Long-Term Study Confirms DST As Suicide-Prediction Tool

Which depressed persons will ultimately commit suicide? A few psychological or behavioral risk factors have emerged with consistency: hopelessness, suicidal ideation, a history of suicide attempts, being unmarried, living alone, and being male. But suppose doctors had a simple, inexpensive, biological test that could also help them identify such persons?

Some studies conducted during the past few years have suggested that such a test indeed exists. And now a new investigation bolsters the argument still further. It was conducted by William Coryell, M.D., a psychiatrist at the University of Iowa College of Medicine in Iowa City, and by Michael Schlesser, M.D., a psychiatrist with Neuropsychiatry Associates in Dallas. They report their favorable study results with the test in the May American Journal of Psychiatry.

Depressed persons in especially great danger of committing suicide, various researchers have suspected, might have abnormally high levels of three interacting hormones that mirror the brain’s and body’s reaction to stress. The hormones are the adrenal glands’ cortisol, the pituitary gland’s adrenocorticotropic hormone (ACTH), and the hypothalamus’s corticotropin-releasing factor (CRF). For instance, when the bodies of persons who had died by suicide were compared with those of persons who had died by other violent means, the former were found to have heavier adrenal glands and also higher levels of CRF in their cerebrospinal fluid. So some of these investigators decided to conduct studies to learn whether the dexamethasone suppression test (DST), which measures abnormally high levels of cortisol in the body (and thereby indirectly also abnormally high levels of ACTH and CRF), might predict suicide in depressed persons. Some of these studies (although not all) produced favorable results, suggesting that the DST might indeed be able to predict suicide.

One drawback of these investigations, however, is that they were not conducted over the long term. Another drawback was that the apparent predictive value of the DST had not been compared with known psychological and behavioral suicidal risk factors—that is, to known psychological and behavioral predictors. So Coryell and Schlesser decided to conduct a study that would both test the value of the DST over a 15-year period and pit the value of the DST against psychological and behavioral predictors.

Between 1978 and 1981, they recruited 78 patients for their study. The patients sought either inpatient or outpatient treatment at the University of Iowa for conditions that met research diagnostic criteria for major depressive disorder, mania, or schizoaffective disorder–depressed type. In addition to providing psychological and behavioral data that indicated whether they might be especially in danger of committing suicide, such as hopelessness, previous suicide attempts, living alone, and so forth, the patients were given the DST.

During the test, each patient took 1 mg of dexamethasone orally at 11 p.m., then provided blood samples the next day at 8 a.m. and/or 4 p.m. Cortisol levels in the blood samples were then determined. A cortisol value greater than 5 μg/dl indicated that dexamethasone had failed to suppress cortisol in the patient’s blood—in other words, that test results were abnormal and that the patient’s cortisol level was abnormally high. Thirty-two of the 78 patients, it turned out, showed abnormal DST results.

Coryell and Schlesser then followed the fate of the 78 patients over the next 15 years to see how many committed suicide, whether the DST results were predictive of suicide, and how DST predictivity compared with predictivity via psychological and behavioral risk factors. (The study did not influence any depression treatments that the patients received throughout these years, although such therapies were noted.)

Of the 78 patients, eight were found to have committed suicide. DST results for seven of the eight had been abnormal. Also telling, survival analyses showed that the estimated risk for eventual suicide in the abnormal DST group was 27 percent, compared with only 3 percent among patients who had normal DST results. When the known psychological and behavioral risk factors for suicide in the 78 patients were examined, none was found to distinguish significantly the eight patients who later committed suicide from those who did not.

Yet another intriguing aspect about the DST that emerged from the study is that its ability to predict suicide seemed to remain robust as the years went by.

If the DST truly pans out after further study as an effective suicide predictor in depressed patients, it is probably something that psychiatrists could use in their clinical practices, Coryell indicated.

The study, “The Dexamethasone Suppression Test and Suicide Prediction,” is posted on the journal’s Web site at http://ajp.psychiatryonline under the May issue. ▪