Can Your Smartphone Predict Mental Health Struggles? New Research Suggests It’s Possible

Data passively collected from smartphone sensors may soon offer clinicians a new way to identify and monitor mental health conditions, according to research led by the University of Pittsburgh. The study, published July 3 in the journal JAMA Network Open, demonstrates a correlation between sensor data – including GPS location, screen time, and call logs – and a range of mental health symptoms.

Beyond Traditional Assessments

Traditionally, mental health assessments rely heavily on self-reporting, which can be subject to recall bias and inaccuracies. “We’re not always the best reporters; we often forget things,” explains Colin E. Vize, assistant professor in the Department of Psychology at Pitt’s Kenneth P. Dietrich School of Arts and Sciences. “But with passive sensing, we might be able to collect data unobtrusively, as people are going about their daily lives, without having to ask a lot of questions.”

The research builds on previous work linking smartphone data to conditions like depression and post-traumatic stress disorder. However, this new study expands the scope, showing correlations with symptoms that aren’t specific to any single disorder. This is significant because many behaviors are associated with multiple mental health conditions, and individuals experience symptoms differently.

How the Study Worked

Researchers analyzed data from 557 participants in the Intensive Longitudinal Investigation of Alternative Diagnostic Dimensions (ILIADD) study, conducted in Pittsburgh in the spring of 2023. Participants shared self-assessment data and data from their smartphones, including:

• GPS data (time spent at home, maximum distance from home)
• Physical activity (walking, running, stationary time)
• Screen time
• Call logs (incoming and outgoing calls)
• Battery status
• Sleep time

Using a statistical analysis tool called Mplus, the team identified correlations between this sensor data and six broad symptom dimensions: internalizing, detachment, disinhibition, antagonism, thought disorder, and somatoform (unexplained physical symptoms). They also examined the “p-factor,” a shared underlying vulnerability across all mental health symptoms.

Key Findings: A Transdiagnostic Approach

The study found that sensor data correlated with both the six symptom dimensions and the p-factor. This suggests that smartphone data can provide insights into a person’s overall mental health state, even if their symptoms don’t neatly fit into a specific diagnostic category. “The disorder categories tend to not carve nature at its joints,” Vize said. “We can think more transdiagnostically, and that gives us a little more accurate picture of some of the symptoms that people are experiencing.”

The Future of Mental Healthcare?

Even as the findings are promising, researchers emphasize that this technology is not intended to replace human clinicians. “A lot of work in this area is focused on getting to the point where we can talk about, ‘How does this potentially enhance or supplement existing clinical care?’” Vize stated. “Because I definitely don’t think it can replace treatment. It would be more of an additional tool in the clinician’s toolbox.”

Currently, the data provides averages and doesn’t offer insights into individual mental health. Behavior is complex and varies widely. However, the potential for passively collected data to inform assessment and treatment is significant, offering the possibility of more personalized and proactive mental healthcare.