Pilot ML Program Cuts Depression Scores by 3.5 Points in 40 Patients, Improving Anxiety and Cognition
Updated
Updated · Nature.com · May 19
Pilot ML Program Cuts Depression Scores by 3.5 Points in 40 Patients, Improving Anxiety and Cognition
5 articles · Updated · Nature.com · May 19
Forty completers in the PerMA pilot trial saw self-rated PHQ-9 depression scores fall by 3.5 points after a six-week personalized lifestyle program, with benefits still present at 12-week follow-up.
The intervention used 2-4 weeks of smartphone surveys and smartwatch tracking to build N-of-1 machine-learning models, then assigned each participant a targeted plan in sleep, exercise, diet or social connection plus weekly coaching.
Anxiety also dropped significantly, clinician-rated depression improved, and objective gains appeared in selective attention, interference processing and working memory; 22 of 40 participants reached remission.
EMA data linked mood improvement specifically to gains in each person's targeted lifestyle domain rather than off-target changes, supporting the intervention's personalized mechanism.
The single-arm, open-label study enrolled 50 people with mild-to-moderate depression, and researchers said the results justify a larger randomized controlled trial; automated assignment matched human coaches in up to 95% of cases.
Can AI's personalized health plans truly outperform generic lifestyle coaching?
With AI nearly doubling remission rates, what is the real timeline for public access?
As AI learns to manage our moods, who will protect our most private data?
Personalized Mood Augmentation (PerMA) Trial Delivers 55% Remission Rate in Depression: Implications for AI and Digital Health
Overview
The PerMA trial, led by researchers at the University of California San Diego, marks a breakthrough in depression treatment by introducing a highly individualized approach. Using data from consumer smartwatches and real-time logs, the team developed personalized mood augmentation plans (iMAPs) that resulted in a 55% depression remission rate and significant reductions in anxiety. These promising outcomes highlight the effectiveness of tailored mental health care and set the stage for future automation, offering a new framework for delivering remote, personalized support to individuals struggling with mental health challenges.