Smartphones sensors can detect major forms of psychopathology and could provide a promising way to identify early symptom deterioration and deliver “just-in-time” tailored interventions, research suggests.
The findings, in the journal JAMA Network Open, provide further evidence that passively collected information from these devices can relate to transdiagnostic dimensions of these mental disorders.
The study further suggests that the devices could one day be used as symptom monitoring tools and lead to more precise and effective treatment.
In an editorial accompanying the study, Massachusetts-based researchers Christian Webb, PhD, from McLean Hospital in Belmont, and Hadar Fisher, PhD, from Harvard Medical School in Boston, suggest that smartphones and wearables could provide low-burden support aligned to a person’s real-time state.
“Passive sensing via smartphones and wearables offers a powerful lens on individuals’ lives beyond the research laboratory and clinic walls,” they maintained.
“When used wisely, it may augment our understanding of psychopathology—capturing aspects of functioning that matter for mental health, enabling personalized monitoring, and potentially prompting earlier interventions.”
For example, smartphone sensors that detected a notable increase in time spent at home, decreased physical activity, and reduced initiation and response to communication could trigger a brief assessment of either depressive symptoms or a lack of appreciation for activities typically enjoyed.
The researchers note that current, in-person clinical observation provides only a narrow snapshot of a patient’s status, with lengthy patient-reported outcome measures only administered episodically.
Digital phenotyping using mobile sensing could address these limitations, collecting key information on symptoms and daily functioning that occur in the days and weeks between clinical encounters that is missed in current monitoring practices.
To investigate further, Whitney Ringwald, PhD, from the University of Minnesota at Minneapolis, and co-workers examined which forms of psychopathology relate to behavior that can be assessed through smartphone sensors.
The team continuously collected data from 557 adults using smartphone sensors for 15 days, following baseline assessments of psychopathology.
The study used six types of smartphone sensors: global positioning system (GPS); accelerometer; motion; screen on/off; battery; and call logs. These were used to derive 27 behavioral metrics.
Results showed that passively sensed behaviors were associated with all six major transdiagnostic psychopathology domains, with the strongest associations for detachment and somatoform symptoms.
Even after accounting for shared variance across domains, all but one domain—thought disorder—retained unique associations with smartphone-derived variables. For example, detachment was linked with lower physical mobility and disinhibition to lower telephone battery charge.
The p-factor, which reflects general impairment and is a general psychopathology dimension, was associated with a distinctive cluster of behaviors that included later bedtimes, reduced physical mobility, more time spent at home, and less telephone charge.
“These findings suggest that smartphone sensors can detect not only domain-specific behavioral patterns but also broad behavioral signatures of general psychopathology, with potential applications for monitoring and intervention,” the editorialists noted.
Ringwald and team concluded: “These results suggest that the findings from this study may advance research on day-to-day maintenance mechanisms of psychopathology and inform development of symptom monitoring tools.”