Menu Close

What a little-known brain region can tell us about depression

A tiny brain region, called the habenula, has been proposed to drive the symptoms of depression, which can have a devastating impact on people’s lives. Despite its small size – just 3mm in diameter – the habenula is thought to play an important role in processing negative experiences. A theory, developed in recent years, suggests that the habenula is hyperactive in depressed people. Our latest study may call this theory into question. Not only did we find that this was not the case, but the habenula responded in exactly the opposite way than predicted in depressed people.

The results of our study show that habenula function is disrupted in depression, but not in the way that we had expected. If the habenula is really responding in such an abnormal way, we need to work out why this is and how it relates to the way that depressed people feel and think, as it could be important for the development of new treatments.

The habenula is an ancient structure (in evolutionary terms) situated deep in the middle of the brain. It is closely connected with other important centres, such as the nuclei that contain dopamine and serotonin neurons, which exert a profound impact on emotion and behaviour through their effects in the rest of the brain. All vertebrate species have a habenula, even the lamprey, which predates the dinosaurs, so it’s long been thought to play a fundamental role in brain function and behaviour.

Even lampreys have a habenula. Dave Herasimtschuk/Flickr, CC BY

Over the past decade, neuroscientists have become increasingly interested in the function of the habenula, in part triggered by the publication in 2007 of a seminal paper on monkeys. Focusing on a specific part of this structure (the lateral habenula) the authors showed a remarkable interaction between the habenula and dopamine neurons.

When they stimulated habenula neurons and simultaneously recorded dopamine neurons, there was a profound reduction in the firing of the latter. Since dopamine neurons are known to fire when unexpectedly good things happen (called a “positive prediction error”) it raised the possibility that habenula neurons might also respond when unexpectedly bad things happen (“negative prediction error”). This yin-yang relationship between habenula and dopamine neurons was confirmed in another paper from the same group two years later.

Learned helplessness in animals

Separately, other groups of researchers were investigating habenula function in rats with “learned helplessness”. Experiments conducted in the 1960s with dogs showed that when they were exposed to random electric shocks over which they had no control they wouldn’t try to escape when the opportunity arose. However, dogs that were able to stop the shocks by pressing a lever did try to escape. This learned helplessness is used as a model for depression.

Rats bred to be particularly vulnerable to learned helplessness were found to have increased activity in the lateral habenula. Conversely, inhibiting the habenula with drugs was shown to protect against learned helplessness. A more recent study showed that the neurons connecting the lateral habenula with dopamine nuclei were overactive in helpless rats.

These and other findings were synthesised into an influential theory of depressive symptoms. If the lateral habenula encodes negative events and turns off dopamine neurons, perhaps a chronically hyperactive habenula could drive certain symptoms of depression, for example anhedonia (loss of interest or pleasure in previously enjoyable activities) or a pessimistic outlook.

However, habenula research in humans has lagged behind animal studies. Consequently, until recently, the hyperactive habenula theory of depression had never been tested directly. In part this is because measuring the tiny habenula in humans is technically very difficult and requires a special high-resolution brain imaging method (to rule out the possibility that any results could be contaminated by activity in nearby regions).

In a paper published two years ago, we used such a method to show that the habenula responded when healthy volunteers expected to receive a painful electric shock over which they had no control. And the more likely they were to get the shock the more the habenula was activated, precisely what would be predicted from animal research. This was the precursor to our recently published study, which directly tested the theory that the habenula is hyperactive in depression.

Testing the theory in humans

We tested 25 people with depression, none of whom were taking any medication, and compared them with 25 never-depressed healthy volunteers. The participants underwent various high-resolution brain scans. Based on the theory described above, we predicted that the depressed participants would have a hyperactive habenula, either in the “resting-state” (lying in the scanner doing nothing) or evoked (by expectation of an electric shock), or possibly both. But our predictions were wrong.

“Resting-state” habenula function was very similar between the groups. Even more surprising, evoked habenula activation showed the exact opposite pattern to the healthy volunteers. As the electric shocks became more and more likely, activation in the habenula actually decreased.

This is just one small study, but if our results are correct the hyperactive habenula theory of depression needs a serious rethink.

Like much research, our study raises more questions than it answers. Does habenula function return to normal when people recover from depression? If participants were able to take action to avoid the shocks would the results look different? And most importantly, can abnormal habenula function predict how a person will respond to antidepressants or psychotherapy?

Given the theoretical, experimental and clinical interest in the habenula, it is likely that we’ll be hearing much more about this pea-sized brain structure in future.

Want to write?

Write an article and join a growing community of more than 171,300 academics and researchers from 4,744 institutions.

Register now