Do you remember how you felt a week ago? Chances are it’s tough to recall your emotional state from moment-to-moment—or how it changed over the day.
That’s a big challenge when it comes to accurate measuring symptom severity and treatment response in patients with major depressive disorder (MDD), says Paola Pedrelli, PhD, the Director of Dual Diagnoses Studies for the Depression Clinical & Research Program at Massachusetts General Hospital.
“Depression assessment is based so much on memory and reporting,” Pedrelli explains. “Distress is subjective—and experience is critical—but it’s hard to research depression and develop new medications when your evaluations are solely based on what people tell you.”
Could digital data provided by smartphones help? That’s the question being explored in the growing field of digital phenotyping—the process of using patient-provided smartphone and activity tracker data to monitor patient symptoms.
In depression, digital phenotyping studies have shown that a reduction in social interactions (shorter and less frequent calls and texts) and decreased activity levels can signal the onset of depression symptoms.
However, these studies have only been able to capture the presence or absence of symptoms, not their severity. By combining smartphone data with physiological data, Pedrelli is hoping to gain better insights into symptom severity and treatment response.
In a recent pilot and feasibility study led by Pedrelli and MIT’s Rosalind Picard, PhD, 31 participants with MDD agreed to wear two E4 Empatica wristbands for eight weeks in addition to providing data from their smartphones.
The wristbands measured electrodermal activity (changes in skin conduction in response to external stimuli), skin temperature, heart rate, motion and sleep, among other measures.
The severity of depression symptoms for each participant was assessed six times during the study using the 28-item Hamilton Depression Rating Scale (HDRS-28).
When the team compared the HDRS-28 assessments to the data from participants’ smartphones and wristbands, they found that–as expected—levels of activity and social interaction were strong indicators of depressive symptoms.
They also found that data on electrodermal activity and heart rate variability provided by the wristbands could be helpful in training machine learning models to measure symptom severity.
“It’s a good first step, but larger studies are needed to further evaluate this approach and inform more accurate models,” Pedrelli says.
She’s hopeful that in the future, data passively collected from smartphones and wristbands could prompt providers to offer quick and personalized support when patient symptoms worsen.
“I think that if we take away a little bit of the responsibility from the patient and doctor to do this continued calling and visiting, we may be able to deliver better treatment in a more efficient way.”
About the Mass General Research Institute
Massachusetts General Hospital is home to the largest hospital-based research program in the United States. Our researchers work side-by-side with physicians to develop innovative new ways to diagnose, treat and prevent disease.
Support our research
Leave a Comment