It has been known for many years that electric current applied to an animal’s head can elicit a convulsion. The procedure of eliciting seizures by electrical stimulation is termed “electroconvulsive shock” (ECS). The term “electroshock therapy” (EST) is used in medical references. In the early 1930s drug-induced convulsions (using insulin or Metrazol, for example) were adopted as a treatment for patients with severe mental disorders. In 1938 Cerletti and Bini demonstrated that ECS provided a more highly controlled and reliable means of eliciting convulsions. ECS gradually replaced drugs in convulsive therapy treatments and is still used to some extent as a treatment for mental disorders. It is considered to be particularly effective in the treatment of severely depressed patients (Ulett et al. 1962). Since Cerletti and Bini introduced ECS there have been numerous extensive clinical and experimental studies of the nature and bases of its behavioral effects.
The ECS convulsion . Rats or other rodents have been used as subjects in the majority of experimental studies of ECS. In general, the pattern of the convulsion obtained with rats is similar to that obtained with larger mammals, including humans. Convulsions are usually elicited by stimulating the rats with 25 to 100 milliamperes of current (AC) for approximately 0.2 to 1 second. The current is delivered through electrodes that are either attached to the animal’s ear or applied directly to the rat’s corneas. The maximal seizure (grand mal) consists of a highly stereotyped sequence of movements. The animal’s hind legs are first drawn up and then extended. This tonic extension lasts for several seconds and is followed by a brief phase of whole-body clonus. The animal then remains in an immobile state for several minutes. A maximal convulsion can again be elicited in approximately ten minutes. For about 15 minutes following the convulsion the animals are usually hyperirritable. Current below that necessary to elicit a maximal seizure may produce violent running, temporary immobilization, or clonic convulsions. Maximal seizures can be prevented by administering to the animals such depressant drugs as ether, phenobarbital, or diphenylhydantoin prior to the ECS stimulation (Toman et al. 1946). Maximal seizure thresholds increase with the age of the animals and, in different laboratory species, vary directly with the weight of the animals.
In the clinical use of ECS, patients are usually given a series of treatments. Most experimental studies with laboratory animals have adopted this procedure. This procedure is, of course, appropriate for research viewed as an experimental analogue of the clinical treatments. However, recent evidence indicates that ECS has numerous behavioral effects and that the effects of a single ECS treatment are quite different from those produced by a series of treatments.
Activity, sexual and maternal behavior . When rats are given a series of ECS treatments (one or more per day for 7 to 25 days) their behavior is markedly affected. Spontaneous activity is depressed, male sexual arousal is impaired, and maternal behavior (for example, nest building and care of the young) is disrupted (Munn 1950; Beach et al. 1955). The animals’ behavior usually returns to normal within a few weeks after the treatments are discontinued.
Learning and retention . Multiple ECS treatments have also been found to impair rats’ maze learning and retention. The degree of impairment is directly related to the complexity of the maze task. However, the impairment appears to be temporary. Little impairment is found if a month elapses between the last ECS treatment and the maze learning or retention tests. There is no impairment of performance if convulsions are prevented by the delivering of the ECS stimulation while the animals are anesthetized with ether (Munn 1950; Russell 1948).
Studies of memory in human patients given a series of ECS treatments yield findings similar to those of the rat studies. Patients typically experience impaired memory for several weeks following the termination of the treatments. This deficiency is fairly general and typically involves difficulty remembering well-learned life-history data as well as recently experienced events. Although the deficits generally disappear within a few weeks, deficits have been observed over a period of months in some patients. Some investigators have suggested that temporary memory impairment may contribute in some way to the therapeutic effectiveness of ECS treatments (Janis & Astrachan 1951).
Conditioned emotional response . A number of studies have shown that a series of ECS treatments is particularly effective in attenuating a learned emotional response (Hunt 1965). In these studies rats were first trained to press a lever for a water reward. They then were presented with a series of trials in which a clicking noise was followed by a painful shock delivered to their feet. Within a few trials the clicking noise elicited a conditioned emotional response (CER) consisting of crouching, urination, defecation, and depressed rate of lever pressing. The rats were then given a series of 21 ECS treatments—three per day for seven days. On subsequent tests the clicking noise failed to elicit the CER. Less of an attenuating effect was found if fewer ECS treatments were given or if several weeks elapsed between the CER training and the ECS treatments. The differential effect of ECS on the lever-pressing response and the CER is not due to the fact that the CER was the last response learned. Similar results are obtained when the CER training is given prior to the learning of the lever-pressing response. The effects of multiple ECS on the CER, like those found with maze learning and retention as well as sexual and maternal behavior, are transient. The CER typically reappears within a month following the treatments. The convulsions appear to play a critical role in attenuating the CER. Multiple ECS treatments do not attenuate the CER if the convulsions are prevented by etherizing the rats prior to each ECS treatment.
It is clear from these studies that a series of ECS treatments markedly affects rats’ behavior. The findings indicate, however, that the effects are for the most part temporary. Effects on sexual arousal and learning and retention generally last for only a few weeks following the treatments. Further, it seems clear that the convulsions are essential for the effects observed. There is little evidence of behavioral effects of multiple ECS treatments when convulsions are prevented by anesthetization of the animals prior to the treatments.
Shortly after the introduction of ECS clinical observations indicated memory loss as a common consequence of such treatments. In addition to a general loss of ability to remember names, events, and personal life history, patients seemed to have amnesia for events that had occurred shortly before each treatment. This phenomenon has been termed “retrograde amnesia.” Systematic studies of memory in patients treated with ECS have confirmed this clinical observation. In an early study, for example, patients were shown a series of pictures prior to receiving an ECS treatment. In tests given the next day they showed poorest retention of the last pictures seen before convulsion (Mayer-Gross 1943).
Evidence that ECS produces retrograde amnesia has been obtained in a large number of experiments with rats. In the first two of such studies (Duncan 1949; Gerard 1955) rats and hamsters, respectively, were given a single ECS treatment after each trial in a learning task. The animals were arranged in different groups and each group was given the ECS treatment at a different time interval after each trial. With the intervals shorter than one hour, the rate of learning increased directly with increases in the length of the interval between the trial and the ECS treatment. Learning was not affected by the treatments given one hour or longer after each trial. Inasmuch as all of the experimental animals were given a series of ECS treatments, it is not possible to discount the possibility that at least some of the effects observed in these studies were due to the cumulative effects of repeated ECS treatments. It is important to note, however, that treatments did not affect behavior unless they were administered shortly after each training trial. Thus, these findings contrast with findings of a general learning and retention impairment in rats given a series of ECS treatments prior to learning or retention tests.
Evidence of the retrograde amnesic effect of ECS has also been obtained in a large number of studies in which the animals were given only a single ECS treatment. In a series of experiments Thompson and his associates (for example, Thompson & Dean 1955) administered an ECS to different groups of rats at different intervals following massed training trials on a visual discrimination problem. The findings were similar to those obtained by Duncan and Gerard. On retraining trials, 48 hours later, animals given an ECS treatment four hours after the training did not react differently from the controls. In animals treated within one hour after the training, efficiency of relearning varied directly with the interval between the training and the ECS treatments.
In other research Thompson and other investigators have shown that the degree of the retrograde amnesia found with ECS depends upon numerous conditions, including the age and strain of the subjects, degree or strength of original learning, and complexity or difficulty of the learning task. The greatest effects are found when the rats are young or brain damaged and when the learning task is difficult. The duration of the retrograde amnesic effect of ECS is typically limited to a few minutes when a relatively simple learning task is used (Glickman 1961).
During recent years most of the interest in ECS has centered on the problem of the basis of the retrograde amnesic effect. Various hypotheses have been proposed.
Brain damage . A number of investigators have suggested that the amnesic effect of the ECS may be due to brain damage produced by the current. Studies of the brains of experimental animals subjected to ECS treatments indicate that there is some evidence that ECS produces some changes in the brains—particularly small hemorrhages. However, the changes are generally minor and reversible (Madow 1956). It could be that the confusion and general impairment of learning and retention found with a series of ECS treatments is due in part to reversible vascular damage. Such effects could not, however, account for the retrograde amnesic effects of ECS. A single ECS treatment has little or no amnesic effect if the treatment is given several hours after training has been terminated. Further, a single ECS does not impair rats’ subsequent ability to learn a new task or perform previously well-learned tasks. Thus, although ECS may produce brain damage, it would be difficult to explain the differential effects of ECS on recent and older memories in terms of brain damage.
Interference with memory consolidation . The most generally accepted interpretation of the retrograde amnesic effect of ECS is that the ECS interferes with the neurophysiological processes involved in storage or consolidation of memory traces (Glickman 1961). Evidence from a variety of clinical and experimental studies has provided strong support for the hypothesis that memory trace consolidation is based upon the perseveration of neurophysiological processes initiated by an experience. In humans retrograde amnesia is a common consequence of head injuries. In experimental studies with laboratory animals retrograde amnesia comparable to that found with ECS treatments has been produced by such treatments as hypoxia, drugs, and audiogenic seizures. The treatments seem to prevent the storage of information acquired during the training trials just prior to the treatment. According to this interpretation the memory loss should be permanent. Available evidence is consistent with this hypothesis. In a study by Chevalier (1965) rats were trained for a task and then given a single ECS. Clear evidence of retrograde amnesia as long as 60 days later was obtained.
Some investigators have assumed that the time required for consolidation to occur after training is indicated by the minimum interval between training and ECS treatment within which no memory impairment is found. However, since different kinds of post-training treatments produce varying degrees of retrograde amnesia, it is more likely that memory storage involves a sequence of processes and that the different treatments are capable of interfering with different processes. For example, with ECS treatments retrograde amnesia is usually obtained only when a few minutes or at most an hour elapses between training trials and treatments. With drugs (for example, Metrazol and puromycin) retrograde amnesia has been found with training-treatment intervals as long as several days (McGaugh &; Petrinovich 1965). As indicated above, the deficits in learning and memory observed following a series of ECS treatments tend to disappear within a few weeks.
Most of the experimental studies of ECS effects on memory have used learning tasks employing aversive motivation—usually punishing shock. Consequently, it has been suggested that the ECS treatments may produce only a selective amnesia for the aversive stimulation. This intriguing hypothesis, however, is not supported by the data. Clear evidence of retrograde amnesia has been obtained in studies using food and water rewards. Further, in studies using shock motivation in discrimination learning tasks, the ECS seems to have its primary effects on the memory of the correct cue rather than on the motivation for responding.
Punishing effects of ECS . The findings of several studies suggest that impairment of performance found following repeated ECS treatments may be due in part to aversive effects of the treatments. For example, Friedman (1953) found that rats’ performance of a lever-pressing response for food reward was depressed after the animals were given a series of ECS treatments in the apparatus. Behavior of the animals suggested that they had learned to fear the apparatus. The animals urinated, defecated, trembled, approached the lever hesitantly, and then ran away from it. The depression of responses and emotional behavior was considerably less marked in animals given the ECS treatments in a dissimilar apparatus. No effects were observed, however, in subjects treated while under ether anesthesia. The behavioral effects appeared to be due to the convulsions rather than to the ECS current.
Other studies have shown that rats tend to stop performing responses that are repeatedly followed by ECS treatments. These findings have lead some investigators to suggest that the learned fear rather than amnesia may be the cause of the impaired performance of subjects given ECS shortly after training. Recent evidence indicates that, paradoxically, ECS treatments have both amnesic and aversive effects; the amnesic effects, however, cannot be explained in terms of the aversive effects. In one study rats were given a single foot shock as they stepped from a small platform to the floor of a table; half of the rats were given an ECS within a few seconds. The next day the rats were placed on a platform again; those that had received the foot shock but no ECS tended to remain on the platform; those that had been given the ECS treatment displayed little evidence of remembering the previous foot shock and most of them readily stepped off the platform. In a subsequent study using a similar procedure rats were given an ECS in the apparatus each day for eight days immediately after they had stepped from the platform. Latencies of stepping from the platform increased gradually over the eight days. Other rats were given only a foot shock, and most of these animals learned to stay on the platform within two trials. The performance of rats given foot shocks followed by ECS was similar to that of rats given only ECS treatments. Neither amnesic nor punitive effects were found when the ECS treatments were given one hour after each trial (McGaugh 1965). The results of these studies indicate that the aversive effects of ECS are found with repeated treatments, whereas retrograde amnesia can be obtained with a single ECS treatment. Thus, the amnesic effects of a single ECS cannot be explained in terms of the punishing effects of the treatment.
The paradox remains, however. What is the basis of the punishment if the treatment produces amnesia for events just prior to the ECS stimulation? One possibility is that the subjects learn gradually to associate the apparatus cues with the aftereffects of the convulsion. Patients given a series of ECS treatments tend to develop a fear of the treatments (Gallinek 1956). The patients readily admit that they do not experience any discomfort during the treatment. The fear seems to be based on the severe disorientation experienced while recovering from each treatment. The finding that the aversive effects in rats are eliminated by administering the ECS while the animals are anesthetized lends support to the interpretation that the aversion is based on the aftereffects of the convulsions. A more complete understanding of the basis of the aversive effects of ECS treatments requires additional research [see Learning, articles onreinforcement and avoidancelearning].
Competing response hypothesis . Lewis and Maher (1965) have proposed still another interpretation of the amnesic effects of ECS. These investigators suggested that through conditioning, the cues in the apparatus elicit behavioral inhibition, that is, a general muscular relaxation and lowered level of activity that competes with and thus interferes with the performance of the previously learned response. Most of the findings of ECS studies, however, are clearly inconsistent with this hypothesis. First, as indicated above, amnesia is found with a single treatment while other behavioral effects, including changes in response latency and freezing and crouching, appear only after a series of ECS treatments. Second, retrograde amnesia is obtained with a single treatment even when the treatment is administered outside of the training apparatus (this, in fact, is the typical procedure). Third, rats given a series of ECS treatments in an apparatus do not appear to be relaxed. They typically urinate, defecate, and tremble. Further, rats will actively avoid a place in an apparatus where they are given repeated ECS treatments (McGaugh 1965). These findings are clearly inconsistent with the conditioned inhibition interpretation of ECS effects.
Contribution of the convulsions . There is clear evidence that the effects of repeated ECS treatments are different from those produced by a single ECS. The effects of repeated ECS treatments seem to be due to the convulsions rather than to the current. Impaired maternal behavior, learning and retention deficits, suppression of a CER, and punishing effects are not found if the convulsions are prevented by the administration of the ECS while the subjects are anesthetized. Recent work indicates, however, that the convulsions are not essential for the production of retrograde amnesia with a single ECS treatment. In unpublished research McGaugh and Alpern obtained clear evidence of retrograde amnesia in animals anesthetized with ether just prior to receiving a single ECS treatment. These findings suggest that the retrograde amnesic effect of ECS is produced by the current rather than the convulsion and are consistent with other recent evidence that retrograde amnesia can be produced by restricted subcortical (and subconvul-sive) electrical stimulation of the brain (Williston et al. 1964).
Although many of the effects of ECS remain to be explained, it is clear that ECS has a variety of behavioral effects and that the varied effects are not readily explained in terms of any single hypothesis. In particular, the retrograde amnesic effects of ECS must be considered separately from the diverse and complex effects found when ECS is repeatedly administered. The results of experimental studies of ECS have, as yet, shed little light on the basis of the therapeutic effectiveness of ECS in human patients. The findings have, however, had considerable influence on theories and research concerned with learning and memory. It may be that the therapeutic effectiveness is due to the learning and memory effects; an evaluation of this hypothesis must await further research.
James L. McGaugh
Beach, Frank A.; Goldstein, A. C.; and Jacoby, G. A. JR. 1955 Effects of Electroconvulsive Shock on Sexual Behavior in Male Rats. Journal of Comparative and Physiological Psychology 48:173–179.
Cerletti, U.; and Bini, L. 1938 L’elettroshock. Archivio generale di neurologia, psichiatria, e psicoanalisi 19: 266–268.
Chevalier, Jacques 1965 Permanence of Amnesia After a Single Posttrial Electroconvulsive Seizure. Journal of Comparative and Physiological Psychology 59:125-127.
Duncan, Carl P. 1949 The Retroactive Effect of Electro-shock on Learning. Journal of Comparative and Physiological Psychology 42:32–44.
Friedman, Merton H. 1953 Electroconvulsive Shock as a Traumatic (Fear Producing) Experience in the Albino Rat. Journal of Abnormal and Social Psychology 48:555–562.
Gallinek, Alfred 1956 Fear and Anxiety in the Course of Electroshock Therapy. American Journal of Psychiatry 113:428–434.
Gerard, R. W. 1955 Biological Roots of Psychiatry. Science New Series 122:225–230.
Glickman, Stephen E. 1961 Perseverative Neural Processes and the Consolidation of the Memory Trace. Psychological Bulletin 58:218–233.
Hunt, Howard F. 1965 Electroconvulsive Shock and Learning. New York Academy of Sciences, Transactions Series 2 27:923–945.
Janis, Irving L.; and Astrachan, Myrtle 1951 The Effects of Electroconvulsive Treatments on Memory Efficiency. Journal of Abnormal and Social Psychology 46:501–511.
Lewis, Donald J.; and Maher, Brendan A. 1965 Neural Consolidation and Electroconvulsive Shock. Psychological Review 72:225–239.
McGaugh, James L. 1965 Facilitation and Impairment of Memory Storage Processes. Pages 240–291 in Daniel P. Kimble (editor), The Anatomy of Memory. Palo Alto, Calif.: Science and Behavior Press.
McGaugh, James L.; and Petrinovich, Lewis F. 1965 Effects of Drugs on Learning and Memory. International Review of Neurobiology 8:139–196.
Madow, Leo 1956 Brain Changes in Electroshock Therapy. American Journal of Psychiatry 113:337—347.
Mayer-Gross, W. 1943 Retrograde Amnesia. Lancet , no. 2:603–605.
Munn, Norman L. 1950 Handbook of Psychological Research on the Rat. Boston: Houghton Miffin.
Russell, Roger W. 1948 Contributions of Research on Infrahuman Animals to the Understanding of Electric Convulsive Shock Phenomena. Journal of Personality 17:16–28.
Thompson, Robert; and Dean, Waid 1955 A Further Study of the Retroactive Effect of ECS. Journal of Comparative and Physiological Psychology 48:488–491.
Toman, James E. P.; Swinyard, E. A.; and Goodman, L. S. 1946 Properties of Maximal Seizures, and Their Alteration by Anticonvulsant Drugs and Other Agents. Journal of Neurophysiology 9:231–239.
Ulett, George A.; Smith, K.; and Biddy, R. 1962 Shock Treatment. Progress in Neurology and Psychiatry 17: 559–571.
Williston, John S. et al. 1964 Disruption of Short-term Memory by Caudate Stimulation. American Psychologist 19:502 only.