PI-004 - DIGITAL HEALTH BIOMARKERS FOR CHARACTERIZING DEPRESSION SEVERITY TRAJECTORIES DURING ANTIDEPRESSANT TREATMENT.

J. Marrero-Polanco¹, M. Chauhan², S. Chesak³, P. Croarkin⁴, W. Bobo⁵, A. Athreya⁴; ¹Mayo Clinic, Rochester, MN, USA, ²Mayo Clinic, Jacksonville, FL, USA, ³University of Minnesota, Rochester, MN, USA, ⁴Mayo Clinic, Rochester, MN, United States, ⁵Florida State University, Tallahassee, FL, USA.

Background: Digital health technologies such as wearables serve as non-invasive remote monitoring probes to estimate individual functioning by measuring sleep, heart rate, and activity levels. Digital biomarkers could serve as quantitative adjunct indicators of the longitudinal course of depressive symptoms in patients treated for major depressive disorder (MDD). This study hypothesized that digital biomarkers would associate with depression trajectories in a group of individuals taking antidepressants.

Methods: In a 12-month wellbeing study conducted at Mayo Clinic, 298 adult participants were issued a smartwatch with an expectation of ≥70% wear time. In addition to sociodemographic data and lifetime history of diagnosis of anxiety and mood disorder at baseline, quarterly assessments of depressive symptoms were measured using Center for Epidemiologic Studies Depression Scale – CES-D. Fifty-two out of 298 participants self-reported taking antidepressants. Of these, 40 reported previous MDD diagnoses, 4 had a CES-D score ≥ 16 (a cut-off defining active depression), and 8 had unspecified indications for antidepressant use but were included for exploratory purposes. Trajectories were identified based on the CES-D cutoff: recovery (depressed to non-depressed, n = 9), persistence (only depressed, n = 21), relapse (non-depressed to depressed, n = 8), and stability (only non-depressed, n = 14). Univariable logistic regression analyses examined relationships between recovery and persistence, relapse and stability, and persistence and stability.

Results: Lower average awake hours (OR = 3.8E-04 [2.3E-07, 0.63]) and lower stress levels (OR = 0.74 [0.59, 0.94]) were significantly associated with recovery from depression. Lower variations in awake hours (OR = 2.0E-03 [4.8E-06, 0.84]) were associated with relapse into depression. Higher average resting heart rate (OR = 1.13 [1.01, 1.27]) was associated with persistent depression.

Conclusion: These findings serve as a promising framework for establishing digital biomarkers for antidepressant response. Future work should be followed up with larger sampling of patients with defined antidepressant indications and ecological momentary assessments in smartwatches, as these holds promise for enhanced precision management of mood disorders such as MDD.