SAN DIEGO —Depression will affect about 15% of Americans at some point in their life, resulting in lost productivity, poor quality of life, and, in some cases, suicide. Antidepressant treatments have been shown effective in approximately 65% of patients but usually require several weeks or even months to take effect. However, the mechanisms underlying antidepressant treatment, as well as the etiology of depression itself, have not been identified.

In recent years, researchers have learned that depression results, in part, from a loss of neurotrophic support leading to atrophy and loss of neurons in the brain. This finding suggests that there is a role in the etiology and treatment of depression for structural alterations as well as neurochemical ones. Antidepressant agents appear to act, at least partially, by inducing neurotrophic effects that reverse the structural changes that occur in depression. Could behavioral modifications such as physical exercise have similar beneficial effects—when used either alone or in combination with antidepressant therapy?

Investigators at Yale University and other institutions have formulated a neurotrophic hypothesis of depression and antidepressant response, which suggests that neurotrophic, neurogenic, and gliogenic effects of antidepressant agents could reverse or block the atrophy of the hippocampus and prefrontal cortex that has been observed in patients with major depressive disorder (MDD). Induction, through exercise, of neurogenesis and of a major neurotrophic factor found in the brain—known as brain-derived neurotrophic factor (BDNF)—indicates that behavioral modification can have similar antidepressant effects and can cause structural remodeling in the adult brain, according to Ronald S. Duman, PhD, Professor of Psychiatry and Pharmacology and Director of the Abraham Ribicoff Research Facilities at the Yale University School of Medicine.

“This is something that we hadn’t thought about much, but behavioral modifications and behavioral therapy can probably lead to an increase in neurotrophic factors and similar neurotrophic changes,” he said at the 2007 Annual Meeting of the American Psychiatric Association. “A combination of behavioral therapy and exercise as well as medication may produce more efficacious structural remodeling and therapeutic intervention than either alone.”

Dr. Duman zeroed in on exercise as one of the factors, along with an enriched (ie, activity-promoting) environment, hippocampal-dependent learning, and antidepressant therapy, that up-regulate neurogenesis and BDNF in the adult hippocampus—countering the process of down-regulation that occurs in depression by such factors as stress, adrenal steroids, drugs, and age.

He cited research by van Praag and colleagues that depicted several settings for experimental mice, including one that was devoid of anything other than the mice themselves. “This is how we typically keep our animals,” he said. “It’s not a very enriched environment, I’ll say. What the authors found was that an enriching environment—lots of toys, running wheels—can produce a very robust increase in neurogenesis … and that the running wheel was something that probably had the most robust effect of all.

“It turns out—and we’ve done these studies now, too—if you give a mouse a running wheel, it will start out running 4 or 5 km a night. By a week to 10 days, it’s running 12 to 14 km a night. Not forced—it just loves to get on that running wheel.”

The reason for this could be that factors such as running and an enriched environment have been found to increase neurogenesis and BDNF in the adult brain, Dr. Duman emphasized. “What happens is that you can see very robust induction of neurons, a very nice induction of BDNF,” he said. “You also have induction of vascular endothelial growth factor [VEGF]. So running has a tremendous effect on trophic support, on neurogenesis in the brain.”

Dr. Duman and his colleagues at Yale have been studying the behavioral effects of running and have found that it results in a robust antidepressant-like effect—a finding that has also been borne out in several clinical studies involving exercise in general. “So this is something else that we’re pursuing,” he said. “We have some studies going on in which we hope to identify some other factors that are induced by exercise but not by antidepressants, raising the possibility that we may have identified factors that could also be targeted for increasing the antidepressant effect.”

Dr. Duman noted the uncertainty regarding the mechanisms underlying the effects of antidepressant agents. “Even though we’ve had these drugs available now for over 50 years—and we’ve used them rather successfully, at least to a certain degree—it’s still unclear what the mechanism of action is,” he said. “That’s something that has interested me and my lab, now, for well over 25 years.”

In Dr. Duman’s view, the neurotrophic hypothesis encompasses the following key observations:

• There are opposing actions of stress and antidepressants on neurotrophic factor expression and adult neurogenesis, occurring mainly in the hippocampal region and on glial proliferation, occurring in the prefrontal cortex and in other regions of the brain.

• The levels of growth markers in serum may be biomarkers of depression and antidepressant response, and they have functional implications.

• There is evidence of gene and environment interactions in mood disorders related to neurotrophic factors.

• Behavioral phenotypes of antidepressants can be either dependent on or independent of neurogenesis, which implies a role for neuroplasticity.

The evidence supporting this hypothesis comes from basic research studies that have used stress as a model of depression. “There are a lot of data now showing that [chronic] stress can lead to atrophy and remodeling of cortical as well as hippocampal neurons,” Dr. Duman reported. “Stress can decrease processes of neurogenesis … in the adult brain, and stress decreases one of the major neurotrophic factors in the brain—BDNF—in the hippocampus.”

On the clinical side of the available evidence, he cited several studies indicating that hippocampal volume is decreased in patients with MDD—a reduction that is related to the duration of the depression. Dr. Duman also pointed to studies showing that antidepressant treatments partially block or reverse the reduction in hippocampal volume seen in MDD.

“The important point about these studies is that this [main finding] illustrates the fact that there are indeed structural changes that occur in depression,” he said. “But what differentiates this kind of structural change from neurologic disorders where you may have damage and loss that is not reversible [is] that these structural changes are, in fact, reversible, at least under some conditions—maybe not all, but at least some of the changes.”

Although most of these studies have focused on the hippocampus, there are now considerable data indicating that structural changes occur in other regions of the brain, including the amygdala and some specific areas of the prefrontal cortex. For example, MDD has been correlated with decreased volumes of both amygdala core nuclei and the subgenual prefrontal cortex.

Evidence of neuronal atrophy and loss in mood disorders has also been derived from postmortem studies, taking the neurotrophic hypothesis to another level, “a cellular level,” according to Dr. Duman.

 “There have been some studies counting numbers of neurons and glia, and there does not seem to be a difference in the number of neurons in either prefrontal cortex or hippocampus,” he said. “But there is a decrease in the number of glia, at least in prefrontal cortex.”

He cited evidence from these studies of a shrinking of neuronal cell body size, as well as data indicating a reduced number of glia in the amygdala and cingulate cortex. “So the data so far demonstrate a loss of the size of neurons, but in terms of numbers there’s actually a decrease in numbers of glia, which could be important to the metabolic support provided to these neurons.

Dr. Duman noted the substantial number of studies supporting the key observations that comprise the neurotrophic hypothesis—from requirements for the interaction of genes and environment to the pathways underlying neuroplasticity. He focused on neurotrophic growth factors implicated in stress and depression—particularly BDNF, which is a member of the nerve growth factor (NGF) family, along with NGF, neurotrophin-3 (NT-3), and NT-4/5. Other groupings include the VEGF family, which includes such members as VEGF-A and VEGF-B; the fibroblast growth factor (FGF) family, whose members include FGF-2 and FGF1-23; and the insulin-like growth factor (IGF) family, which includes insulin, IGF-1 and IGF-2.

Dr. Duman noted the substantial number of studies supporting the key observations that comprise the neurotrophic hypothesis—from requirements for the interaction of genes and environment to the pathways underlying neuroplasticity. He focused on neurotrophic growth factors implicated in stress and depression—particularly BDNF, which is a member of the nerve growth factor (NGF) family, along with NGF, neurotrophin-3 (NT-3), and NT-4/5. Other groupings include the VEGF family, which includes such members as VEGF-A and VEGF-B; the fibroblast growth factor (FGF) family, whose members include FGF-2 and FGF1-23; and the insulin-like growth factor (IGF) family, which includes insulin, IGF-1 and IGF-2.

He then proceeded to summarize how stress influences the expression of BDNF in the hippocampus. “You can see there are many different types of stressors, different lengths of stress, and … almost every case that’s been reported has shown that stress decreases the expression of BDNF… in hippocampus and, in some cases, the prefrontal cortex as well,” he said. “It’s one of the most consistent effects, which really hadn’t been reported before for the action of stress. In contrast to that, antidepressant treatment increases the expression of BDNF in hippocampus.”

Each class of antidepressants shares the ability to increase the expression of BDNF, including such SSRIs as sertraline and fluoxetine; a norepinephrine-selective compound, desipramine; and a monoamine oxidase inhibitor, tranylcypromine. By contrast, nonantidepressant agents have not been shown to influence BDNF expression.

Given this evidence that BDNF was regulated by both stress and antidepressant treatment, Dr. Duman and his colleagues sought to determine whether there was any functional relevance to this regulation, by examining different behavioral models of depression and antidepressant response. He cited learned helplessness, a model in which researchers expose animals to a stressor that they cannot escape and subsequently test the animals. Following the exposure, these animals typically lose the ability to escape something that normally they would very readily be able to escape, Dr. Duman observed.

“An animal that is exposed to an inescapable stressor fails 25 out of 30 times to escape that stressor,” he said. “If you administer desipramine or tranylcypromine for several days, you reduce that, block that”—in other words, almost back to the point at which the animal started prior to the stress.         

—Fred Balzac

Return to table of contents