CNS Spectr. 2008;13(8):682-687
Faculty Affiliation and Disclosure
Dr. Belmaker is assistant director at Beer-Sheva Mental Health Center at Ben Gurion University of the Negev in Israel.
Faculty Disclosure: The author does not have an affiliation with or financial interest in any organization that might pose a conflict of interest. This article was presented as a lecture given in Tel Aviv, Israel, on November 21, 2007. Parts of this article have been previously published in Hebrew.
Acknowledgment: The author would like to dedicate this article to the memory of William Pollin, MD, his mentor at The National Institute of Mental Health, whose skepticism never reduced his enthusiasm.
Submitted for publication: April 7, 2008; Accepted for publication: July 23, 2008.
Please direct all correspondence to: Robert H. Belmaker, MD, Beer-Sheva Mental Health Center, PO Box 4600, Beer-Sheba, Israel; Tel: 972-7-640-1602, Fax: 972-7-640-1621; E-mail: firstname.lastname@example.org.
• Monoamines in animal models influence multiple behaviors.
• The placebo-drug difference in controlled studies suggests that perhaps only one third of depressions are monoamine related.
• Cortisol is not consistently elevated in depression and, when it is, it is a small elevation compared with a stressor such as public speaking, for example.
• The dexamethasone suppression test is abnormal only in melancholic depression, whereas psychosocial stress is most associated with chronic, milder depression.
Along with the development of selective serotonin reuptake inhibitors there has been a tremendous widening of the definition of depression and an impressive decrease in the placebo-drug difference in controlled studies. In the early 1960s, about one third of depressed patients improved with placebo and two thirds with active compounds. Current controlled studies suggest that the situation has certainly not improved. The Sequenced Treatment Alternatives to Relieve Depression Study found that response rates to new compounds after the failure of the first antidepressant are low. The monoamine hypothesis of depression was formulated in the mid 1960s based on the antidepressant efficacy of the monoamine reuptake inhibitors, monoamine oxidase inhibitors, and the depressogenic effects of reserpine as a monoamine depleter. However, no monoamine-related finding has been found that is diagnostic for depression. A second major hypothesis regarding depression has been the stress cortisol hypothesis. However, blood cortisol levels are not diagnostic of depression. Psychiatric clinicians are convinced that there are patients for whom antidepressants have made the difference between life and death. However, physicians may generalize unjustifiably based on single dramatic cases to a much larger diagnostic group. Perhaps there are many causes of different types of human sadness, and perhaps only some of these involve mechanisms related to monoamines. Thus, perhaps only some kinds of depression are responsive to monoamine affecting antidepressants.
No one in 1950 would have predicted the development of the soon to be discovered antidepressant and antipsychotic compounds. No developments in basic neuroscience preceding these discoveries heralded their discovery, which was very much serendipitous.1 The pipeline of pharmaceutical companies has been thinning out in recent years and there seems to be fewer and fewer major discoveries of the kind that have made psychopharmacology what it is today. The reasons for this were reviewed by Klein.2 However, it has become quite fashionable to talk about imminent breakthroughs and major new treatments and perhaps whole new paradigms based on basic sciences. Deep brain stimulation, neurogenesis stimulators, and glutamate receptor antagonists all have reasonable bases in neuroscience and could provide revolutionary new approaches. Whether they will do so or not is almost impossible to predict. The incredible complexity of the brain makes it difficult to know at what point advances in neuroscience understanding will bring human treatments.
The 50 years since the first antidepressants, imipramine and iproniazid were introduced, have been characterized by a plethora of new antidepressants, almost yearly. Table 1 shows the introduction of these compounds from 1950–2000. Given that imipramine inhibits the reuptake of serotonin and noradrenaline it was a reasonable hypothesis that selective serotonin reuptake inhibitors (SSRIs) might be cleaner or more effective drugs than imipramine. Most experts feel that this has not turned out to be the case. While a common belief suggests that SSRIs have fewer side effects than tricyclic antidepressants, this is to a large extent dependent on what side effect scale is used; in one that weights heavily on anticholinergic side effects, such as dry mouth, constipation, and blurring of vision, SSRIs are clearly superior. However, if a scale is devised that measures gastrointestinal distress, anxiety, insomnia, and sexual dysfunction then SSRIs come out looking poorly. Community studies do not always show that patient satisfaction is greater with SSRIs or that dropout rates of treatment are lower. Turner and colleagues3 suggest that along with the development of SSRIs there was a tremendous widening of the definition of depression and an impressive decrease in the placebo-drug difference in controlled studies.
Many new antidepressants have become blockbuster new compounds but have not necessarily been clear improvements over previous treatments. Looking back on the summary of the early 1960s that ~33% of depressed patients improve with placebo and 66% with active compounds, current controlled studies suggests that the situation has not really improved. Most senior clinicians looking over Table 1 will not see marked progress over the last 50 years. It is difficult to find a table such as Table 1 in a textbook as most texts present the compounds not chronologically but by mechanism of action at the synapse, suggesting great progress. The great variety of compounds may not reflect a more powerful clinical armamentarium. Some compounds may be effective after a previous antidepressant has failed, but the extensive Sequenced Treatment Alternatives to Relieve Depression (STAR*D) Study4 found that response rates to new compounds after the failure of the first antidepressant are low.
As a basic understanding of depression the monoamine hypothesis of depression was formulated in the mid 1960s based on the antidepressant efficacy of the monoamine reuptake inhibitors, monoamine oxidase inhibitors and the depressogenic effects of reserpine as a monoamine depleter.5 Figure 1 summarizes recent research on this hypothesis. This hypothesis continues to stimulate excellent research, and each time technology opens a new window into the brain, evidence6 is found supporting the concept that monoamines are related to depression. However, no finding has been found that is diagnostic for depression. Figure 2 illustrates some results of a study of monoamine oxidase density in vivo in the brain of depressed patients using a kind of positron emission tomography scan.6 While monoamine oxidase is 30% increased on the average in many different regions of the brain in depressed patients and this effect is statistically significant, Figure 2 shows an overlap between depressed patients and controls and this test could not be used in any way as diagnostic of depression. This despite its great technological advancement and the fact that the patients of the study were carefully chosen to have diagnosable depression. Table 2 summarizes the problems with the monoamine hypothesis today.
A second major hypothesis over the last 50 years regarding depression has been the stress cortisol hypothesis. The fact that stress seems to predispose to depression has always been obvious to clinicians although stress can predispose to many psychological and physical disorders as well. Since cortisol is a major hormone of the stress system, it seemed logical to study cortisol and its regulation in depressed patients. However, early on, it was clear that the relationship of stress to depression is not a one-to-one relationship. Many patients develop depression “out of the blue” with no obvious precipitating stress and many people have terribly stressful lives without developing a depression. Moreover, sometimes banal stresses, such as public speaking or parachuting, can increase blood cortisol by orders of magnitude greater than those found in depressed patients. As in the case of the monoamine hypothesis, each technological advancement that has led to more sophisticated tests of the cortisol theory has found some evidence to support the relationship of cortisol to depression. For example, the adrenal cortex that secretes cortisol is slightly enlarged on sophisticated scans of depressed patients. The anterior pituitary, where adrenocorticotropic hormone is secreted, is slightly enlarged as well. The hippocampus, which is known from animal studies to be sensitive to excess cortisol and reacts to such excess by atrophy, is reduced in size on brain imaging studies of depressed patients. Cortisol-releasing factor (CRH), which is a small peptide released by the hypothalamus to stimulate the anterior pituitary to secrete adrenocorticotropic hormone is increased according to one report in the CSF of depressed patients. Moreover, the CRH system and its receptors are present in many regions of the brain outside the hypothalamus and seem to mediate higher-level stress responses and stress perception in the human brain. There are reports that CRH receptors and expression are increased in the brain of depressed patients even above the level of the hypothalamus. Some evidence also suggests that depressed patients or individuals prone to depression perceive threat differently as measured by blood-flow responses of the amygdala and cingulate gyrus as determined with functional magnetic resonance imaging.
Figure 3 summarizes the hypothalamic-pituitary-cortisol system in depression. As with the monoamine hypothesis, it is important to note that blood cortisol levels are not diagnostic of depression or of any other psychiatric abnormality. Usually articles on this subject give only mean values and those with a large enough population report that the P value is <.05 or .01, and is, therefore, statistically significant. Figure 4 is adapted from one of the few articles showing the actual data. It can be seen that cortisol levels while significantly elevated at both their low point and high point of the day in the depressed patients are greatly overlapping with normal controls and plasma cortisol could not be used as a diagnostic test for depression. Table 3 summarizes the problems with cortisol hypothesis of depression.
An impending storm in depression research that may relate to the future of depression psychopharmacology has to do with the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition attempt to diagnose depression without relation to the psychosocial context in which it occurs. DSM-IV and DSM-III before it based the diagnosis of depression on low mood distinct from that in previous periods of life and accompanied by a number of psychomotor and psychophysiological symptoms. This atheoretical approach without a psychosocial context was partly based on the fact that it was difficult in research in the 1970s to define the relationship of depression to preexisting stress or trauma. The one exception in DSM-IV is the bereavement exclusion as discussed by Wakefield and colleagues.7 Recent bereavement has been accepted by DSM-IV as a condition that can lead to symptoms that meet the criteria for depression without being diagnosed as major depression. However, Wakefield and colleagues7 showed that numerous other psychosocial stresses do not behave differently than bereavement in terms of the symptomatology and prognosis of depression. Therefore, Wakefield and colleagues7 suggested that bereavement not be an exclusion or that other psychosocial stresses constitute a similar exclusion.
Horwitz and Wakefield8 extended their argument into a critique of the modern psychiatric view of depression. They stated that psychiatry has extended the boundaries of depression so widely that all of human sadness has been labeled and lost its meaning as a normal human emotion essential for people to deal with. This philosophical critique comes at the same time as recent influential reviews on the decreasing utility of antidepressants in clinical trials.3,9 Every psychiatric clinician is convinced that there are patients for whom antidepressants have made the difference between life and death. However, it is well known that physicians sometimes extrapolate single dramatic cases to a much larger diagnostic group.10 Could this be the case in depression? Perhaps there are many causes of different types of human sadness, and perhaps only some of these involve mechanisms related to monoamines. Thus, perhaps only some kinds of depression are responsive to monoamine-affecting antidepressants.
The future of depression psychopharmacology is that we will be spending much time in the near future dealing with this issue. The discouraging results reported in the STAR*D community trial of the outcome of depression treatment have contributed to this stocktaking. The slowly accumulating body of literature proving the effectiveness of certain kinds of psychotherapy for depressed patients has also undermined the self-confidence of a narrow psychopharmacologic approach. Table 4 summarizes some thoughts about future directions in psychopharmacology. Of course, it is difficult to separate wishes from predictions.
As a final statement, I would like to highlight a critical piece of advice that I received from William Pollin, MD, early in my career. It seems that patients would be best served by reducing the extent to which opinion leaders in our field have been devoted to the hypotheses that they are studying. The art of psychiatric research over the long-term is to avoid religious devotion to the unproven hypothesis without becoming cynical or giving up the effort.
1. Ban TA. The role of serendipity in drug discovery. Dialogues Clin Neurosci. 2006;8:335-344.
2. Klein DF. The loss of serendipity in psychopharmacology. JAMA. 2008;299:1063-1065.
3. Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R. Selective publication of antidepressant trials and its influence on apparent efficacy. N Engl J Med. 2008;358:252-260.
4. Nelson JC. The STAR*D study: a four-course meal that leaves us wanting more. Am J Psychiatry. 2006;163:1864-1866.
5. Belmaker RH, Agam G. Major depressive disorder. N Engl J Med. 2008;358:55-68.
6. Meyer JH, Ginovart N, Boovariwala A, et al. Elevated monoamine oxidase a levels in the brain: an explanation for the monoamine imbalance of major depression. Arch Gen Psychiatry. 2006;63:1209-1216.
7. Wakefield JC, Schmitz MF, First MB, Horwitz AV. Extending the bereavement exclusion for major depression to other losses: evidence from the National Comorbidity Survey. Arch Gen Psychiatry. 2007;64:433-440.
8. Horwitz AV, Wakefield JC. The Loss of Sadness: How Psychiatry Transformed Normal Sorrow into Depressive Disorder. New York, NY: Oxford Press; 2007.
9. Lacasse JR, Leo J. Serotonin and depression: a disconnect between the advertisements and the scientific literature. PLoS Med. 2005;2:e392.
10. Groopman J. How Doctors Think. New York, NY: Mariner Books; 2008.
11. Carroll BJ, Cassidy F, Naftolowitz D, et al. Pathophysiology of hypercortisolism in depression. Acta Psychiatr Scand Suppl. 2007;(433):90-103.
12. Yehuda R, Teicher MH, Trestman RL, Levengood RA, Siever LJ. Cortisol regulation in posttraumatic stress disorder and major depression: a chronobiological analysis. Biol Psychiatry. 1996;40:79-88.