SAN FRANCISCO—For the past three years, a symposium on cerebrovascular disease and late-life depression has been held at the American Association for Geriatric Psychiatry’s annual meeting. The emphasis in previous years has been on the concept of vascular depression—specifically, as George Alexopoulos, MD, put it, “Does brain destruction associated with aging—or with conditions that aging brings with it, such as vascular conditions—contribute to the development of depression?” This year, however, the organizers of the symposium elected to focus instead on new findings relating neuroanatomic, immunologic, and neurotransmitter changes to late-life depression.

ANATOMIC CHANGES IN DEPRESSION

The first speaker, David Steffens, MD, offered a brief review of findings from studies that have examined the anatomic abnormalities associated with late-onset depression (usually defined as depression first occurring after age 60). These studies have found that alterations in subcortical white matter are associated with executive dysfunction and reduced speed of cognitive processing; moreover, subcortical gray matter lesions have been linked to increased depression severity and to relatively poor treatment response and long-term outcome. More recently, Dr. Steffens, who is Assistant Professor of Psychiatry at Duke University Medical Center in Durham, North Carolina, conducted a study in which he and his colleagues performed magnetic resonance imaging on 88 elderly patients with depression and 47 nondepressed age-matched controls. Statistical parametric mapping revealed that, compared to controls, the depressed group had increased lesion severity in the white matter of the medial orbital prefrontal area, a region important in planning and other executive functions. Moreover, the severity of depressive symptoms correlated with the severity of lesions in this area, Dr. Steffens said. The findings point to a link between late-life depression and executive dysfunction, a theme Dr. Alexopoulos would revisit in his talk.

Similarly, in recent years, several studies have suggested that late-life depression is a risk factor for dementia. The link between the disorders is not well understood, but Dr. Steffens noted that the two conditions might be related by a common process of neurodegeneration or perhaps by a gradual worsening of cerebrovascular disease. He described a second study in which he and his colleagues examined neuroanatomic alterations in 182 elderly patients with depression who did not have dementia at the beginning of the study. Controlling for such variables such as age, sex, and baseline Mini-Mental State Examination (MMSE) scores, the investigators found that the volume of subcortical gray matter lesions at baseline significantly predicted the subsequent development of dementia. A preliminary analysis suggested that the burden of white matter lesions at baseline also predicted dementia risk.

Dr. Steffens emphasized that these studies constitute only a preliminary assessment of the relationships among changes in brain anatomy, symptoms of depression, and dementia; additional longitudinal studies are needed to clarify the nature of these relationships, as are intervention studies examining factors such as blood pressure control, cholesterol levels, and the prophylactic use of aspirin.

THE POTENTIAL ROLE OF SYSTEMIC ILLNESS

The symposium’s next speaker, Jeffrey M. Lyness, MD, discussed evidence suggesting a link between systemic illness, such as atherosclerosis, and the subsequent development of depression. Dr. Lyness, who is Associate Professor of Psychiatry at the University of Rochester Medical Center, outlined a preliminary model relating these factors but emphasized that it is only a first step toward a complete understanding of the link between depression and systemic illness; although many of the individual components of the model have been documented in preclinical and clinical studies, Dr. Lyness and colleagues are only beginning to collect pilot data to test the full model.

“First of all, we need to remember that atherosclerosis is a systemic condition that is really quite complex,” Dr. Lyness explained. “It’s more than just lipids and platelets.” Central to the atherosclerotic process are complex immune and inflammatory responses in the affected vessels. One of the many substances produced during this process is the cytokine interleukin-1ß (IL-1ß). Dr. Lyness and colleagues have been particularly interested in IL-1ß because it has been linked to the so-called “acute sickness response”—the subjective feeling of sickness that accompanies many illnesses. In addition, IL-1ß is able to cross the blood-brain barrier. Animal studies have demonstrated that IL-1ß stimulates central monoamine activity and have suggested that chronic stimulation of monoamine systems by IL-1ß leads to a “burned out” state of diminished monoamine functioning. “So, you start off with atherosclerosis systemically and you wind up with hypofunctioning monoamine systems, presumably including norepinephrine, serotonin, and other systems that are involved in depression,” Dr. Lyness said.

Dr. Lyness described a small pilot study designed to begin evaluating this model in humans. Thirty-seven patients age 50 or older, all with coronary artery disease, underwent an extensive evaluation. Surprisingly, although plasma IL-1ß levels were related to patients’ overall degree of illness, they were not significantly related to scores on the Hamilton Rating Scale for Depression (HAM-D) or to whether patients met diagnostic criteria for depression, findings that are at odds with the IL-1ß model. Nonetheless, the investigators are currently recruiting patients for a larger study that will include a control group of patients who do not have coronary artery disease.

Dr. Lyness noted that even if a correlation between depression and IL-1ß status is established, the issue of causality might remain unclear. For example, although chronic illness may promote depression by increasing the production of IL-1ß, it is also possible that preexisting depression exacerbates atherosclerosis, stimulating cytokine production.

SEROTONIN RECEPTOR FUNCTION IN DEPRESSION

Positron emission tomography (PET) imaging is usually thought of as a tool to study the metabolic activity of various brain regions. But PET scanning can also provide an effective means of examining abnormalities in neurotransmitter functioning, noted Yvette I. Sheline, MD, Associate Professor of Psychiatry, Radiology, and Neurology at Washington University School of Medicine in St. Louis. For example, PET can be used to test the hypotheses that people with depression exhibit abnormal function of serotonin (5-hydroxytriptamine [5-HT]) receptors and that treating early-onset depression with antidepressant medications leads to improved mood in part by down-regulating 5-HT receptors. Conversely, one might hypothesize that treatment has little impact on receptor function in late-onset depression, which is associated with poor outcome.

To explore these questions, Dr. Sheline and colleagues used a highly specific ligand (altanserin) to examine postsynaptic 5-HT_2A receptors in 44 patients with major depression and 29 controls. PET imaging was conducted at study entry and, for depressed patients, again after eight weeks of treatment with sertraline or desipramine. Dr. Sheline noted that recruitment for the study was difficult: Although the investigators wanted to find patients with vascular depression, they also wanted to exclude those who were using ß-blockers or calcium channel blockers, or who had serious concomitant medical conditions, such as diabetes. Still, they managed to assemble a cohort that included 19 depressed patients older than age 50; 11 of these had early-onset depression and eight had late-onset depression. For the entire sample of depressed patients, the mean HAM-D score was 23 at baseline and 11 after treatment.

“Before we could really look at late-life depression, we first had to understand the effects of aging” on 5-HT receptors, Dr. Sheline noted, as previous postmortem studies suggested that at least half of the brain’s 5-HT receptors are lost with aging. The wide range in patient age in the current study (21 to 86) provided the investigators with an opportunity to examine this issue, albeit in a cross-sectional rather than longitudinal manner. An analysis of the PET images revealed a large “decline” in the number of 5-HT_2A receptors with age: The rate of loss was roughly 16% per decade, with a maximum total loss of about 70%. Dr. Sheline noted that this decline is not simply due to a loss of brain tissue; although brain mass decreased somewhat with age, this change was far too small to account for such a large decline in the number of 5-HT_2A receptors. The analysis also suggested that the decrease in the number of receptors is not linear; it is fastest during early and middle adulthood and becomes more gradual after age 50.

The investigators next examined differences in brain 5-HT_2A receptors between depressed patients and control subjects. Depressed patients had significantly reduced 5-HT_2A binding in portions of the cortex that belong to the limbic system, which regulates emotion and mood states, but not in portions of the brain outside the limbic system, Dr. Sheline reported. Moreover, receptor binding was lower in patients with late-onset depression than in those with early-onset illness. Contrary to expectations, however, there was no significant change in the number of receptors after eight weeks of treatment in either the early- or late-onset depression groups. Thus, Dr. Sheline said, the very large and widespread decrease in 5-HT_2A receptor number that occurs with aging appears to overshadow any down-regulation from drug treatment, at least in the short term.

DEPRESSION AND EXECUTIVE DYSFUNCTION

The final speaker, Dr. Alexopoulos, discussed the relationship between depression and impaired executive function in older adults, suggesting that both phenomena might be connected to the same age-related changes in brain structure. Moreover, he said, patients with this dual presentation may represent a subgroup with specific treatment needs.

The common link between depression and executive impairment may be abnormalities in the striatofrontal system, noted Dr. Alexopoulos, who is a Professor of Psychiatry at Weill College of Medicine in New York City. Previous research has suggested that stroke of the basal ganglia or the frontal pole contributes to both depression and loss of executive function, and a similar profile is associated with subcortical disorders such as Parkinson’s disease and Huntington’s disease. Structural neuroimaging studies have found increased white matter lesions in frontal subcortical regions of patients with late-life depression; severe lesions tend to be associated with executive dysfunction. Finally, functional imaging of depressed adults has revealed hypometabolism of the left caudate nucleus and the cingulate gyrus, as well as abnormal activation of prefrontal regions, the amygdala, and paralimbic structures.

When these abnormalities result not only in depression but in executive dysfunction, the prognosis tends to be substantially poorer, Dr. Alexopoulos said. Studies conducted by his own group have shown that depressed patients with executive dysfunction are more likely than those with normal executive function to have a poor treatment response (one trial used nortriptyline, the other, citalopram), and two new studies have shown that executive dysfunction is associated with an increased risk of early relapse.

To further explore the clinical characteristics of these patients, Dr. Alexopoulos and his colleagues recently assessed executive dysfunction and psychomotor retardation (believed to be an expression of striatofrontal abnormalities) in 126 depressed, nondemented elderly persons. Patients were divided into two groups according to their scores on the initiation/perseveration portion of the Mattis Dementia Rating Scale; those who scored below the median value for healthy individuals were classified as having executive dysfunction and the remainder were considered cognitively intact. Depression severity and MMSE scores were similar between the two groups; however, patients with executive dysfunction were more likely to exhibit reduced interest in activities and to have psychomotor retardation. They also had higher scores on measures of paranoia and depersonalization. “This presentation, I would argue, is at least theoretically consonant with what you would expect in a person with striatofrontal dysfunction,” Dr. Alexopoulos said.

The investigators also found a significant interaction between levels of depression and executive function with regard to patients’ activities of daily living. That is, the severity of depression was a more important determinant of impairment in patients who also had executive dysfunction than in those with depression alone. Interestingly, a follow-up study has found an inverse relationship between depressive symptoms and initiation/perseveration scores over time: When depression begins to improve, the executive dysfunction worsens. In contrast, there was no association between overall cognitive dysfunction and severity of depression over time.

Recent evidence, then, suggests that “depression with executive dysfunction should be considered a viable entity that requires further investigation,” Dr. Alexopoulos concluded. Because poor treatment response seems to be associated with this clinical profile, he added, clinicians may wish to consider nonstandard approaches; although there are no data yet to support their use, dopaminergic agents and psychostimulants may be worth investigating.

—Mark Bowes, PhD

Return to table of contents