Psychiatric Annals


Hypotheses Regarding the Mechanism of Action of Electroconvulsive Therapy, Past and Present


Understanding how ECT works should provide remarkable insights into the biology of mood disorder and the other conditions that are responsive to ECT.


Understanding how ECT works should provide remarkable insights into the biology of mood disorder and the other conditions that are responsive to ECT.

How electroconvulsive therapy works is, of course, a question worth pondering. The answer should provide remarkable insights into the biology of mood disorder and the other conditions that are responsive to electroconvulsive therapy (ECT). This article will review several of the more prominent past hypotheses regarding the mechanism of action of ECT and the reasons why they are no longer tenable. Recent research findings will be summarized, regarding the response of the brain to electrically induced seizures that would need to be reconciled by any hypothesis seeking to explain the remarkable response of affectively disordered patients to this treatment modality. The paper concludes with a modem hypothesis based on an amalgamation of leading investigators' work in this field.


The therapeutic benefit of electroconvulsive therapy is based solely on its placebo effect. This is a dramatic treatment often administered in an operating room-like setting by staff who have strong expectations that it will benefit the patient. These are circumstances ideally suited to promote the maximum placebo response. However, several double-blind prospective studies13 employing a placebo control group of sham treatment (general anesthesia without electrical stimulation) have consistently demonstrated a significant therapeutic response above and beyond that of placebo. The only study of this design4 that demonstrated an equivalent response between active treatment and sham treatment employed what is now generally thought of as subtherapeutic electrical dosing for nondominant unilateral treatment.

It also appears that in a more severely ill group, the favorable response to treatment is more significant. Clearly something more than a placebo response is occurring. These same data can be used to refute a hypothesis that is closely allied to the placebo-effect hypothesis. That is, that ECT satisfies the depressive patients need for punishment as a means of expiation for consciously or unconsciously perceived wrongdoing.


ECT causes anatomical brain injury resulting in psychic numbing similar to that seen in patients who have undergone frontal lobotomy. This hypothesis, of course, has taken on the form of an accusation and is charged with emotional and political overtones. When treated as fact, it has been responsible for the banning of ECT in certain jurisdictions in this country.

The belief that ECT causes permanent brain injury is based on studies in animals subjected to repeated electrical stimulation and on autopsy studies performed on patients who died during or shortly after a course of ECT. The cat study of Hartelius,5 although carefully controlled, can be criticized for its use of exorbitant electrical dosing and failure to use anesthesia, which would blunt the cardiovascular response to repeated seizures. The applicability of work in the cat to humans, of course, is also problematic. Autopsy studies cannot control for either preexistent pathology or post-mortem changes. As detailed elsewhere in this issue, modern neuro-imaging studies have failed to reveal significant change in the human brain following a course of ECT.


ECT induces amnesia. The patients simply forget why they have become depressed. Retrograde and antegrade amnesia are well-documented side effects of ECT.e·7 They are more likely to occur in bilateral treatment than in unilateral nondominant treatment.8·9 However, there is no correlation between the amnestic effect of ECT and its therapeutic benefit. It is not uncommon today, with the use of briefpulse unilateral stimulation, that the patient will complete a successful course of ECT without experiencing anything other than mild amnesia surrounding the actual treatment. Yet the benefits of treatment will be sustained for several weeks to months after completion of the ECT course. The converse is also true that patients who fail to respond to ECT may exhibit memory dysfunction. Thus, there appears to be no association between amnesia and therapeutic benefit of ECT.


ECT causes the release of an asyet-unidentified peptide, antidepressiv It is said that a hypothesis that cannot be tested is not a hypothesis at all. but a guess. No reliable animal model exists for acute depression that would afford a before-and-after means of examining this issue. Furthermore, virtually every neurotransmitter system studied is perturbed by ECT. However, the significance of these changes is unclear and antidepressin is yet to be discovered. Designing a research study to test this hypothesis in humans has been problematic.


What do we know about the effects of stimulus intensity, lead location, and alterations in the neurophysiology of the brain on treatment outcome in ECT? In a series of elegant clinical studies performed over the last decade, the group at the New York State Psychiatric Institute, under the direction of Harold Sackeim,10"13 have made many basic observations about the brain's response to exogenously induced seizures. Among their more important findings are:

* Seizure threshold increases and seizure duration diminishes during a course of ECT.

* Right unilateral stimulation at or slightly above seizure threshold is significantly less beneficial than similarly administered bilateral treatment.

* Nonresponders tended to be patients whose seizure threshold did not increase and whose seizure duration did not diminish.

Sackeims finding of the relative ineffectiveness of right unilateral therapy employing electrical doses at or slightly above seizure threshold has recently been confirmed by Abrams et al.14 These findings suggest that contrary to long-held belief, electrophysiologic events occurring in the motor strip have very little to do with the mechanism of action of ECT but simply serve as an indicator that a seizure has been induced. The situation would appear to be infinitely more complex. Location and intensity of stimulus administ7-ation has a definite influence on the clinical response, suggesting that the path along which synchronous electrical activity is propagated, which neural structures are stimulated, and in what sequence are key to an understanding of how ECT works.

Post et al15 found that electroconvulsive stimulation has a similar effect to that of carbamazepine on the phenomenon of amygdala kindling. Tortella and LongHi demonstrated that a transferable antiseizure substance, possibly an endorphin, is produced in the experimental animal by a series of electroconvulsive stimuli. Anticonvulsant agents such as carbamazepine and valproic acid are effective in the treatment of affective disorder. These findings all suggest a confluence of the activities of a putative endogenous antiseizure system and the pathophysiology of affective disorders. A number of central nervous system elements are known to affect neuronal excitability and synchronization.


Effective anticonvulsant therapy results in heightened activity of the brain's endogenous antiseizure system and the neurotransmitters important to the function of this system. A by-product of this heightened activity is a correction of the abnormalities responsible for the symptoms of major depression.


Clinical studies conducted in the last 20 years have refuted all of the original hypotheses offered to explain the mechanism of action of ECT. The benefits of ECT appear to be closely linked to heightened activity of an endogenous antiseizure system in the brain. Collaboration between neuroscientists interested in the mechanism of action of ECT and those concerned with the neural biology of seizure disorder might prove to be useful in elucidating how ECT achieves its remarkable benefits in affectively disordered patients.

However, as was emphasized in a recent Scientific American article,17 "the human brain is the most complex object in our universe." It isn't clear how close we are to understanding this small but highly clinically relevant part of this complexity.


1. West ED. Electric convulsion therapy in depression: a double-blind controlled trial. Br Med J. 1981; 282(311:355-357.

2. Crow TJ, Johnstone EC. Controlled trials of electroconvulsive therapy. Ann N Y Acad Sci. 1986; 462:12-29.

3. Freeman CPL. Basson JV. Crighton A. Doubleblind controlled trial of electroconvulsive therapy (ECT) and simulated ECT in depressive illness. Lancet. 1978; 1(8067):738-740.

4. Lambourn J, Gill D. A controlled comparison of simulated and real electroconvulsive therapy. Br. J Psychiatry. 1978: 133:514-519.

5. Hartelius II. Cerebral changes following electrically induced convulsions. Ada Psychialr Scand Suppl. 1952; 77(suppl):92-119.

6. Ottosson JO. Memory disturbance after ECT- a major or minor side effect? In: Dunlop E, ed. Proceedings of the First International Congress of the Academy of Psychosomatic Medicine. New York: Excerpta Medica; 1967:161-168.

7. Squire LR. Slater PC. Chance PM. Retrograde amnesia: temporal grathent in very long-term memory following electroconvulsive therapy. Science. 1975: 187:77-79.

8. Halliday AM. Davison K. Browne MW, et al. A comparison of the effects on depression and memory of bilateral ECT and unilateral ECT to the dominant and non-dominant hemispheres. Br J Psychiatry. 1968; 114:997-1012.

9. Fleminger JJ, de Horno DJ, Nair PN. et al. Differential effect of unilateral and bilateral ECT. Am J Psychiatry. 1970: 127:430-136.

10. Sackeim HA. Decina P, Portnoy S, Neeley P, Malitz S. Studies of dosage, seizure threshold, and seizure duration in ECT. Biol Psychiatry. 1987; 22:249-268.

11. Sackeim HA, Decina P, Prohovnik I, et al. Anticonvulsant and antidepressant properties of electroconvulsive therapy: a proposed mechanism of action. Biol Psychiatry. 1983;18:1301-1310.

12. Sackeim HA, Decina P, Prohovnik I, et al. Dosage, seizure threshold, and the antidepressant efficacy of electroconvulsive therapy. Ann N Y Acad Sci. 1986; 462:398-410.

13. Malitz S. Sackeim HA, Decina P, Kanzler M, Kerr B. The efficacy of electroconvulsive therapy: dose-response interactions with modality. Ann N Y Acad Sci. 1986; 462:56-64.

14. Abrams R. Swartz CM, Vedak C. Antidepressant effects of high dose right unilateral electroconvulsive therapy. Arch Gen Psychiatry. 1991; 48:746-748.

15. Post RM,Putnam F, Contel NR, Goldman B, Electroconvulsive seizures inhibit amygdala kindling: implications for mechanisms of action in affective illness. Epilepsia. 1984; 25:234-239.

16. Tortella FC, Long JB. Endogenous anticonvulsant substance in rat cerebrospinal fluid after a generalized seizure. Science. 1985; 228:1106-1108.

17. Fischbach GD. Mind and brain. Sci Am. 1992; 267(3):48-57.


Sign up to receive

Journal E-contents