According to the National Alliance on Mental Illness and the World Health Organization, depression affects 16 million Americans and 322 million people worldwide. New evidence suggests that the COVID-19 pandemic is further worsening the prevalence of depression in the general population. With this trajectory, it is evident that more effective strategies are needed for therapies that address this critical public health problem.
In a recent study, published in the June 9, 2021 online edition of Translational Psychiatry Nature, researchers at the University of California, San Diego School of Medicine used a combination of modalities, such as measuring brain function, cognition, and lifestyle factors, to generate individualized predictions depression.
The machine learning and personalized approach took into account several factors related to an individual’s subjective symptoms, such as sleep, exercise, diet, stress, cognitive performance and brain activity.
“There are different reasons and underlying causes of depression,” said Jyoti Mishra, PhD, lead author of the study, director of NEATLabs and assistant professor in the Department of Psychiatry at UC San Diego School of Medicine. . “Put simply, today’s healthcare standards are all about asking people how they feel and then writing a prescription for drugs. These first-line treatments have been shown to be only mild to moderately effective in large-scale trials.
“Depression is a multifaceted illness, and we need to approach it with personalized treatment, whether it’s therapy with a mental health professional, more exercise, or a combination of ‘approaches.”
The month-long study collected data from 14 participants with depression using smartphone apps and wearable devices (such as smartwatches) to measure the mood and lifestyle variables of the patient. sleep, exercise, diet and stress, and has associated them with cognitive assessments and electroencephalography, using scalp electrodes to record brain activity.
The goal was not to make comparisons between individuals, but to model predictors of daily fluctuations in each person’s depressed mood.
Researchers have developed a new machine learning pipeline to systematically identify distinct predictors of bad mood in each individual.
For example, exercise and daily caffeine consumption appeared to be strong predictors of mood for one participant, but for another, it was sleep and stress that were more predictive, while in a third topic, the main predictors were brain function and cognitive responses to rewards. .
“We shouldn’t approach mental health as one size fits all. Patients will benefit from a more direct and quantified insight into how specific behaviors may fuel their depression. Clinicians can use this data to understand how their patients might feel and better integrate medical and behavioral approaches to improve and maintain mental health, ”said Mishra.
“Our study shows that we can use readily available technology and tools, such as mobile phone apps, to collect information from people with or at risk for depression, without significant burden to them, and then use that information to design personalized treatment plans. . “
Mishra said the next steps are to examine whether personalized treatment plans driven by data and machine learning are effective.
“Our findings may have broader implications than depression. Anyone seeking greater well-being could benefit from information quantified from their own data. If I don’t know what’s wrong, how do I know how to feel better? “
Co-authors include: Rutvik Shah, Gillian Grennan, MariamZafar-Khan, Fahad Alim, Sujit Dey, all with UC San Diego; and Dhakshin Ramanathan with UC San Diego and the VA San Diego Medical Center.
The research was funded, in part, by the University of California at San Diego and startup grants from the Center for Mental Health Technology at UC San Diego and the Sanford Institute for Empathy and Compassion.
Disclosure: Shah, Dey and Mishra have filed an invention disclosure for “Personalized Depressed Mood Machine Learning Using Wearable Devices.”