For some people with depression, finding the right medication can be a process of trial and error lasting for months or even years, which can worsen symptoms.
But what if doctors, when diagnosing someone with depression, could assess exactly how depression is affecting a patient鈥檚 brain and prescribe a treatment that gets it right the first time?
Scientists may be a step closer to that reality, thanks to new research that has identified six subtypes 鈥� or 鈥渂iotypes鈥� 鈥� of major depression via brain imaging combined with machine learning. The study, published Monday in , also tested how three of those biotypes responded to different antidepressants and therapies.
鈥淭here are currently no tests available to help pinpoint what type of depression (people) have, or, I think especially importantly, what treatment might be most suitable for them,鈥� said Dr. Leanne Williams, the study鈥檚 senior author and the Vincent V.C. Woo Professor of Psychiatry and Behavioral Sciences at the Stanford University School of Medicine in California. 鈥淭he current situation is we rely on a person to tell us what they鈥檙e experiencing and for the physician or therapist to observe symptoms and come to a diagnosis.鈥�
About 280 million people worldwide and 26 million people in the United States have depression, which is a leading cause of disability. Some 30 per cent to 40 per cent of people with depression do not experience symptom improvement after trying one treatment, according to the study. And about 30 per cent of people diagnosed with depression go on to experience when the disorder doesn鈥檛 improve after multiple treatment attempts.
鈥淭hat motivated this study 鈥� to have a whole new way to more quickly get the right treatment, to figure out the right treatment for each person the first time,鈥� said Williams, who is also the director of the Stanford Center for Precision Mental Health and Wellness. Williams lost her partner to a decades-long struggle with depression in 2015 and for more than 20 years on individualized mental health care.
Mapping depression in the brain
The authors used data from 801 adult participants who were previously diagnosed with depression or anxiety, and 137 healthy control group participants. The authors used functional MRI 鈥� magnetic resonance imaging 鈥� to measure participants鈥� brain activity when they were at rest doing nothing, focusing on brain regions already known to play a role in depression, and on the connections between those regions. They also monitored brain activity when participants, who were in their mid-30s on average, engaged in various tests that evaluated their cognitive and emotional functioning.
The authors also randomly assigned 250 of the participants to receive behavioral talk therapy or one of three commonly used antidepressants: venlafaxine, escitalopram or sertraline.
The six biotypes of depression the authors found include one characterized by hyperactivity in cognitive regions, which was associated with more anxiety, negative bias, threat dysregulation and anhedonia than other biotypes. Threat dysregulation refers to how people manage their reactions to their fears, such as social interactions, Williams said. Anhedonia is the or enjoyment of life experiences.
Participants with this biotype also performed worse on executive function tasks that assessed how well they could manage thoughts or behaviors, make decisions or suppress distraction, Williams said. They also had the best response to the antidepressant venlafaxine.
Another biotype was marked by higher levels of brain connectivity in three regions associated with depression and problem-solving. Those with this biotype also made errors in executive function tests but did do well on cognitive tasks. They found their symptoms better alleviated by behavioral talk therapy, which teaches skills for better addressing daily problems.
The higher connectivity in these brain regions might be what helped participants with that biotype more readily adopt new skills, study coauthor Dr. Jun Ma said .
There was also a biotype distinguished by lower levels of activity in the brain circuit that manages attention. This biotype was linked with more mistakes on tasks requiring sustained attention, and less of a chance of improving with therapy. People with this biotype may need medication for the dysfunction first so they can gain more from therapy, said Ma, the Beth and George Vitoux Professor of Medicine at the University of Illinois Chicago.
The authors also found a biotype characterized by high emotional reactivity, meaning the brains of participants in this group were more affected by emotional inputs such as their own emotions or people鈥檚 facial expressions, Williams said. One other biotype was associated with lower activity in cognitive brain regions and less connectivity in emotional regions, meaning these participants had difficulty responding to cognitive information and regulating negative emotions.
Those last two biotypes didn鈥檛 respond to the medications or therapy, which suggests other options may be needed for people with those types, Williams said. 鈥淚n other studies, we鈥檙e finding they respond to some of the newer treatments being developed.鈥�
The sixth biotype identified didn鈥檛 differ from brain scans of the same region in people without depression. Williams said she thinks this finding could mean the full range of brain biology underlying depression hasn鈥檛 been entirely discovered.
鈥淒epression is a lot of different things with a lot of different causes, biological changes and treatments,鈥� said Dr. Richard Keefe, professor emeritus of psychiatry and behavioral sciences at Duke University Medical Center in North Carolina, who wasn鈥檛 involved in the study.
The study 鈥渢akes a positive step in the direction鈥� of figuring these things out, Keefe added via email.
Hurdles and next steps
The study, though 鈥渟ophisticated and very well done,鈥� does have several key problems, including the low number of people enrolled in treatment, said Dr. Jonathan Alpert, the Dorothy and Marty Silverman Chair of the department of psychiatry and behavioral sciences at Montefiore Medical Center in New York City. 鈥淚t needs to be thought of as a very preliminary study that needs to be replicated.鈥�
Additionally, more diverse samples are needed, said Alpert, who wasn鈥檛 involved in the study and is a professor of psychiatry, neuroscience and pediatrics at the Albert Einstein College of Medicine. Most participants were White, and two per cent were Black.
But the most important next step is a study that tests the authors鈥� hypothesis 鈥� that if patients have particular biotypes, they鈥檒l do better on a specific treatment 鈥� and tracks participants over time, said Alpert, chair of the American Psychiatric Association鈥檚 Council on Research.
The 250 treatment participants weren鈥檛 randomized on the basis of their biotypes. So, what Alpert recommends the authors do next is assign people to treatments based on their biotypes and see whether those participants have better outcomes from that method than if they had been assigned to a treatment according to clinical judgment without knowledge of their biotype.
Another issue is that the study investigated only one form of psychotherapy and three medications; in the real world, there are many of each, Alpert said. The medications were also all serotonin-based, but there are a few other classes of antidepressants.
Studies can only do so much at a time, Alpert acknowledged, but addressing these shortcomings incrementally would help continue advancements toward precision psychiatry.
What these findings mean for you
The study鈥檚 methods and findings are years away from being applied to direct patient care, experts said, but there is funding for such efforts.
鈥淪ince 2009, the National Institute of Mental Health has been invested in using basic science, including functional brain imaging like in this study, for identifying the causes of mental illness through approaches that delve deeper than the traditional diagnostic approaches,鈥� Keefe said.
This month, an US$18.8 million grant as part of the National Institutes of Health鈥檚 Individually Measured Phenotypes to Advance Computational Translation in Mental Health Initiative. The grant supports a five-year project involving 4,500 participants, which is centred on the development of a better diagnosis and treatment tool for depression biotypes.
The new study鈥檚 approach has begun to be experimentally implemented at a Stanford clinic, Williams said.
鈥淲hen we use it in that setting, we can effectively double the chance of someone getting better,鈥� she said, taking the odds from around 30 per cent of people improving with the traditional approach to about 75 per cent with the more precise method.
This method isn鈥檛 intended to replace or be the primary choice for assessments of individual cases of depression, Williams said. It鈥檚 another piece that can be added to the puzzle that also includes symptom information, clinical interviews and more.
For now, people with depression should know 鈥渢here is continued progress being made鈥� toward efficiently getting patients effective treatment, Alpert said. If you鈥檙e struggling, talk with a mental health professional about your options.
One powerful effect these findings could have immediately is to reduce stigma, Williams said. For people who think their depression is just due to them 鈥渘ot trying hard enough,鈥� she added, understanding the disorder through the lens of objective measures of brain function could be 鈥渄eeply helpful."
