Imagine that tomorrow you wake up with a persistent cough and a sore throat. You go to see your doctor and she asks about your symptoms, but also takes your temperature and blood pressure, and maybe takes a throat culture or blood sample to send to the lab. Most of the medical field operates in this way, but not mental health.

Recent work by our lab is trying to change this. We’re looking for “biomarkers,” physical tests that could help in the diagnosis of particular mental illnesses and help to inform treatment strategies. The goal of these tests is not to replace people's self-report about how they are feeling, but to rather to provide additional information to have a more complete picture of what is going on when someone has a mental illness (or is at risk for developing one).

Previous studies from our lab and others have identified patterns of brain activity that are related to risk for PTSD and depression. And we have been particularly excited by the results from one of our recent projects. My colleagues and I found that people with a certain pattern of brain activity had worsening anxiety over the six months following their brain scan while others did not.

This study looked at activity in three different brain regions, using a tool called fMRI to measure changes in blood flow as an indirect measure of brain function (the general idea is that if a brain area is more active, it needs more blood). In the fMRI scanner, participants completed separate tasks to activate each of the three brain regions.

Games, faces, and memory

In the first task, participants played a guessing game in which they were shown the back of a playing card and asked to guess if the card was going to be greater or less than five. The participants had no way of knowing which it would be, but when they guessed correctly they were informed that they would increase their chances of winning $10 at the end of the game. Part of the brain called the ventral striatum, a small region deep in the brain that is important for motivation and action, was particularly active when people learned that they had guessed correctly. But not everyone’s ventral striatum was the same – some people’s had a lot of activity while other’s had only a little. We wanted to know what these individual differences might mean.

The second task involved looking at pictures of fearful and angry faces and matching up which faces were the same. We looked at brain activity in a part of the brain called the amygdala, an almond-shaped region involved in recognizing and reacting to important aspects of our environment, and saw a similar result: some people had a lot of activity in this brain region when looking at the threatening faces compared to looking at shapes, whereas other people had less activity.

The third task involved holding numbers in memory and performing simple mathematical calculations with them versus adding and subtracting numbers that were on a screen the whole time. Doing math from memory was harder and tapped into the brain's prefrontal cortex, a region behind your forehead that is involved in planning and problem-solving, as well as emotion regulation. Once again, there was variability between participants in the activity of this region.

I was eager to untangle the significance of our new data. The the culmination of my years of graduate school research, it combined several lines of the inquiry from our lab that had been ongoing for decades. It took several months to process and analyze the data, and then it was finally time to look at the results.

I first looked to see if the three brain regions together (our biomarker) could predict whether people would become more or less depressed and anxious over the next six months. But it turned out that the biomarker didn't help with this. We were disappointed, to say the least.

After further discussion, however, I realized that our biomarker might not predict both depression and anxiety, but only one or the other. Symptoms of anxiety and depression often go hand-in-hand, and in previous studies we had looked at them together because we hadn't had a large enough sample size to pull them apart. But now we did, so I went back to the lab and analyzed the data looking separately at depression and anxiety symptoms. And it worked!

Biomarkers in mental health

The biomarker predicted whether or not people had increased anxiety over the next six months. High threat-related amygdala activity and low reward-related ventral striatum activity were associated with worsening anxiety over the next six months when combined with low prefrontal cortex activity. But if people had high activity prefrontal cortex activity, they were more likely to be resilient and instead showed decreased anxiety over time.

We found this exciting for a number of reasons. First of all, it added to a growing literature suggesting that biomarkers might be useful in the mental health field. Since we hadn't asked people to regulate their emotions on any of the tasks, we think the activity we measured might be a marker of generally how well these parts of the brain function for a range of things – like when a doctor measures your blood pressure as an index of your overall cardiovascular health.

Second, we were excited because we had identified a combination of biomarkers that could help augment the information people provided about their symptoms in identifying worsening anxiety. If this finding holds up in follow-up studies, these types of approaches could eventually be used by clinicians to help guide treatment decisions.

We were surprised, however, that the results turned out to be only for the specific type of anxiety – the type where people notice a lot of physical symptoms (heart racing, sweating, etc.) and not for general symptoms of anxiety and depression. We’re still not sure why the results were specific to this type of anxiety, so it will be important to try and replicate the results to find out more. Another important caveat is that the participants were college students, so the results might be different if we looked at a different sample of participants such as older adults. Additionally, we looked at symptoms of anxiety and didn’t have a chance to see if people would actually meet criteria to be diagnosed with an anxiety disorder, so that could be another important follow-up.

This was an exciting start, but we still have a lot to learn. In particular, we want to know if we can do anything to help increase prefrontal cortex activity for people who might be at risk of developing anxiety. We hope to do this through giving people psychotherapy (and measuring change in their brain activity) and/or using a type of treatment where doctors use magnets to directly stimulate the brain, an approach that is already being used to treat severe depression. Eventually these types of approaches could help to figure out who would benefit from one type of therapy over another, and could help improve our ability to treat or prevent worsening anxiety.