STANFORD -- The first direct evidence that stress can shrink a crucial part of the human brain is being compiled by scientists using new, high-resolution magnetic resonance imaging (MRI) scans, according to a Stanford expert on stress and the brain.

In a review article in the Aug. 9 edition of the journal Science, biological sciences Professor Robert Sapolsky said that the work of several research groups shows links between long-term stressful life experiences, long-term exposure to hormones produced during stress, and shrinking of the part of the brain involved in some types of memory and learning.

Sapolsky studies the effects of stress and stress hormones on wild baboons in Africa and on rats in his Stanford laboratory. He is the author of a popular book, Why Zebras Don't Get Ulcers, on the physiology of the stress response. He said that for 20 years, he and other stress physiologists have wished for a direct way to study the effects of stress on the human brain.

Research by Sapolsky and others has shown that some of those hormones, called glucocorticoids, spell bad news when brain cells are exposed to them for a long time ­ at least in the brains of rats. Glucocorticoids can cause rats' brain cells to shrivel, as the dendrite branches that they use to communicate with other neurons wither away. Prolonged exposure can kill the neurons or make them vulnerable to destruction during a brain injury or stroke.

The researchers also know that long-term exposure to stress hormones is a fact of life for some animals. Studying a troop of wild baboons, Sapolsky has shown that the same glucocorticoids that flood the bloodstream during a stressful event remain at high levels for months or years if the baboon has a stressful life ­ for example, if he's always in fear of an attack by the dominant male in his troop.

The parallels are obvious with stressful human lives. But there has been no way to poke inside the living human brain and see if our neurons are more robust than a rat's, or if stress hormones actually can damage our brains.

Now, thanks to improvements in MRI, researchers can make clear images of specific parts of the brain. In his Science article, titled "Why Stress Is Bad for Your Brain," Sapolsky summarizes what has been found so far as scientists tune up high-resolution MRI to take pictures of the hippocampus. The hippocampus is the region of the brain responsible for explicit, declarative memory ­ for knowing a fact like an address or the name of a friend, and knowing that one knows it. Its neurons are rich in glucocorticoid receptors; this is the region where animal studies have shown that stress hormones can damage neurons.

Because depression can raise the levels of glucocorticoids in the blood, Yvette Sheline and her colleagues at Washington University in St. Louis compared the hippocampi of people who had recovered from long-term, major depression with controls matched to them by age, education, gender and height. They found that the people with a history of depression had smaller hippocampi ­ averaging as much as 15 percent smaller in volume.

Tamara Gurvits and Roger Pitman of Harvard found that this region of the brain was 26 percent smaller in Vietnam veterans with post-traumatic stress disorder than in combat veterans without stress disorder. Douglas Bremner of Yale found a 12 percent atrophy in the left hippocampus of adults who suffered from post-traumatic stress disorder because of childhood sexual abuse. Sapolsky stressed that these findings do not prove that stress caused the brains to shrink ­ long-term prospective studies would be necessary to establish that, perhaps measuring the brains of soldiers before they go into combat and re-measuring them long afterward. Still, the evidence is mounting that stress affects human brains. "Each of the studies has some weaknesses," Sapolsky wrote in his Science review, "but they are countered by complementary strengths in the other studies."

None of the depression or trauma studies directly shows that glucocorticoids are the culprits in damaging human brains. However, a related study of people with Cushing's syndrome sheds some light on the topic.

Cushing's syndrome is caused by a tumor that stimulates the adrenal glands to produce large quantities of glucocorticoids. When Monica Starkman of the University of Michigan at Ann Arbor scanned the brains of people with this disorder, she found that the hippocampus had atrophied in the same areas where glucocorticoids were being over-produced. In a neighboring region of the brain that has few glucocorticoid receptors, there was no atrophy.

Is stress-linked brain damage permanent? Sapolsky said that in rats, short-term exposure to glucocorticoids causes neurons to shrink, but they rebound when levels of the hormones return to normal. Long-term exposure causes irreversible damage. The same effect seemed to occur in the study of people with Cushing's syndrome. Their brain atrophy was reversed when the tumor was removed, stopping the overflow of glucocorticoids.

However, rats lose some of their neurons permanently if they are exposed to the stress hormones over a long time. In a similar fashion, people with depression and post-traumatic stress disorder may have suffered permanent neuron loss. The studies detected atrophy in their brains many years after the damage must have occurred.

Sapolsky's advice to people leading everyday, stressful lives is not to strain themselves by worrying about stress too much.

"This fits what we know about animals," Sapolsky said. "It suggests that massive, massive amounts of stress or glucocorticoid exposure may have an effect on this part of the brain in humans. "But there's absolutely no evidence that ordinary stressors cause this much damage. And there's no evidence that stress causes Alzheimer's disease [which also strikes the hippocampus]," Sapolsky said.

If further studies support the link between glucocorticoids and brain damage, he said this may pose a difficult dilemma for people who take similar substances to treat autoimmune and inflammatory disorders like arthritis. Steroidal anti-inflammatory drugs like cortisone or prednisone are glucocorticoids.

In the doses that most people take ­ a small amount of cortisone cream to treat a rash or a nasal inhalant to keep an allergy at bay ­ these drugs probably pose no risk, Sapolsky said. However, some people with severe autoimmune diseases require large daily doses of anti-inflammatories over many years to keep their condition under control.

"This may mean there will be tough clinical decisions [for these people], with major minuses as well as major pluses to using these drugs," he said. "It is important to say that at this point, we don't know whether [anti-inflammatory] drugs have this effect. Studies are needed to find that out," Sapolsky said.