The effects of abstinence and of re-training on the chewer of coca-leaf




Pages: 41 to 47
Creation Date: 1969/01/01

The effects of abstinence and of re-training on the chewer of coca-leaf *


In a previous paper [ 4] it was demonstrated that long-term chewers of coca-leaf show relative signs of chronic brain damage when compared with controls of similar background who have taken little or no coca-leaf. The psychological deficit was found only through tests specifically designed to reveal signs of brain damage and was not apparent in the subjects' daily life and work, so that it must be viewed as of minor importance in comparison with the much greater handicaps which socio-economic conditions impose on both users and non-users of coca in the Andean region. However, when the economic situation is improved - which one hopes will be soon - then the matter could become more important. In particular, these earlier findings raise the question whether the long-term use of coca would prevent the habitués from taking advantage of new opportunities, particularly if these opportunities involved learning new skills.

For that reason it was decided to return to the site of the earlier study and to investigate the chronic cocauser's response to a learning situation, again using matched controls from the same background but adding the further variable of having half of the users continue on their customary daily amount of coca while the other half were in enforced abstinence. This latter modification was intended to place beyond doubt the inference derived from internal evidence in the previous study, that the deficit was due to the chronic and not merely to the recent use of the drug; and through the use of some of the original tests in the learning situation it was hoped that the earlier differences between the habitués and their controls would be confirmed and expanded. Furthermore, by observing the abstinence group closely, it was hoped that conclusions could be drawn regarding withdrawal effects and the subjects' ability to do without the drug. From earlier researchers [ 1] and from the social scientists working with our own research [ 2] the impression had been gathered that coca could quite easily be dispensed with when economic conditions made this worthwhile, but the matter called for stricter investigation.

This work was supported by a grant from WHO.


The investigation was carried out in the same sugargrowing area of Northern Argentina as had been the site of the previous study, using this time the empty wing of a plantation hospital. Subjects and controls were field laborers matched by age, literacy and birthplace as on the previous occasion, but, in order to reduce the number of variables to be allowed for, recent immigrants from Bolivia were excluded and only persons with more than twenty years' residence in the area were included. Some of the subjects had participated in the previous investigation two years earlier, thus enabling the consistency of the test results to be assessed 1 and all were initially screened to ensure that they fell into the appropriate category for the experiment and that extraneous physical or mental defects were absent. Chronic chewers were required to have used the leaf regularly for at least ten years, with an average daily dose of not less than about 30 g; most of them had used more, and for longer. Controls were required to have taken the leaf only occasionally, as on social occasions, and never to have been observed to chew it while at work. For both groups the information given was checked with plantation supervisors and storekeepers, who knew very well which men used the leaf during their work and how much each family consumed in a week, and this resulted in the exclusion from the experiment of three men who had originally been selected for it. As explained in the earlier paper, there was no pressure in this locality either towards using the leaf or towards abstaining from it, at least as far as the field laborers were concerned, and we found no evidence of deliberate falsification of answers regarding the habit.

Table 1 shows the composition of the groups.

Performances on this second experiment were very similar to those exhibited two years previously, with the same peculiarities reappearing in the same individuals, thus suggesting that the tests are consistent. Average performance was slightly better than in the earlier period, but too slight to make any difference to the present trials.


Characteristics of studied subjects


No. of subjects

No. tested two years previously

Mean age

Mean daily consumption of coca-leaf (reported)

Duration of habit (years)

50 g.
6 2 42
45 g.
4 2 36
45 g.
6 3 42
40 g.
5 1 40
5 3 40

Subjects and controls were admitted to the hospital in two groups, each containing approximately equal numbers of continuing-users, abstainers and non-users. The total stay in hospital of each group was ten days. All subjects were given initial medical examinations, and some physical tests (urinalysis, etc.) were continued each day throughout their stay to reinforce the hospital atmosphere. All were paid their normal wages during their time in hospital. The continuing-users were supplied with coca each day in whatever amount they desired, but none of them took more than what they had previously reported to be their customary ration, and it was necessary to instruct them that they must consume it all, thus supporting the impression that this is not a drug which induces dependence and increasing consumption. Abstainers and controls were told that they must not take any coca during their stay and that the urine tests would show whether they had done so or not. Irregular visits were made to the wing throughout the twenty-four hours to ensure that there was no breach of this rule but none was found, the abstainers being apparently able to tolerate lack of the leaf quite easily, with some of them saying that they would not have been chewing it in such a situation even if they had been allowed to, since they only took it to permit hard work to be done and to suppress hunger. On the other hand there was some discontent at the fact that alcohol was not permitted.

The learning situation to which they were exposed was the repeating of a single battery of psychological tests thrice during the ten-day period, and exercising daily with a type of skill game involving manual dexterity and pattern recognition (so that experience with it was quite relevant to the test battery) which the subjects found amusing. The test battery was built up of those items in the earlier study which had yielded the greatest differences between coca-users and controls, namely an auditory memory test, two pairs of tests drawn from the Army Beta Examination, the Knox Cubes used as an attention test, the Seguin Formboard, and three of the Wechsler Block Designs. Indications of how these were modified for local use are given in the previous paper [ 4] . Small incentive payments were offered in addition to the subjects' regular wages in order to induce them to try to beat their former scores. Originally it had been intended that all subjects should work daily at a simple task which would be new to all of them and in which improvements in speed and accuracy would be easily measured, but supply of the necessary materials failed at the last minute and this had to be abandoned.

Because coca is used to suppress hunger and because cocaine interferes with sleep in naive subjects, the amount of food eaten by each subject and the number of hours he slept were recorded daily. In addition, since it was unclear from our earlier investigations whether the leaf was used to facilitate social intercourse or not, the men were organised into small mixed groups each evening and one of us (Rios) regularly noted how much each man spoke and whether he showed irritability, somnolence, etc.


The assumption on which the present experiment was based was that all subjects would improve their scores on the test battery as they acquired additional experience with it and with the training game, but that the different groups would not improve by the same amount. The controls were expected to perform better than both groups of coca-users initially, but if this difference were due to the dulling effect of recently consumed coca on the latters' attention, then the abstainers could be expected to catch up with the controls by the end of the experiment, assuming that there was no withdrawal effect to confuse the issue, whereas the continuing-users should not. However, if the coca had induced a chronic deficit, as was suspected, and if this affected learning, then the degree of improvement in performance should be less in the users than in the controls, with the abstainers possibly doing less well than the continuing-users, since there had been some evidence that the immediate effect of coca might compensate, through inducing greater alertness, for some of the chronic damage. Finally, if the drug induces marked withdrawal effects, then the abstainers could be expected to show a decline in scores - especially in tests of attention - that would be absent with the continuing-users.

Regarding social interaction, we expected that even though the subjects' work supervisors had been unable to see any difference between coca-users and nonusers, in the more closely and professionally observed conditions of our experiment a difference might appear. We hypothesised that the coca-users, especially the continuing-users, would be less socially involved and more apathetic than the controls, whereas it seemed likely that the abstainers would show some irritation, with shortness of temper and increased movement as is common when many habitually taken drugs, including nicotine, are withdrawn.


The expectations that the controls would initially score better than either group of coca-users and that all three groups of subjects would improve their performances were confirmed, thus supporting both the original findings regarding a psychological deficit and the design of the present experiment. From the combined data charted on Figure I it can be seen that there is only one point at which these two expectations are not fulfilled, namely on the first performance of the abstainers on the Similarities tests from the Army Beta. On that occasion some of them scored unusually high on this item, but were unable to keep it up on subsequent trials and had dropped to their expected position by the end of the series.

Figure I. - Relative scores of coca-users and controls on first and third trials with six psychological tests, data being reduced to comparable scale and corrected for differences in literacy between the groups. (Vis. = Knox Cubes; Aud. = Auditory Memory Test; Com. = Army Beta Figure Completions; Form. = Seguin Formboard; Sim. = Army Beta Similarities; Wec. = Wechsler Blocks.

Full size image: 12 kB, Figure I

(_____ Controls ;- - - Continuing-users; ................ Abstainers)

The hypotheses that the users might catch up with the controls after their training, or alternatively that there would be a consistent difference between the continuing-users and the abstainers, on the other hand, are not supported. The coca-chewers are clearly able to learn, but when all tests are taken together their performances show no consistent signs of overtaking those of the controls and one might argue that they do not even learn as well, since it would seem much easier to improve from the very low levels at which some of the coca-users start than to improve from the more normal levels of the controls. The point is doubtful, however. Regarding abstention, also, there is no evidence in Figure I to suggest either that this improves performance or that it detracts from it, since there are two types of test on which the abstainers perform better than the continuing-users, three on which they perform worse, and one on which the two groups are about the same. Taking all the tests together, therefore, it would appear that long-term coca-users can learn through training but that they cannot catch up with corresponding non-users in performance, and that their recent consumption of coca is not a factor in the results.

Figure II. - Relative scores of coca-users and controls on the first and third trials with the Seguin Formboard. (A = using preferred hand only; B = using other hand only; C = using both hands.)

Full size image: 7 kB, Figure II

(______Controls; - - - Continuing-users; ................ Abstainers)

When these results are looked at in more detail, however, some modification of conclusions is suggested. In the first place there is one group of subtests on which the coca-users do appear to be able to catch up with the controls, while there is another on which they fall increasingly behind. In the second place abstention seems to benefit a particular category of task, but to interfere with another. The subtests on which the coca-users seem able to catch up with the controls are those in which manual dexterity plays a considerable part, in which both hands are permitted to be employed, and in which there is no memorising and no serious abstract thinking required. This picture is found with the first and, more doubtfully, the second of the Wechsler Block Designs, and with the bimanual run on the Seguin Formboard; and on these subtests it is also the case that the continuing-users of coca perform substantially better than the abstainers. Figure II shows the results with the Seguin Formboard, a test in which cut-out shapes must be matched with designs on a board and in which speed is the major element in scoring. There are three trials, first with one hand, then with the other, and then with both, the same shapes and board being used each time and being visible throughout the test. Accordingly, there is no memorising to be done, and nimbleness at handling the pieces can compensate for slowness at recognising a rotated form. As Figure II indicates, both groups of coca-users are slower than the controls when the first, preferred, hand is used, but are less behind on the second, and when both hands are employed the continuing-users have completely caught up. In the tests two years previously the same finding did not appear, but there were differences between the earlier and the later experiments' settings which could be important here, and since the catching-up this time proves true regardless of what age group of subjects is considered, or whether they are literate or illiterate, there are grounds for thinking that it is more than an accident.

Figure IIL - Relative scores of coca-users and controls on the first (A), second (B) and third (C) designs with the Wechsler Blocks, three trials being shown for each design. Scale indicates quarter-minutes required to complete design, with incorrect completions being counted as five minutes. Accordingly, the best performance has the lowest score.

Full size image: 9 kB, Figure IIL - Relative scores of coca-users and controls on the first (A), second (B) and third (C) designs with the Wechsler Blocks, three trials being shown for each design

( - - Controls ;--- Continuing-users; ..........Abstainers)

On the other test in which manual dexterity and speed play substantial roles, the picture starts out by being apparently similar, but then sharply diverges. As Figure III shows, the coca-users get significantly lower scores on the first design in the Wechsler Blocks test than the controls do, but by the second trial the continuing-users have caught up, and by the third trial, at the end of their hospital stay, the abstainers have caught up as well. On the second subtest the pattern is confused through the controls failing to improve their scores as expected, but it seems likely that the continuing-users would have caught up with them in any event, given time, although the abstainers have fallen behind. On the third subtest, however, the controls improve their performance greatly, while the two groups of coca-users are much poorer. Since, once again, the general relationships which Figure III shows repeat themselves regardless of whether one takes only the older men or only the younger, only the literate or only the illiterate, the differences seem likely to be genuine. Although both the Wechsler Blocks and the Seguin Formboard tests focus on pattern recognition, manual dexterity and speed, there are at least two important differences in the ways they present themselves, and these could produce a difference in results.

These results with the manual tests remained much the same regardless of whether the subjects were literate or illiterate, but the results with the two tests exploring memory and attention (Fig. IV) show that here literacy is important, and in an interesting fashion. Among me control subjects the literate score very much better on the Knox Cubes and the Auditory Memory tests than do the illiterate, both on immediate and on delayed recall, and repetition of the tests at three-day intervals does not change this. Among the coca-users, on the other hand, the literate are no better than the illiterate, regardless of whether they are abstainers or continuing-users. The illiterate controls perform no better, on average, than the illiterate coca-users; the literate controls perform much better than the literate coca-users. Accordingly, these tests would appear to tap two types of memory, one of which is developed only by formal training in school and is attacked by long-term coca consumption while the other is relatively untrained and is not affected by the coca. Age does not appear as a relevant variable in these results, although the previous experiments, which used more subjects and a wider age range, suggest that it could be relevant. Similarly, judging from the earlier findings, conditions of upbringing might be relevant even though not explored in the present study.

Figure IV. -- Relative scores of coca-users and controls on the Immediate and Delayed subtests of the Knox Cubes and Auditory Memory tests (combined), literate and illiterate subjects being distinguished.

Full size image: 10 kB, Figure IV

( - - Controls ;--- Continuing-users; ..........Abstainers)

On the manual tests, as was mentioned, the continuingusers of coca almost always performed better than the abstainers and there is reason for thinking that this difference is significant. On the tests involving memory such a difference is absent, but on one of the pairs of subtests from the Army Beta examination the positions of the two groups reverse, and since the superiority of the abstainers can be found equally in the literate and the illiterate, in the older and the younger, it is possible that this is more than an accident. The subtests in question call for the recognition of errors in the repeating of a symbol, design or group of numbers, and the subjects can go as slowly as they like. The impression which we have from the data is that the continuing-users went through these subtests more rapidly than the abstainers, but with less care.


Our expectations that coca would affect social interaction, sleep patterns, or appetite were not borne out, and none of our hypotheses here were confirmed. On physical examination there was nothing to distinguish the three groups, a finding which merely repeated what had been noted in our earlier report. Regarding sleeping, there was no evidence that the long-term coca-users slept either more or less than the controls, and they appeared neither more nor less rested in the morning. An unaccustomed chewer of coca, according to the hospital staff, would tend to suffer from insomnia after his indulgence, but the experimental subjects would chew a wad of leaves till perhaps 10 p.m. and be asleep within a half hour of ceasing. Regarding appetite, we had expected that since coca was used by the local population to suppress hunger when travelling or working in the fields, the continuing-users might eat less than the other experimental subjects. However, they ate amply, and explained their behaviour by saying that coca did not kill hunger but merely made it more easy to bear and to forget.

Social interaction had been formally scored by means of simplified sociograms of the evening group meetings, and less formally by noting each subject's daily behaviour on the ward and in the training and testing situations. During the evening meetings, when members of all three groups would be together, there was no evidence that one had more vocal members than another, or initiated contact more often, or received more attention from the others, although there were the expected variations between individuals. The abstainers were not more short-tempered or agitated than the rest, and neither were they more somnolent. There were, naturally, some subjects who were less cooperative than the rest and two who made a little trouble because they were not allowed alcohol, but the use or non-use of coca did not differentiate them. Most of the men's time was spent playing cards, and although no formal note was taken of who won and lost the impression was that the coca-users did not lose more often than the controls. Matters were made a little abnormal by the fact that the coca-chewers were forbidden to offer the leaf to the others, something they would normally do in a social setting, but the resultant differences were probably fewer than if the same conditions had applied to cigarettes.


Of the two sets of findings, intellectual and social, presented here, the latter are not necessarily less important than the former. When one asks whether the coca habit may interfere with its followers' ability to grasp improved socio-economic possibilities, it is quite as important to consider its effects on social interaction as its effects on learning. Accordingly, the fact that the coca-users showed themselves to be quite as cooperative and as easily interacting as the controls is of some importance, and of even more importance is the observation that temporary deprivation of their coca-leaf created no disturbance. Previous researchers, including the sociologists in our own team, have reported similar impressions, but always these impressions were derived at second hand (e.g. through the subjects' own statements and from supervisors), or they related to unrepresentative samples (e.g. young army recruits or men promoted to technical work). The present findings relate to representative field workers in middle age and are based on direct and formal observation in a closed setting. The only factor which makes the conclusion at all doubtful is the fact that our subjects were observed under what were for them unusually comfortable circumstances, since they were being well fed and paid standard wages to sit around most of the day playing cards. Even if these conditions had made them more sociable than usual, however, one would have expected there to be some detectable difference between the users and non-users, if coca had had any significant effect on social interaction, and we found none. Moreover, in the evening discussions, which were sometimes led around to the subjects of coca and of alcohol, it was confirmed, as our sociologists had deduced, that there are in this region no religious or cultural associations with coca which would lead to the habit being retained after the manifest reason for adopting it, i.e. the fact that it makes fatigue, pain and hunger more easy to bear, had been rendered less important. From this angle, therefore, there seems little reason to fear coca or to prohibit its use prior to an improvement in social conditions.

From the angle of intellectual performance, however, reasons for fearing the long-term effects of the coca habit remain. Our results confirm that coca-users fall behind control subjects in some, though not all, of our tests, and that they are probably suffering from some form of brain damage. For public health and social planning in the Andes region it is of some importance to assess how great this damage is, and which mental functions it would interfere with in real life.

The battery of tests originally used by us was intended to reveal brain damage regardless of its nature or location, and covered a wide range of functions all of which had been found to be affected in one or other form of organic mental disorder. In our earlier experiment the coca-users showed a deficit on nearly all items, thus pointing to a general organic brain syndrome but not to any more localised deficit. In the present experiment the data are more discriminatory and permit a number of deductions to be made. In the first place, there are three aspects of mental functioning on which the cocausers seem relatively unimpaired, namely simple untrained memory, hand-mind coordination, and elementary concept formation. Regarding memory, we find that the coca-users perform no worse than the controls on tests of immediate and delayed recall, both visual and auditory, when both groups had had no schooling. This finding is in accord with the clinical observation that the coca-users were no less accurate than the controls in knowing the day of the week, the details of their family history, etc., and it thus distinguishes chronic coca poisoning from the results of cerebro-arteriosclerotic disease and alcoholic dementia. Regarding brain-muscle coordination the same applies, for whereas chronic alcohol poisoning and many other forms of brain damage result in impairment here, chronic coca poisoning probably does not. The coca-users initially perform worse than the controls in manual tasks, but improve rapidly, and although the previous study indicated a correlation between duration of coca-chewing and decline in the Seguin Formboard test results, it seems likely that this relates more to the question of understanding or conceptualising the task than to its execution. The subjects displayed no ataxia, facial incoordination or signs of poor muscle control clinically. Finally, the tests did not indicate any special difficulties in verbalisation, or constrictions of the visual fields, or disturbances of perception such as would point to a particular locality of the brain being affected.

Where the coca-users did fall behind the controls was at three points- in the more efficient memorising associated with literacy, in the third subtest with the Wechsler Blocks, and in the more difficult sections of the Army Beta subtests. Considering the Army Beta items, the coca-users prove to perform adequately with the figures to which they can easily give a name, as in recognising that a hammer is not an ax, but to perform poorly when the whole cannot so easily be named, as when comparing two geometric designs. This impaired approach is typical of what Goldstein has called abstract as opposed to concrete thinking [ 3] . With the Wechsler Blocks, both coca-users and controls appear to have tackled the first and second designs by trialand-error methods, in which manual dexterity is able to compensate for a certain degree of mental difficulty, and hence their scores on these show little consistent difference. On the third design, however, the controls improve their speed and accuracy greatly, whereas the coca-users do not, and the probable explanation for this is that the controls have nearly all grasped the principle underlying the test while the coca-users are continuing with their trial-and-error approach. This is also an established sign of difficulty in the approach to abstract thinking, and the fact that on each trial at least four coca-users fail completely to produce the correct design leads further to the same conclusion. Finally, although we can trace no research concerning recall which yields the type of result which we obtained with the literate and illiterate controls (since such research on illiterates is very rare in any case) it seems possible that here again an aspect of abstract thinking is involved. Simple memory is usually fairly concrete, but schooling, even if only for two or three years, provides people with training in linking discrete concrete elements into abstract patterns or wholes which thereby permit each element to be recalled in a given order or on a given cue. While our results regarding recall are still something of a mystery, therefore, it seems very possible that the three respects in which the coca-users show definite deficits are all aspects of a general deficit in the ability for abstract thinking, whereas their normal performance on the simpler elements in each test, and in the manual tasks, indicate a retention of normal concrete thinking.

Defects of this sort have been described in some lobotomised patients, in patients with lesions of the frontal lobes, and occasionally in patients with lesions in other cerebral lobes if the symptoms associated with the localised area are sufficiently mild to permit such vaguer defect to be recognised. There is some indication that it may be related to loss of brain tissue, regardless of what lobe the loss affects, provided it is extensive enough, and this would open the way to seeing how it could be produced by a toxic process affecting the whole brain without apparently concentrating on any system. According to Goldstein, who is still the authority on certain aspects of brain damage, the patient with a deficiency in abstract thinking typically "appears passive, lacks initiative and decision, has great difficulty.., in making a choice, in finding out the essentials of a situation..." [ 3] . This sounds exceedingly like the descriptions which have been given of the heavier cocachewers in the high Andes and in Peru, even though our Argentinian subjects appeared less affected. Moreover, Goldstein notes that the patient with this defect is usually cooperative and peaceful, but that "retraining promises much less success in patients of this type" than in patients with a more dramatic but more localised deficit ( op.cit.). Our coca-chewing subjects are almost certainly less damaged than Goldstein's patients were, and the prognosis is therefore better, but on the other hand one must think of complete populations affected, to a greater or lesser degree, in this way.

The deficit is clearly deriving from the chronic use of coca, not from its recent use, but there is still a little that can be said regarding this recent use. On tests which involve manual dexterity, the continuing use of coca appears to provide subjects with a greater alertness or speed than if the subjects are habitu6s but abstain from the drug. When abstract thinking is called for as in the case of the more difficult items in the Army Beta test, on the other hand, it seems possible that continuing use of coca diminishes performance, although the evidence is unclear. And in concrete thinking not involving manual dexterity, continuing use of coca seems to make no difference.

In terms of social action, these conclusions suggest that as long as the habit persists, and for some years thereafter, older men will be handicapped in their ability to grasp the principles behind innovations which might be offered to them; innovations, for instance, connected with new farming systems, or the accounting methods needed by a marketing cooperative, or some new organisation of local government. To fully grasp such innovations needs an appreciation of the principles behind them, and it seems likely that the coca-users will be unable to acquire this. But on the other hand the coca-users would probably be quite able to grasp and thence to benefit from new techniques in farming and in industry, provided these were shown them in sufficiently concrete form and provided they could retain before their eyes an example of what they must do. The coca-user appears, on the basis of our findings, to be quite capable of learning industrial work and of performing it with acceptable efficiency, provided he is properly trained, provided the task is suitably broken down into simple stages, and provided he has not to remember too much. The same should be true for agricultural innovations, provided he can be shown the concrete results and is guided in his efforts instead of being expected to make or grasp long-range plans himself.

If the latter type of social change were introduced and did produce some alleviation of present miseries, the evidence from our study area and from elsewhere suggests that the coca habit would decline without police action, although some official discouragement would be desirable to accelerate the process. Until such change is introduced, however, it should be realised that present conditions of life scarcely permit the application of abstract thinking and that coca is probably doing as much good as harm, since it appears to help men to earn a living under highly difficult conditions, both social and psychological. While the habit is definitely undesirable, therefore, the well-being of the habitués needs to be improved at the social level before one is justified in expending gratuitous effort at eradicating it.



N. L. Fine: Coca chewing, a social versus a nutritional inter-pretation. Mimeograph, 1960.


D. Goddard & S. Goddard: The social conditioning of the use of coca among field labourers in Northern Argentina. Report prepared for the World Health Organization, 1967.


K. Goldstein: Aftereffects of Brain Injuries in War. New York, 1942.


J. C. Negrete, & H. B. M. Murphy: Psychological deficit in chewers of coca leaf, Bulletin on Narcotics, 1967, XIX (4), 11-18.