Stress, memory and social support

By Sylvain Comeau

Mac Gill Reporter, September 26, 2000

Psychiatry professor Sonia Lupien

Specifically, the project looks at the hormone cortisol, which occurs naturally in response to stress.

"Cortisol is a stress hormone, and there are receptors for cortisol in the hippocampus, the region of the brain responsible for memory," McGill psychiatry professor Sonia Lupien explains. "Long-term exposure to these hormones can cause atrophy of the hippocampus, leading to memory impairment." Lupien adds that hippocampal damage is one of the early signs of Alzheimer's.

Lupien, a Douglas Hospital researcher, is working with McGill colleagues N.P.V. Nair (psychiatry), Michael Meaney (biology and psychiatry) and Jens Pruessner (psychiatry), as well as Charles Wilkinson of Washington University.

This cortisol-memory loss link was first demonstrated in 1968. In the late 1980s, Meaney at the Douglas performed the first animal model experiments. In the early 1990s, Lupien and her colleagues started to fill in the gaps of that early work, to see if the animal models were valid for humans.

"We followed a group of elderly people for five years, and we showed that about 30 per cent of the elderly population secretes too much cortisol; those who were exposed to it for 5 years showed memory impairment and atrophy of the hippocampus."

That study also demonstrated that the memory impairment may be reversible in the elderly; it is only irreversible when the damage has progressed past a "point of no return." The results of that study were published in 1998.

In the August 2002 issue of the Journal of Clinical Endocrinology and Metabolism, the team published the results of the latest stage, which provides proof of the cortisol-memory link. They conducted an experiment with two groups of elderly people who had been involved in the five-year study. One group had a moderate level of cortisol, and another had high levels of cortisol as well as impaired memory. Both groups were given metyrapone, a steroid that prevents the secretion of the hormone. Their memories were tested, and they then received a dose of hydrocortisone, a drug that restored cortisol to previous levels.

These results were compared with those obtained using placebos. The elderly people with moderate cortisol levels suffered memory loss after receiving the metyrapone: restoring cortisol levels also restored their memory. Those with a high level of cortisol saw no memory improvement, although restoring their cortisol level caused an even greater memory impairment.

"For those who are not at the 'too late' stage (of atrophy of the hippocampus), reducing cortisol pharmacologically should benefit their memory. However, someone who produces normal amounts of cortisol might suffer memory impairment if they had it reduced, because this hormone is necessary to life. We need it for many brain functions, including memory."

Lupien emphasizes that stress is not the villain here, but rather the excess of it, specifically the excess of cortisol. "We need stress in our lives; without it, we couldn't even wake up in the morning."

However, the memory-impaired group in the study probably suffered from too much stress in their lives.

"When we inhibited the hormone in these people, it did nothing, so they were in the 'too late' stage. Since their memory was further impaired by the replacement of cortisol, in real life that would show they are very sensitive to little stressors. Their memory is easily affected at that point."

Future drug therapies to emerge from this work will focus on cortisol reduction before permanent damage can occur.

"We have shown that we can move memory up and down with changes in this hormone. We did it because we know that at one point in life, some elderly people produce too much cortisol. The question now is: what is the time window? If we can help these people at the time when they start producing too much of it, decreasing it at that point should be beneficial. This will be the third stage of our research." To that end, the team is examining people from 18 to 77 years to pinpoint the age-related time window.

Drug treatments probably will not emerge for at least another five years, when the study is completed. The goal then would be to identify and aid "at risk" individuals.

"If we can identify certain cortisol levels which reliably indicate memory impairment at different ages, that could be detected by a simple blood test."

Thus, routine screening for cortisol related memory problems could become the norm in the near future.

While a pill may one day fix the problem, Lupien points out that improvements to the social conditions of the elderly may help today. The team recently conducted a research study, not yet published, showing a direct link between memory and social support.

"We know that 30 percent of elderly people start to produce too much of this hormone; to find out why, it was natural to look at their environment, since we are talking about a stress hormone. We tested cortisol levels in about 70 elderly people every day for a year. We found that the only environmental factor related to high cortisol levels was a lack of social support.

"This finding makes sense, because social isolation in our elderly is something very common, and it is well known that social support is important in dealing with stress."

Changing a stressful environment may be more effective, and even cheaper, than any drug treatments.

"We can eventually have an effective drug, but if lack of social support is the problem, maybe changing social policies would be much cheaper. And it would address a root cause, rather than a symptom."

The environment study was part of a search for more "soft" variables to explain memory loss. Another key intangible emerged in a recent experiment, conducted with Jens Pruessner, a Montreal Neurological Institute researcher.

"We conducted brain scans on 23 elderly people, measuring the volume of their hippocampus; at the same time, we measured a number of personality variables. We were asking, can the way you are also be a predictor of what is going to happen to your memory? We found that the factor strongly associated with the volume of your hippocampus, later in life, is self-esteem."

This finding dovetails with the social support discovery.

"It makes sense because someone with high self-esteem will be more out-going and tend to find more social support...so we started by showing that this hormone could be bad for your brain, and now we are backtracking: trying to find the origin."

The project is sponsored until 2006 by a Canadian Institutes of Health Research (CIHR) grant.