Until recently, the laboratory had relatively little to offer in psychiatry. Laboratory tests were used mainly to diagnose or exclude organic illness. For example, in one study about 5% of patients with dementia had organic diseases such as hyponatremia, hypothyroidism, hypoglycemia, and hypercalcemia; about 4% were caused by alcohol; and about 10% were due to toxic effects of drugs. A few psychiatric drug blood level assays were available, of which lithium was the most important. In the 1970s, important work was done suggesting that the neuroendocrine system is involved in some way with certain major psychiatric illnesses. Thus far, melancholia (endogenous psychiatric depression or primary depression) is the illness in which neuroendocrine abnormality has been most extensively documented. It was found that many such patients had abnormal cortisol blood levels that were very similar to those seen in Cushing’s syndrome (as described in the chapter on adrenal function) without having the typical signs and symptoms of Cushing’s syndrome. There often was blunting or abolition of normal cortisol circadian rhythm, elevated urine free cortisol excretion levels, and resistance to normally expected suppression of cortisol blood levels after a low dose of dexamethasone.

Because of these observations, the low-dose overnight dexamethasone test, used to screen for Cushing’s syndrome, has been modified to screen for melancholia. One milligram of oral dexamethasone is given at 11 P.M., and blood is drawn for cortisol assay on the following day at 4 P.M. and 11 P.M. Normally, serum cortisol levels should be suppressed to less than 5 µg/100 ml (138 nmol/L) in both specimens. An abnormal result consists of failure to suppress in at least one of the two specimens (about 20% of melancholia patients demonstrate normal suppression in the 4 P.M. specimen but no suppression in the 11 P.M. specimen, and about the same number of patients fail to suppress in the 4 P.M. specimen but have normal suppression in the 11 P.M. sample). The psychiatric dexamethasone test is different from the dexamethasone test for Cushing’s syndrome, because in the Cushing protocol a single specimen is drawn at 8 A.M. in the morning after dexamethasone administration.

The Cushing’s disease protocol is reported to detect only about 25% of patients with melancholia, in contrast to the modified two-specimen psychiatric protocol, which is reported to detect up to 58%. Various investigators using various doses of dexamethasone and collection times have reported a detection rate of about 45% (literature range, 24%-100%). False positive rates using the two-specimen protocol are reported to be less than 5%. Since some patients with Cushing’s syndrome may exhibit symptoms of psychiatric depression, differentiation of melancholia from Cushing’s syndrome becomes necessary if test results show nonsuppression of serum cortisol. The patient is given appropriate antidepressant therapy and the test is repeated. If the test result becomes normal, Cushing’s syndrome is virtually excluded.

Various conditions not associated with either Cushing’s syndrome or melancholia can affect cortisol secretion patterns. Conditions that must be excluded to obtain a reliable result include severe major organic illness of any type, recent electroshock therapy, trauma, severe weight loss, malnutrition, alcoholic withdrawal, pregnancy, Addison’s disease, and pituitary deficiency. Certain medications such as phenobarbital, phenytoin (Dilantin), steroid therapy, or estrogens may produce falsely abnormal results.

At present, there is considerable controversy regarding the usefulness of the modified low-dose dexamethasone test for melancholia, since the test has a sensitivity no greater than 50% and significant potential for false positive results.

Besides the overnight modified low-dose dexamethasone test, the thyrotropin-releasing hormone (TRH) test has been reported to be abnormal in about 60% of patients with primary (unipolar) depression. Abnormality consists of a blunted (decreased) thyrotropin-stimulating hormone response to administration of TRH, similar to the result obtained in hyperthyroidism or hypopituitarism. However, occasionally patients with melancholia have hypothyroidism, which produces an exaggerated response in the TRH test rather than a blunted (decreased) response.

One investigator found that about 30% of patients with melancholia had abnormal results on both the TRH and the modified dexamethasone tests. About 30% of the patients had abnormal TRH results but normal dexamethasone responses, and about 20% had abnormal dexamethasone responses but normal TRH responses. The TRH test has not been investigated as extensively as the modified dexamethasone test.

A more controversial area is measurement of 3-methoxy-4-hydroxyphenylglycol (MHPG) in patients with depression. One theory links depression to a functional deficiency of norepinephrine in the central nervous system (CNS). 3-Methoxy-4-hydroxyphenylglycol is a major metabolite of norepinephrine. It is thought that a significant part of urinary MHPG is derived from CNS sources (20%-63% in different studies). Some studies indicated that depressed patients had lower urinary (24-hour) excretion of MHPG than other patients, and that patients in the manic-phase of bipolar (manic-depressive) illness had increased MHPG levels. There was also some evidence that depressed patients with subnormal urinary MHPG levels responded better to tricyclic antidepressants such as imipramine than did patients with normal urine MHPG levels. However, these findings have been somewhat controversial and have not been universally accepted.