GCs and the Health of the Human Hippocampus

Work in the past decade suggests that these findings regarding the deleterious effects of GCs and stress apply to the human hippocampus. These findings have been most striking in a number of realms.

Cushing’s Syndrome. Cushing’s syndrome can arise from any of a number of different types of tumors and results in vastly elevated levels of circulating GCs. Magnetic resonance imaging (MRI) studies by Monica Starkman and colleagues of Cushingoid patients have shown selective atrophy of the hippocampus. More strikingly, it is the same patients who have the most severe GC hypersecretion, the most severe hippocampal atrophy, and the most profound cognitive deficits.

Post mortem studies to uncover whether this atrophy is due to neuronal loss, atrophy, or compression remain to be carried out. However, when the GC hypersecretion is corrected by removal of the tumor, the hippocampal gradually returns to normal size, in agreement with the model of reversible dendritic atrophy.

Exogenous GC Administration. Synthetic GCs are frequently administered in different realms of clinical medicine. Most cases involve transient exposure to fairly low concentrations, often through routes that do not result in much access of the steroids to the brain. However, there are hundreds of thousands of individuals on long-term, systemic highdose GC treatment to control various autoimmune or inflammatory diseases.

Difficult case-controlled work by Pamela Keenan and colleagues demonstrates that independent of the disease, treatment with such steroids increases hippocampal-dependent cognitive deficits. The mechanisms underlying this effect (e.g., whether it is due to dendritic atrophy, inhibition of neurogenesis, and so on) are unknown.

Clinical Depression. Major clinical depression is an archetypical example of a stress-related disease; as but three pieces of evidence, exogenous GC treatment increases the risk of depression, major stressors do as well, and about half of patients with major depression present with GC hypersecretion. Yvette Sheline and colleagues have observed that a history of prolonged depression is associated with selective hippocampal atrophy on MRI.

In this observation, since replicated, the more prolonged the depression, the more atrophy. While the necessary post mortem cell counting has not been carried out, the atrophy appears to be permanent in that it is demonstrable in individuals long after the depressions have resolved; whether this is due to loss of neurons or inhibition of neurons that would otherwise have been born is not clear.

Some researchers have suggested a scenario of reverse causality, in which individuals who (for other unrelated reasons) have a smaller than average hippocampus are more at risk for clinical depression. However, such volume loss is not demonstrable shortly after diagnosis, and more prolonged depressions are associated with more volume loss; both findings argue against reverse causality.

Posttraumatic Stress Disorder (PTSD). Probably the most contentious studies concerning GC effects on the human hippocampus concern PTSD. A large numbers of studies, with the first coming from J. Douglas Bremner and colleagues, demonstrate volume loss in the hippocampi of individuals with PTSD. Furthermore, the more volume loss, the more severe the cognitive impairments, and there is little evidence that either tends to reverse over time. The controversial issues are threefold.

First, are GCs the cause of the volume loss? This question is central to the second issue as well, but the first version of this question focuses on whether GC levels are elevated in PTSD. While it might seem intuitive that this is the case, many (but not all) studies suggest lower than normal levels and that such hyposecretion may even precede and predispose toward succumbing to PTSD.

The studies showing hyposecretion, pioneered by Rachel Yehuda and colleagues, also suggest an enhanced sensitivity to GCs in these patients. This would then shift the issue from whether excessive GC levels play a role in the volume loss to whether excessive sensitivity to GCs plays a role.

The second controversy concerns when the volume loss occurs. There are three possibilities:
(1) volume loss occurs in response to the trauma itself;
(2) volume loss occurs in response to the biology of the prolonged posttraumatic period (i.e., the period of the PTSD); and
(3) a small hippocampus (arising for reasons unrelated to stress or GCs) precedes and predisposes toward PTSD.

This issue is not resolved, and, at present, there are some striking findings supporting, as well as arguing against, each of these possibilities. One reasonable solution is that from the standpoint of hippocampal neurobiology, PTSD is a heterogeneous disorders, with all three occurring on occasion. Such heterogeneity is certainly seen in the clinical profile of the disorder.

The final issue is whether the volume loss is due to loss of neurons, blockage of birth of new neurons, atrophy of dendritic processes, loss of glia, and so on. As with the other realms of studies of the human hippocampus, resolving this issue requires difficult quantitative postmortem studies.

Obviously, many issues remain to be resolved and are likely to have some important consequences in neurology, psychiatry, and gerontology. Most broadly, the interrelated phenomena of GC neuroendangerment, inhibition of neurogenesis, dendritic atrophy, and neurotoxicity shift the realm of stress-related disease to the nervous system.

For nearly 70 years, stress physiology has taught that cardiovascular function, metabolism, and immunity can be sensitive to the adverse effects of stress. This newer literature demonstrates that this vital and fragile portion of our bodies, the brain, can be sensitive to stress as well.

Further Reading: Gould, E. and Gross, C. (2002). Neurogenesis in adult mammals: some progress and problems. Journal of Neuroscience 22, 619-623.
McEwen, B. (2002). Sex, stress and hippocampus: allostasis, allostatic load and the aging process. Neurobiology of Aging 23, 921.
Sapolsky, R. (1999). Stress, glucocorticoids and their adverse neurological effects: relevance to aging. Experimental Gerontology 34, 721.
Sapolsky, R. (2000). Glucocorticoids and hippocampal atrophy in neuropsychiatric disorders. Archives of General Psychiatry 57, 925