summary: A pioneering and first-of-its-kind study demonstrates that a personalized machine learning-powered lifestyle coaching program can nearly double remission rates for mild to moderate depression. This study tracks how individual behavioral factors uniquely predict depressed mood states.
By building an Individualized Mood Enhancement Plan (iMAP) using consumer smartwatches and real-time data logging, researchers achieved a 55% depression remission rate and significant reductions in anxiety, providing a highly effective framework for personalized telemental health care.
important facts
- Iteration failure: More than 21% of U.S. adults have depression. Standard clinical guidelines recommend general adjustments to sleep, exercise, and diet, but these one-size-fits-all recommendations are often overwhelming for people with depression and fail due to large individual differences.
- 2 week biometric audit: To build a data-driven baseline, 50 adults with mild to moderate depression wore smartwatches that tracked their heart rate and body movements while recording local daily indicators of sleep quality, diet, and social interactions up to four times a day.
- iMAP strategy: The Neural Engineering and Translation Laboratory (NEATLabs) at the University of California, San Diego ran this personalized data through a machine learning model to isolate the main lifestyle triggers that led to depression in each participant. The health coach then combined these insights with customized evidence-based behavioral treatments to create an individualized mood enhancement plan (iMAP).
- Double clinical benchmarks: Standard behavioral interventions have an average remission rate of only 30%. In contrast, the algorithmic iMAP approach Remission rate 55%This means that more than half of the cohort no longer met clinical criteria for depression after 6 weeks.
- Peripheral Health Dividend: As a result of a 6-week remote video coaching program that not only cures local symptoms of depression, 36% reduction in anxiety symptomssignificantly improved self-reported quality of life and improved scores on simple memory and attention tests.
- Lasting therapeutic effect: A follow-up audit confirmed that the cognitive and psychological benefits gained during the 6-week training block persisted for a full 3 months after the active intervention formally ended.
sauce: UCSD
More than 21% of adults in the United States experience depression, which significantly impacts their quality of life. Dr. Jyoti Mishra, associate professor of psychiatry at the University of California, San Diego School of Medicine, says many people with mild to moderate depression can improve their symptoms by making adjustments to their daily habits, including sleep, exercise, diet and social interactions.
However, depression varies greatly from person to person, so a one-size-fits-all lifestyle approach is not very effective.
In a first-of-its-kind study, Mishra and her team developed a machine learning-powered lifestyle coaching program based on data about participants’ moods and daily habits collected via personal devices.
They found that participants who implemented the program had significantly reduced symptoms of depression after six weeks. The results of this study provide a promising approach for remotely delivering personalized depression treatment tailored to each individual’s circumstances.
This study NPP – Digital Psychiatry and Neuroscience.
For two weeks, 50 adults with mild to moderate depression wore smartwatches that tracked their heart rates and exercise levels. They also tracked their mood and answered up to four short questions per day about their sleep quality, diet, activity level, and how often they talked to friends and family.
The team developed a machine learning model specific to each participant based on this data to discover which lifestyle factors best predicted an individual’s depressed mood. Each participant then worked with a health coach to implement an Individualized Mood Enhancement Plan (iMAP).
“Our goal was to understand the main lifestyle factors that cause depression, which vary from person to person, and whether targeting those factors through personalized coaching could actually make people feel better,” said Mishra, co-director of the Neural Engineering and Translation Labs (NEATLabs) at the University of California, San Diego.
Over the next 6 weeks, participants worked with their coach to implement iMAP.
“Each person in the trial received a different behavioral treatment that has already been established in the literature depending on key predictors,” Mishra said. “Some were working on cognitive behavioral therapy programs for insomnia, while others were implementing diet-based interventions to maximize the physical activity they already do in their daily lives, strengthen social connections, and feel healthy.”
After working with a coach through short video calls for six weeks, participants:
- It has been reported that symptoms of depression have been significantly reduced. Fifty-five percent of participants no longer suffered from depression after treatment, as measured by the Patient Health Questionnaire-9 (PHQ-9), a standardized depression screening test.
- They reported a 36% reduction in anxiety symptoms as measured by the Generalized Anxiety Disorder-7 (GAD-7) screening test.
- They reported a significant improvement in their quality of life.
- They scored higher on simple memory and attention tests.
Additionally, the researchers found that treatment effects persisted for three months when participants continued to be followed after the intervention ended.
“Clinical trials have shown that most current interventions are only about 30% effective on average in remitting depression. Here, we see that effectiveness nearly doubled by targeting key lifestyle predictors with data-driven, personalized coaching,” Mishra said.
Mishra believes this intervention may be more effective because it departs from common behavioral health recommendations.
“We all know we need to eat healthier, get eight hours of sleep, and exercise 150 minutes a week,” she says. “But I think personalized insights can be more empowering than these general guidelines because they’re less overwhelming. When a person is in a state of depression, it’s impossible to change everything in your life. You’re just trying to survive and function day to day.”
Although small, this study provided the first evidence that digital monitoring, insights gained from machine learning, and brief, personalized weekly coaching delivered remotely may be a promising integrated approach for addressing mild to moderate depression in large populations. Larger controlled studies of this individualized treatment approach are needed to validate the results.
Other co-authors on the study include Jason Nunn, Susanna Purpura, Satish Jaiswal, Khotan Afshar, Vojislav Maric, James K. Manchanda, and Charles T. Taylor of the University of California, San Diego. Dakshin Ramanathan of the University of California, San Diego and VA San Diego Medical Center;
Funding: This study was funded in part by a seed grant from the Hope for Depression Research Foundation.
Answers to key questions:
a: When a person is depressed, they are just trying to function on a daily basis and are operating in survival mode. It can be very overwhelming to give them a huge, generalized list of lifestyle changes. Depression varies greatly depending on an individual’s biological characteristics, so what works for one person may not necessarily work for another. By targeting only a single trigger based on your data, your Certain low moods get rid of the fatigue of trying to change everything at once.
a: The model tracks patterns between biometric movements and active mood logs over two weeks. Calculate fundamental correlations to identify key risk factors. If the algorithm determines that a lack of social interaction is a direct predictor of your low mood, the coach will provide therapy to strengthen your social connections. If sleep is the main factor, go into the cognitive behavioral therapy track for insomnia.
a: Data shows healing is built to last. This clinical trial revealed that after weekly video coaching calls ended, depression and anxiety were significantly reduced, and memory and attention remained completely stable during a three-month follow-up period.
Editorial note:
- This article was edited by the editors of Neuroscience News.
- Journal articles were reviewed in full text.
- Additional context added by staff.
About this AI and depression research news
author: suzanne bard
sauce: UCSD
contact: Suzanne Byrd – UCSD
image: Image credited to Neuroscience News
Original research: Open access.
“Personalized Machine Learning Interventions to Optimize Lifestyle Behaviors in Depression: A Pilot Study” by Jason Nan, Suzanna Purpura, Satish Jaiswal, Houtan Afshar, Vojislav Maric, James K. Manchanda, Charles T. Taylor, Dhakshin Ramanathan, and Jyoti Mishra. NPP—Digital Psychiatry and Neuroscience
DOI:10.1038/s44277-026-00062-3
abstract
A personalized machine learning-based intervention to optimize lifestyle behaviors in patients with depression: A pilot study.
There is a great need for personalized, data-driven interventions for depression. Here, we leveraged N-of-1 machine learning (ML) to optimally target a behavioral lifestyle intervention for depression. Fifty patients with mild to moderate depression were enrolled in a single-arm, open-label, personalized mood enhancement (PerMA) pilot clinical trial (NCT05662254).
Participants completed a 2-week digital monitoring phase using smartphone-based Ecological Momentary Assessment (EMA, 4 times per day) and tracking of mood and lifestyle factors (sleep/exercise/diet/social connections) via smartwatch.
A personalized ML model was generated from these data to identify the lifestyle factors that best predicted an individual’s mood, and the results were translated into an Individualized Mood Enhancement Plan (iMAP) that participants implemented once a week for six weeks with the guidance of a health coach.
Intervention completers (n = 40) showed a significant reduction in depressive symptoms (primary outcome self-assessment PHQ9 -3.5 ± 3.8, Cohen’s d = -0.89, CI [−1.25 −0.53]p<0.001; clinician-rated HDRS −7.2 ± 6.8, d = −1.03, CI [−1.41 −0.65]p < 1E-6), and the effect persisted up to 12 weeks of follow-up. Comorbid anxiety was also significantly reduced (GAD7: d = −0.85, CI [−1.2, −0.49]p < 0.001) and improved quality of life (d = 0.68, CI [0.33, 1.02]p < 0.001).
Furthermore, objective cognitive measures including selective attention influenced depression (d = 0.51, CI [0.18, 0.84]p < 0.001), interference processing (d = 0.53, CI) [0.2, 0.85]p < 0.01) and working memory (d = 0.66, CI) [0.31, 0.99]p < 0.001) showed significant enhancement.
EMA tracking confirmed that improvements in depressed mood were specifically predicted by individually targeted lifestyle improvements (β = 0.4 ± 0.09, p < 0.0005). Finally, decision algorithms and large-scale language models (LLMs) can match human coach-driven iMAP assignments with up to 95% accuracy.
The PerMA trial presents a personalized lifestyle intervention approach for depression, which merits scale-up and RCT trials to establish clinical efficacy.
PERMA was registered with ClinicalTrials.gov under registration number NCT05662254.
