Dr. Madhukar Trivedi Interview

Before we get into your research and its implications, I want to ask a very basic question. What is depression?

Let me give you a few different kinds of answers to that.

Depression is a common medical illness, a brain disease, that affects many, many people. Approximately 1 in 8 or 1 in 9 Americans will have depression at some point in their lives.

Major Depressive Disorder (MDD), which has been the focus of much of my research, is a clinical illness that is associated with brain dysfunction, and is characterized by symptoms that can include a sad or depressed mood; anhedonia, which is an absence of pleasure or the inability to enjoy things that used to give one pleasure; hopelessness or helplessness; suicidal ideation; and psychomotor agitation or retardation, among other symptoms. It becomes diagnosable when a certain number of these symptoms are present almost all the time for more than two weeks.

MDD is episodic in nature. For some people it can last a short time, and never or rarely re-occur. For other people these episodes can last for a very long time, even months or years, and for a large proportion of patients, recovery isn’t complete even between episodes.

Approximately 1 in 8 or 1 in 9 Americans will have depression at some point in their lives.

What causes depression?

It’s a multi-factorial illness. We know, for instance, that there can be a genetic predisposition to depression. It is also influenced by one’s environment, which may mean that it’s triggered by specific events in the present that are stressful or painful but can also mean that the brain has been influenced by past environmental factors to be more vulnerable. Depression in adolescence and adulthood, for instance, may be the result of early childhood experience, including trauma and stress.  

There is some really interesting evidence emerging, as well, that certain types of depression can be caused by inflammation in the brain.

It’s also worth saying, when thinking about causation, that what we now call “depression,” or “Major Depressive Disorder,” may in fact be a conglomeration of different disorders. Our assumption is that over time, as our understanding of depression improves, we will end up identifying subtypes and subgroups, and that that disaggregation will enable us to be much more precise in the treatments that we recommend.

Think about it in comparison to how the field of oncology has developed. We moved from the idea of cancer, broadly, to organ-specific cancers, to where we are now, where in many cases we are able to specify the type of lung cancer that a given patient has, defined by the molecular definition of that particular cancer. Or with breast cancer, where in some cases there is genetic profiling of the cause, as with the BRCA gene.

That’s the direction in which we hope to move the field of psychiatry.

I want to go back to something you said when defining depression. You called it a “brain disease,” and a “brain dysfunction.” What do you mean by that?

It means that when you are depressed there are circuits in the brain that are not functioning appropriately. In people who have depression, these circuits are interpreting external and internal stimuli in distorted ways.

It’s incredibly complex, and many different areas and circuits in the brain are likely to be involved. We have the most evidence, however, about what we call the reward circuit and the emotion processing circuit.

In the imaging studies of people experiencing depression, as compared to healthy controls, we can see physiological changes in those two circuits, and changes in the patterns of brain activity both in how they are responding to things that should give them a sense of reward or satisfaction—that’s the reward circuit—and in those areas involved in high level emotional regulation and interpretation.

Is there a chicken and egg question here? Are we experiencing all the emotions that we commonly associate with depression, and then it’s that experience that produces these changes in the brain? Or does the brain dysfunction come first, and cause the emotions?

That’s a good question, and we don’t know the answer yet. The EMBARC initiative is designed, in part, to answer it. It involves following large cohorts of people who have depression as well as those whom we know are at risk for depression as a result of their family histories.

We’re doing this at multiple sites around the country, and we are providing the patients really comprehensive sets of tests and measures, as well as variety of treatments, to observe not just what a static picture of their depression looks like, but to see how the brain changes over time as they recover, and to begin to understand what treatments are best for what kind of person, depending on the nature of their disorder.

My own guess is that the circuits come first. There is a dysfunction there that ultimately leads to the subjective emotional experience of depression.

It sounds like the vision you have, for the future of psychiatry, is real precision medicine. That someone can go into a doctor’s office, presenting with some of the symptoms of depression, and after doing the right series of tests, the doctor can offer a customized treatment for the person.

Yes, but to be honest we are just at the beginning of that move toward precision medicine. We are not there yet. Currently the choice of medication, or therapy, isn’t based on any kind of lab test or brain test. It’s based on clinical factors rather than labs. You look at the patient and their history, look at the different treatments they’ve had in the past, and do the best you can to match the patient to the treatment.

Our main goal at the Center is to get to a point where we will develop these tests that will enable precision. But we are not there yet. They conducted the Framingham Heart Study 65 years ago. We have not done the same thing for depression.

What we have right now are some very good signals, from the research, about more precise matching of patients to treatments, but it’s not clinical yet.

What signals?

We’re starting to develop a picture of a sub-group of patients, which might be as high as 15-20 percent of the total group of depressed patients, who seem to have an inflammatory, immune-response type of depression. For this group, there’s strong evidence that traditional antidepressants may not be the best treatments. The better intervention may to be focus directly on the inflammation.

Some patients who have this inflammatory response do better, for instance, with exercise than with traditional antidepressants. Or we may use medications that are currently being used to treat inflammatory diseases, things like non-steroidal anti-inflammatory drugs, what are called NSAIDs, or Anti-TNF drugs.

There is another, smaller groups of patients in whom we see changes in an area of the brain called the dorsal lateral prefrontal cortex. There is some evidence that the type of depression that corresponds to that responds well to a treatment called Transcranial magnetic stimulation, or TMS. It involves putting a magnet outside of the skull and giving what are basically magnetic pulses, at low amplitudes, to certain parts of the circuit. It changes the circuit itself, and that affects the behavior and emotion.

This is part of the purpose of EMBARC, to take these promising signals, from the research, and move them closer to clinical practice. And to discover other signals and relationships.

The goal is to use all this information to develop an index, or an algorithm, which will help clinicians understand what type of depression particular patients have, and then find the best match for them in terms of treatment.

Tell me more about EMBARC. It sounds interesting.

EMBARC is a large-scale, multi-modal, multi-site study, one of the biggest and most comprehensive of its kind. Its goal is to address the question we just talked about, to identify and clarify sub-types of depression, and to understand which patient with which subtype will do better with which treatment.

To do this we are using a wide array of imaging tools, including fMRIs and Diffusion Tensor Imaging, which can assess white matter changes in the brain. We are looking at electrophysiology through EEGs, electroencephalograms. We are putting the patients through computer driven tasks to look, for instance, at how they respond to rewards and measuring reaction time to stimuli. We are keeping track of clinical symptoms like anhedonia, anxiety, suicidal ideation, diminished work productivity, and others.

The goal is to use all this information to develop an index, or an algorithm, which will help clinicians understand what type of depression particular patients have, and then find the best match for them in terms of treatment.

This index is what you call the DTRI?

Yes, what we’re calling the Depression Treatment Response Index, or DTRI.

We want every person to have a biological signature that is their own, and that you could plug into the DTRI to help match them with an optimal treatment.

Are you doing genetic testing as well? Does that figure in?

We will eventually look at the genetic markers, but that was not funded for this round. We didn’t prioritize it because the signals from genetics have been less encouraging to this point. Part of it is that full genome sequencing is expensive, and so the studies aren’t there yet. As the sequencing comes down in price it will be easier to incorporate the genetic and epigenetic information into the broader picture.

We chose our measures based on previously studied measures that showed some relationship, and also to some degree on what the theoretical frameworks predict might be useful. It’s not comprehensive or complete, but it’s a very good start. I imagine we or others will add additional tests to improve precision down the road.

This is just the beginning. We are 50 years behind the curve in terms of looking for the signatures for depression.

Do you imagine a situation, a few decades from now, in which everyone gets all these tests simply as a routine thing, and then has a fully developed biosignature as part of their medical record? Or would that be too expensive, and the full work-up would be for people who have some initial indications of a disorder?

I suspect the latter scenario. We will develop tests that are manageable, feasible, and hopefully not very expensive that we can give to everyone to identify risks for, and early indicators of, the illness. Then following that first round of more universal tests, we would do more complex assessments of people who had some indication of risk or illness.

The cost issue isn’t entirely straightforward, by the way. If we say a 50-year-old needs a colonoscopy, we don’t say that it’s too expensive, even though it is expensive. We’ve decided that universal screening at that age makes sense.

I know you’ve done research on treatments other than anti-depressants. Can you talk a little about some of the existing alternatives, and maybe some newer types of treatments that may be coming?

Traditional antidepressants act on the serotonin or norepinephrine systems, so any mechanism involved in depression that is outside of those may not respond to those drugs.

The ketamine group of drugs, where we’ve seen a lot of promise recently, works through a different mechanism. It acts on the NMDA receptor.

There isn’t a drug company that is selling exercise, but in a sense it is a novel mechanism. It has all sorts of effects that we believe can be efficacious with depression, including neurogenesis and anti-inflammatory effects. More work needs to be done to understand the relationship, but I think we’re likely to see more emphasis on exercise as a treatment.

Transcranial magnetic stimulation, which I mentioned before, is another novel mechanism. There is a lot more work that needs to be done in general with neuro-modulation, where we’re targeting different parts of the brain directly.

I know scientists hate to make predictions, but I’ll ask you to make one anyway. Where are we going to be a few decades from now, in terms of our ability to diagnose and treat in a precise way? What’s your best guess.

There’s a famous line that prediction is a dangerous business, especially about the future. My guess, however, is that we will be able to identify maybe 2-4 subtypes of depression that don’t account for 100 percent of the cases, but some significant percentage of them. We will be able to do some relatively simple tests, and if you do belong to one of these categories then your treatment trajectory will be very different. I’m also very optimistic about neuro-modulation, and our capacity to get more precise with that.

The other hope is that we may be able to identify, well in advance, those who are at risk for developing some subtypes of depression. Then we can put some interventions in place to help prevent development of the disorders, or intervene early if they begin to develop.

So what would that look like, trying to prevent depression from even happening in the first place? How would you do that?

We are doing a study now, which is already in about 15 high schools, where we connect with every incoming freshman, and are able to ask them if their parents have depression or bipolar disorder.

What we’re going to be able to do, with that information, is identify people who are at risk, primarily because of a family history, and then follow them for 10 years, doing a full EMBARC type analysis four times a year.

As some of those people develop symptoms, and others don’t, we can develop signatures, and we can even offer some resilience training. And learn from that as well.

I’m surprised you’re able to get participation from schools, and parents and kids, around such a sensitive topic.

When we get a chance to talk to people who have had experience with depression or bipolar disorder, either in themselves or their family, and we explain to them what we are trying to accomplish, most people are so incredibly helpful. They have experienced the uncertainty and difficulty that these disorders create, and they are so willing to help. They are the most remarkable partners.

What does resilience training look like? Do we know how to prevent depression?

Right now we don’t have a very well defined package of what needs to be done. We have some indications, though. Managing and reducing stress, improving responses to bullying, improving interpersonal skills, increasing mindfulness, avoiding drug and alcohol use—there are good indications that these lower the risk of developing depression.

If someone is at high risk for bipolar disorder, I would tell them to maintain their regular sleep-wake cycle – in other words, try to go to bed and wake up at fairly set times. We know that changing your sleep-wake cycle is a big risk factor for people with bipolar disorder. If you are messing with that, you are increasing the risk. And this is often a challenge with teenagers, especially when they are out of school, because they want to stay up very late and then sleep in.

Physical activity likely helps. We know that middle school kids who are physically active have lower rates of depression. There are some very promising indicators, but we need to know more. That’s why we’re doing the studies.

On the one hand, it’s hard to see how you could just tell kids to do things like reduce stress, or exercise more, or not drink or smoke, and expect them to do that. On the other hand, if you really knew that you were at high risk for developing depression, or bipolar disorder, you might be more motivated, particularly if you’ve seen the damage they can do in your own family.

Exactly. It’s not a universal intervention. It’s for kids where we really know that they’re at risk. Imagine what you and I would do if we knew both our parents had diabetes, and two of our uncles had died from diabetes. We would be on the treadmill five days a week.

In the same way I think we can be hopeful that some of the people, including young people, who’ve seen the effects of mental illness in their family would be willing to change their behavior to give themselves a better chance of staying healthy.