Using mHealth for Early Detection of Depression Relapse: Insights from Potential End Users
February 18th, 2025
Tina Coenen
Conquering my first conference
As is tradition, a considerable part of our imec-mict-ugent research group goes to ETMAAL every year. This 24 hours of communication science conference took place in the picturesque city of Bruges this year. EMUlab was represented by Aleksandra and me; we discussed our most recent research findings. Whereas Aleks is a seasoned expert when it comes to presenting and networking at conferences, I had yet to discover this side of academia.
In this blog post, I would like to share the first research that I presented on ETMAAL.
Background and relevance of this research
Depression is a tough battle, and for many, it’s not a one-time fight. People who have experienced depression are at high risk of relapse, making early detection crucial for better mental health outcomes. That’s where mobile health (mHealth) technology comes in. With real-time tracking and support, mHealth apps have the potential to catch early warning signs and help prevent relapses.
Our interdisciplinary research team is working on the DEDICAT project (“DEpression’s DIgital foreCAsting Tool”), to design and develop an mHealth app aimed at detecting the early signs of
depression relapse. To make sure that the app can truly help the people who need it most, we involved potential end users in the design process from the very beginning.
Listening to Potential End Users
To understand what users want (and don’t want) from a mental health app, we conducted three focus groups with 17 participants, all of whom had experienced at least one depressive episode. Before the discussions, they tracked their mood for a few days with our ecological momentary assessment (EMA) app. This experience gave them a firsthand look at how self-monitoring could work in practice.
During the focus groups, we explored their expectations, concerns, and must-have features for an ideal mHealth app. The app we’re developing uses a combination of active mood tracking (like EMA) and passive data collection from smartphones and smartwatches (such as screen time, location, movement, heart rate variability, etc.). A machine learning model then analyses this data to detect potential signs of relapse.
What They Want and Don’t Want in a Mental Health App
From these discussions, we identified five key themes that can shape the future of depression relapse monitoring apps:
Positivity in Design
Participants stressed the importance of a positive and uplifting interface. The app should not look – dare I say it – depressing. Colourful visuals, encouraging affirmations, and journaling features were seen as helpful tools for boosting well-being.
More Than Just Tracking—Actionable Support
Users don’t just want to track their mood—they want the app to help them take action. By combining mood data with context, they hoped to gain self-awareness and insights into their mental health. Visualizing their well-being metrics could help them recognize patterns, triggers, and coping strategies. Personalized advice, coping techniques, and intervention suggestions were also high on their wish list. The app, they felt, should be something they could rely on during difficult times.
Transparency Builds Trust
Privacy is a big concern when it comes to mental health apps. Users want to know exactly how their data is being collected, stored, and used. Clear explanations of how machine learning predictions work are crucial for building trust in the system.
Avoiding the Burden of Self-Monitoring
While self-monitoring can be helpful, it also comes with risks. Participants highlighted concerns about privacy burdens related to passive monitoring. The idea of their smartphone or smartwatch continuously collecting data about their behaviour made some feel uneasy. They wanted more control over what data was collected and how it was used, ensuring that their personal information remained secure and their boundaries were respected. Additionally, some participants feared that constantly tracking their mood might make them dwell on negative emotions, leading to overthinking or even worsening their mental state. They emphasized that feedback should be supportive, non-judgmental, and honest about its accuracy.
Customization is Key
Users expressed the need for flexible settings, including options for controlling data collection, choosing how often they complete self-assessments, and deciding whether or not they want to see their mental health feedback.
Designing for Real Impact
By incorporating these insights into the design of our mHealth app, we hope to create a tool that not only detects early signs of depression relapse but also provides meaningful support in a way that is both helpful and emotionally safe. The goal is to empower users, offering them control over their mental health journey while minimizing any potential downsides of digital self-monitoring.
As technology continues to evolve, so does its role in mental health care. By listening to real users and addressing their needs, we can ensure that mHealth solutions truly make a difference in people’s lives.
For those who want to know more, the full paper is available as a preprint via this link.