In Practice: Chicken and Egg

By Peter D. Kramer in In Practice

Does depression cause brain differences, or do brain differences cause depression? A scientist whose past research pointed to the latter conclusion has just published findings that reverse the direction.

The prevailing contemporary model for depression suggests that in vulnerable people, repeated stress gives rise to adverse changes in the brain; depression is itself a stressor. The primary evidence for this hypothesis comes from rodent studies, where early deprivation and later mild stress cause what look like mood changes - and shrinkage in areas of the brain that correspond to our hippocampus and prefrontal cortex. Human studies have tended to be correlational: patients who have suffered more days of mood disorder have more differences in brain volume. These findings are ambiguous. Perhaps a person with a small hippocampus is more prone to depression.

Evidence for the second theory, in which small hippocampal size creates vulnerability, came from the laboratory of the German psychiatrist , Thomas Frodl. He found that over the course of a year, chronic depression did not predict hippocampal change, but small hippocampal size predicted chronicity of depression. That's what makes a new study particularly intriguing. Looking now at a three-year interval, Frodl has found that depression does lead to loss of volume in critical regions of the brain.

The study contrasted 38 patients hospitalized for depression with 30 matched controls and used MRI scans coded by new software technology that distinguishes changes throughout the brain, not just in areas pre-selected by researchers. Over the next three years, the patients, and not the controls, showed a "decline in gray matter density" in the hippocampus, prefrontal cortex, and elsewhere. There was no part of the brain where the depressed patients showed increased volume or density. Patients whose mood disorder had remitted showed less volume decline. Overall, Frodl concluded, this prospective study in humans supports the findings from animal models in which stress and depression-like disorders cause brain change.

In the just-published study, antidepressants did not have a separate effect - it was getting better that conferred some protection. But the month before, in September, Frodl contributed an analysis showing that in some patients who took antidepressants for the whole of three years "hippocampal volumes increased significantly." A critical interactive factor may have been hippocampal size at the start of treatment. (It should be said that this study had a high attrition rate; its conclusions are necessarily speculative.) In the September paper, Frodl concluded that "a relatively small hippocampal volume may be a vulnerability factor for a bad treatment response in major depression."

Both may be true: depression attacks the hippocampus, and a small hippocampus impairs resilience. In Against Depression, I wrote that the evidence favored this combined hypothesis - so the theory is not new. But Frodl's current research does point to movement in the field toward the conclusion that depression or stress in the presence of depression injures the brain. That model suggests a need for vigorous early intervention and extended treatment. Recovery seems to offer some protection. Although the evidence is hardly clear-cut, long-term antidepressant use may as well.