at least two separate studies and in different samples [
Observational studies of Happify Health users demonstrate that usage over 8 weeks is associated with more than a 27% increase in positive emotions, and high-use participants see even greater improvement [
Therefore, in this study, our goal is to examine whether Happify Health use over the course of 8 weeks also helps to improve diabetes-specific outcomes. To do so, we plan to compare changes in HbA1c levels among adults with poorly managed type 2 diabetes and elevated depressive symptoms who have completed 8 weeks of activities on Happify Health or 8 weeks of a sham digital intervention. Secondary outcomes include depression, positive affect, and other diabetes-specific outcomes such as medication adherence, diabetes-related distress, and diabetes-related self-care activities. In addition, although many other studies (particularly mindfulness and PP interventions) did not examine long-term effects, we plan to explore the long-term effects of Happify Health use on both primary and secondary outcomes at 3, 6, and 12 months postintervention.
We plan to recruit participants by capitalizing on the existing process used to draw new users to Happify Health, including advertisements on Facebook and other social media sources. To attract individuals with type 2 diabetes specifically, targeted advertisements will also be posted on websites that connect potential participants with research studies and clinical trials (eg, Research Match) and on websites relevant to people living with type 2 diabetes.
Interested participants will be directed to a web-based survey to determine eligibility. The survey questions will include questions on age, location, previous Happify Health usage, diagnosed chronic illnesses, self-reported HbA1c level, the Patient Health Questionnaire (PHQ) [
Participants will qualify for the study if they are aged 18 years or older, currently living in the United States, have never used Happify Health before, and have a current diagnosis of type 2 diabetes. In addition, as we are targeting participants with poorly managed diabetes and recent recommendations are that below 7% is a reasonable glycemic goal for most adults with type 2 diabetes [
This study is an RCT (NCT04068805) with an initial target sample size of 400 participants (200 participants per condition). Recruitment will continue until 400 participants have successfully completed pretesting and have been randomized to condition.
Although previous RCTs using Happify Health had response rates ranging from 56% to 72% for an 8-week posttest [
Participants will receive three types of compensation throughout the study. After the baseline assessment (for which participants will not be compensated), participants will be compensated with Amazon gift cards valued at US $15 for completing each assessment. As separate compensation for completing daily assessments (as part of the 8-week intervention and during the 3-, 6-, and 12-month follow-ups), participants will earn US $1 for each daily assessment, for a possible total of US $56 for completing daily assessments during the intervention and US $28 for completing the week-long daily assessments at postintervention and during the 3-, 6-, and 12-month follow-ups. Finally, participants will be compensated with US $5 for obtaining and reporting each HbA1c recording after baseline. Thus, participants will be compensated with a total of US $164 if they complete all assessments at all waves of data collection.
All assessments will be administered via Happify Health, and participants who do not complete measures will be sent email reminders; however, participants may withdraw from the study or skip assessments at any time. To link participant data across assessments and other data collections, each participant will be assigned a unique study ID number; participants will otherwise remain anonymous throughout the study. In addition to the primary and secondary outcome measures described in the following sections, other exploratory measures will also be included that are not reported here.
Participants will self-report their HbA1c levels, a measure of the average blood glucose levels over the past 3 months, 5 times throughout the study: at baseline, week 12 (3 weeks postintervention), week 24 (3 weeks after the 3-month postintervention assessment), week 36 (3 weeks after the 6-month postintervention assessment), and week 57 (at the 12-month postintervention assessment). At each assessment, participants will also report the date when they received this HbA1c reading to verify that it falls in the correct time frame.
Diabetes distress scale [
Measures of Medication Adherence Scale [
Summary of diabetes self-care activities [
PHQ-9 [
Participants will also complete the positive affect subscale of the Patient-Reported Outcomes Measurement Information Systems [
We will assess daily levels of diabetes-related distress with a single item (ie, How would you rate your diabetes-related distress, on average, over the past 24 hours?) rated on a scale ranging from 1 (
We will assess participants’ daily regimen adherence using a single item (ie, Which of the following activities in your diabetes regimen did you complete over the past 24 hours?) where participants check any of the following options that apply to their situation: monitored blood sugar, ate according to healthy eating diet, engaged in physical activity, took diabetes medication as prescribed, or did not follow my diabetes regimen at all.
We will assess participants’ daily mood with a single item (ie, “How depressed did you feel, on average, over the past 24 hours?”) rated on a scale ranging from 1 (
Participation in this study will take place over approximately 14 months, including an 8-week intervention (or corresponding control) delivered via Happify Health as well as follow-up assessments at 3, 6, and 12 months postintervention. To ensure that HbA1c levels reported postintervention reflect average glucose levels only after starting the intervention, all participants will begin the study within 3 weeks of obtaining their most recent HbA1c levels. Participants who indicate that their most recent HbA1c levels were obtained before that when contacted by phone will wait to begin the study until they obtain their following HbA1c measurement and researchers confirm that they meet the inclusion criterion.
Once participants are ready to begin the study, they will be directed to complete 1 week of daily assessments without exposure to either condition. Participants who complete fewer than 5 of these assessments will be disqualified from the study. Those who complete at least five of the daily assessments during this qualifying period will be directed to download the mobile app and complete the regular onboarding questionnaire for Happify Health; participants will be randomly assigned to either the positive emotion regulation skills intervention group or the control group upon completing these questions. Following randomization, participants will be prompted to complete the baseline assessment.
Participants will then be instructed to begin using their assigned version of Happify Health. Although participants will not be given explicit instructions on how often they should use the platform, they will be encouraged to engage with the platform daily. Participants will receive daily push notifications on their mobile device to remind them to access the platform and answer daily questionnaires, and they will also receive weekly emails as part of the Happify Health platform to help keep them engaged with the program. In addition, to improve use rates and participant retention, research staff will call participants to inquire about problems when a participant has not engaged with the Happify Health platform at all for 1 week.
After 8 weeks, participants will be instructed to complete the postintervention assessment; this assessment will be identical to the baseline with the exception of HbA1c assessment, which will occur 3 weeks later via the Happify Health app. After completing this assessment, participants will receive an email instructing them to answer daily questions for 1 week and to remind them that we will contact them again in 3, 6, and 12 months for follow-up assessments and that they should continue to use their assigned version of Happify Health as they see fit. Participants will then receive emails instructing them to complete each of the 3-, 6-, and 12-month assessments, which will also be accompanied by 1 week of daily assessments at each follow-up period.
Schedule of activities for prescreen, intervention period, and follow-up assessments.
Assessments | Time | |||||||||||
|
Prescreen | Week 1: qualifying period | Week 2: baseline and first week of intervention | Weeks 3-9: intervention | Week 10 | Week 12 | Week 21: 3-month follow-up | Week 24 | Week 33: 6-month follow-up | Week 36 | Week 57: 12-month follow-up | |
|
||||||||||||
|
HbA1ca | Xb | N/Ac | X | N/A | NA/ | X | N/A | X | N/A | X | X |
|
||||||||||||
|
Diabetes distress scale | N/A | N/A | X | N/A | X | N/A | X | N/A | X | N/A | X |
|
Measures of Medication Adherence Scale | N/A | N/A | X | N/A | X | N/A | X | N/A | X | N/A | X |
|
Summary of diabetes self-care activities | N/A | N/A | X | N/A | X | N/A | X | N/A | X | N/A | X |
|
Patient Health Questionnaire | X | N/A | X | N/A | X | N/A | X | N/A | X | N/A | X |
|
Positive affect subscale | N/A | N/A | X | N/A | X | N/A | X | N/A | X | N/A | X |
|
Daily assessments | N/A | X | X | X | X | N/A | X | N/A | X | N/A | X |
aHbA1C: hemoglobin A1c.
bX: assessment administered.
cN/A: not applicable.
Participants assigned to the positive emotion regulation skills intervention will receive full access to the Happify Health platform. However, their version will feature a diabetes-related track focusing on building skills for greater happiness, reducing stress, and coping better with diabetes (
Screenshot of the featured track description for Developing Affective HeaLth to Improve Adherence intervention.
The featured diabetes-related track was designed by Happify Health and the second author, who is an expert on emotion and diabetes. The track is based on activities and content from the Developing Affective Health to Improve Adherence intervention originally developed by Moskowitz et al [
This part contains 8 different activities and exercises focusing on savoring. For example, in 1 activity,
This part contains 9 different activities and exercises focusing on reframing negative thoughts. For example, in
This part contains 11 different activities that build on parts 1 and 2 but includes exercises promoting goal orientation and optimistic thinking. For example, in
This part contains 11 different activities that promote resilience and helping others. For example, in
Participants must complete one part before they can begin the following part, and they do so by earning either a silver medal (earned by completing all but 4 activities) or a gold medal (earned by completing all but 3 activities). Thus, participants will take varying lengths of time to complete a track, depending on their level of engagement with the platform. As participants will have access to the full Happify Health platform, if they complete the intervention track before the 8-week study period has elapsed or choose to change tracks before finishing, they will have access to all other available Happify Health tracks and instant-play activities (where they can choose to complete certain activities outside of dedicated tracks).
Participants assigned to the control condition (
Screenshot of the track description in sham digital intervention.
We plan to analyze the changes in primary and secondary outcomes across the 2 conditions using hierarchical linear modeling (HLM). We will compute the change trajectories on each outcome over the course of the study assessments (ie, at baseline, immediately postintervention, and at 3-, 6-, and 12-month follow-ups). These trajectories will then be averaged together within each condition, so we can compare the trajectories for each outcome across the 2 conditions to determine if there was greater improvement in the positive emotion regulation skills intervention group relative to the control group.
We plan to use participants’ daily assessments in 2 ways. First, we will compute growth trajectories for each participant using their daily assessments to explore when people tend to experience improvements in diabetes-related distress, diabetes regimen adherence, and mood during the intervention. Second, as retrospective reports can sometimes lead to inflated or inaccurate information compared with daily assessments [
Previous research on the impact of health care interventions on people with poorly managed diabetes suggests that effects may be most effective among people with especially poor glycemic control or HbA1c levels above 9.5% [
This study was submitted and approved for ethical review by IntegReview (protocol HLS-07), an independent institutional review board.
Recruitment for the trial is expected to begin during the second quarter of 2020, and participants will begin the 8-week intervention as they are recruited and consent to participate in the study. Consequently, the first wave of data collection is expected to be complete approximately 8 weeks after recruitment, with an estimated intervention completion date for all participants in the fourth quarter. We plan to conduct data analysis for postintervention assessments once this initial data collection is complete, and these results may be shared at conferences or submitted for publication while we continue to collect data for the 3-, 6-, and 12-month follow-up assessments. Owing to the high cost associated with this research, we plan to conduct preliminary analyses once 20% of our target sample has completed the 8-week intervention. If we find no trends suggesting Happify Health use predicts primary outcomes, we plan to discontinue the study.
Recruitment, data monitoring for participant safety, data storage, protocol implementation, and research administration (participant follow-up, etc) will be under the purview of EB, GK, JS, and AP as employees of Happify Health. All data analyses will be conducted by JM and IK.
Previous research suggests that many of the costs associated with diabetes are preventable when individuals with a diabetes diagnosis manage their condition more effectively [
Previous research demonstrated that 8 weeks of Happify Health use improved subjective well-being among people with chronic physical conditions, including type 2 diabetes, at the same rate as people without these conditions [
This study also plans to include 3-, 6-, and 12-month follow-up assessments, providing important information about the longitudinal effects of this intervention. Longitudinal research on the impact of psychological interventions on people with diabetes remains limited [
Previous research has also been characterized by small samples, with sample sizes for in-person interventions ranging from 23 to 139 [
However, one limitation of this study is that we are targeting a specific group of people with poorly managed diabetes and elevated depressive symptoms. Furthermore, because participants will have to successfully complete a run-in period before starting the study, our participants are also likely to differ from those who do not participate in terms of motivation, conscientiousness, etc. Thus, it is unclear whether our findings will generalize to a broader population of individuals with type 2 diabetes.
Another limitation of this study is that the intervention does not specifically focus on diabetes-related content. That is, participants in the positive emotion regulation skills intervention group may choose to complete the featured track, which includes a description cueing them to think about the impact of negative emotions on their diabetes, but activities within this featured track do not refer specifically to diabetes or incorporate diabetes-specific behavioral strategies. Furthermore, participants are not required to complete, or even begin, this track and may choose from any other available tracks that do not refer to diabetes whatsoever. Consequently, our intervention is similar to mindfulness or PP interventions that are broader in focus [
Given the buffering effects of psychological well-being on diabetes outcomes [
cognitive behavioral therapy
hemoglobin A1c
hierarchical linear modeling
mindfulness-based stress reduction
Patient Health Questionnaire
positive psychology
randomized controlled trial
The authors would like to thank Allison L Williams for her assistance in planning this study.
EB contributed to designing the study and wrote the initial draft of this protocol. JM and AP contributed substantially to the study aims, scope, and design of the study. GK contributed substantially to the design of the study. JS contributed to designing the study and provided general manuscript support for this protocol. IK contributed to the study design.
EB, GK, JS, and AP are employees of Happify Health.