Introduction
Materials and methods
Results
Discussion
Acknowledgements
Author: Alfred A. BUCK, Tom T. SASAKI, Jean J. HEWITT, Anne A. MACRAE
Pages: 23 to 32
Creation Date: 1970/01/01
The effects of coca chewing on the health of the user and on society in general has long been a controversial question. Medical opinions concerning the effect of coca leaf chewing on health generally fall into three categories: (1) that coca chewing is harmful; (2) that it is harmful to some extent, but an unimportant problem at a time when many other social and health problems have priority; and (3) that coca-chewing is useful.
Dr. Pablo Oswaldo Wolff, with the assistance of the Division of Narcotic Drugs of the United Nations, undertook the task of assembling an annotated bibliography on the effects of chewing the coca leaf. He reviewed 119 items, and classified and assessed medical and non medical opinions on this subject (1). In their report, the Commission of Enquiry on the Coca Leaf of the United Nations concluded that in the present state of knowledge the effects of coca chewing are explainable by the action of cocaine (1). Some of the conclusions of this report were the basis for three hypotheses which were tested by this study.
Coca, by diminishing the sensation of hunger, has an unfavorable effect on the nutritional state of the habitual chewer. Malnutrition and conditions in which nutritional deficiencies are important disease determinants occur more frequently in coca chewers than among control subjects.
Coca chewing leads to a state of relative indifference which can result in inferior personal hygiene.
The work performance of coca chewers is lower than that of comparable non-chewers.
The investigations on coca were carried out within the framework of a comprehensive epidemiologic study of four contrasting Peruvian villages by a research team, including epidemiologists, a social anthropologist, an entomologist, a sanitary engineer, public health nurses, laboratory scientists and social workers. A more detailed description of the methods and achievements of the study is published elsewhere (2).
The study was supported in part by funds of the Medical Research and Development Command, United States Army.
A more detailed report of the work described in this article appeared in the American Journal of Epidemiology, Vol. 88, No. 2.
Relatively best suited for an investigation of the effects of coca chewing on health among the four selected populations was the community of Cachicoto, because the age, sex and ethnic distribution of the coca users permitted the conduct of a controlled study. In addition, coca ( Erythroxylon coca) was grown locally and freely available.
Cachicoto is a community with 492 residents. It is located in the Montaña at an altitude of 2,400 feet along the Rio Monzon in the District of Monzon, Province of Huamalies, Department of Huanuco. Plantations and the central core of the village have been in existence here for over 50 years.
The total population of Cachicoto residing within limits defined by a specially prepared map was enumerated in a house-to-house census conducted by the research team. The distribution of the population by age and sex is shown in table 1. Two major ethnic groups are represented in the community, i.e., Mestizos and Quechua Indians. Twenty-nine of the 115 households in the village belonged to Quechua Indians, who numbered 115 and represented 23.4 per cent of the total population of Cachicoto. Quechua [Indians are relatively new migrants who have come from their original home in the high Andes to the Montaña for resettlement and to find new opportunities for work on their own farms in jungle clearings or on plantations. Their assimilation into the predominantly Mestizo population has been slow. Generally their economic status is inferior to that of the Mestizo. Status differences between Quechua Indians and Mestizos exist also, as measured by such criteria as education and more subtle qualities as having "culture" or style of life.
Wearing of shoes, an important factor in the prevention of endemic hookworm infections, although more frequent among the Mestizos is related more to status than to health. Almost all of the persons who stated that they wore shoes revealed also that they removed them while walking in the almost daily rains, or when engaged in activities close to home, the predilected site for indiscriminate defecation.
Male Chewers |
Female Chewers |
Total Chewers |
|||||||
---|---|---|---|---|---|---|---|---|---|
Age |
Number |
No. |
percent |
Number |
No. |
percent |
Number |
No. |
percent |
0-9
|
93 | 0 | 0 | 79 | 0 | 0 | 172 | 0 | 0 |
10-19
|
56 | 2 | 3.6 | 44 | 2 | 4.5 | 100 | 4 | 4.0 |
20-29
|
28 | 7 | 25.0 | 37 | 6 | 16.2 | 65 | 13 | 20.0 |
30-39
|
36 | 11 | 30.6 | 24 | 7 | 29.2 | 60 | 18 | 30.0 |
40+
|
26 | 12 | 46.2 | 28 | 6 | 31.4 | 54 | 18 | 33.3 |
TOTAL
|
239 | 32 | 13.4 | 212 | 21 | 9.9 | 451 | 53 | 11.8 |
* No information was available from 41 residents.
Altogether there were 53 coca chewers of whom 28 were Quechuas and 25 Mestizos. Each coca chewer was matched with a control subject of the same sex, ethnic group and ten-year age group. The selection of controls was achieved from random selections of IBM cards which contained only the vital statistics of an individual and were pre-sorted for sex, ethnic affiliation and the corresponding ten-year group. It is emphasized that each phase of the study was conducted independently by the various specialists. The complete data became available only at the end of the study; it was then that the individuals who chewed coca were matched with control subjects who had given negative responses to the specific coca questions in the routine interviews.
Matching was achieved for 51 of the 53 coca chewers; two persons had to be excluded because of lack of individuals with the corresponding characteristics in the control population. Once matched, all of the relevant data for the study were combined on one IBM card for each individual. Whenever a particular measurement or test was unavailable from either member of a matched pair, the pair was withdrawn.
Of the 51 matched pairs both partners had had health interviews, 49 had physical examinations and haemoglobin determinations, in 46 haematocrit readings were available, 44 had completed tuberculin tests, 41 stool examinations and from 34 pairs a venous blood sample was available.
The two groups of chewers and controls were also comparable with respect to their residence histories. Forty-seven (92.1 per cent) of the coca chewers were born in Huanuco, the Department in which Cachicoto is located, compared with 42 or 82.3 per cent of the controls. The average time of residence in the village was 9 years for the chewers and 9.8 years for the controls.
Differences between the two groups existed with regard to their religious affiliation, there being significantly more Protestants in both ethnic control groups. Coca chewing, smoking, drinking of intoxicants and dancing are not condoned by the Evangelists of Cachicoto, and conversion from the traditional Catholic belief to this Evangelical denomination requires the individual to abandon the coca habit.
A detailed description of the general methods of the study is published elsewhere (2). Social anthropologic information was obtained by routine interviews of heads of households and by the key informant technique (3). Quechua Indians were identified from interviews with heads of households, ability of the respondent to speak the Quechua language, and by identification of Indian residents by community leaders from a complete list of households.
The medical information was obtained by four methods: interviews, physical examinations, laboratory tests and skin tests. Household interviews were conducted by two experienced social workers. The physical examinations were standardized and restricted to systematic observations and measurements of selected physiologic variables and pathologic signs. The results were recorded on precoded forms. Routine measurements included body weight and height, as well as skinfold thickness of the back of the upper arms at the mid-posterior midpoint between the acromion and olecranon with the elbow in 90° flexion. The average of 2 measurements was recorded. Also recorded were gross deformities and incapacitations, skin diseases, selected nutritional deficiencies, specific categories of eye infections and diseases, types of oedema, presence of hepato-splenomegaly, and scars resulting from smallpox disease and vaccinations.
Two tuberculins, PPD-S and PPD-G, were used in the skin tests. One tenth of a milliliter, equivalent to five TU of each antigen, was injected intracutaneously into the skin of each forearm; the tests were then read 48 hours after the injection of the antigen.
Haemoglobin determinations were made in the field with a haemoglobinometer (Spencer-A0 1000 D, American Optical Company). The haematocrit was determined by the capillary tube method (4). Stool specimens for parasitologic examinations were collected from most individuals in the study. A portion of all specimens was returned to Baltimore and examined by the merthiolate-formalin-ether concentration method (5). All protozoa and helminth eggs or larvae were identified and recorded. In addition, rough estimates of the numbers of hookworm eggs were tabulated.
The serologic examinations included the following performed by the procedures described. Total serum protein and cholesterol determinations were made with spectrophotometric methods (6, 7). Electrophoretic assays were performed by the Microzone-Analytrol method (8, 9). Flocculation tests included the Venereal Disease Research Laboratory (VDRL) cardiolipin flocculation test (10), the cercarial antigen schistosome flocculation test (11), and the trichinosis charcoal card test (12). The fluorescent treponemal antibody test (FTA-200) was applied routinely (13).
Hemagglutination inhibition (HI) tests for arboviruses included the following antigens selected for Peru: Group A: Eastern Equine encephalitis, Western Equine encephalitis, Mayaro (14). Group B: Yellow fever A 317, St. Louis encephalitis 1182-58, Dengue II A336 and Ilheus (15). Group C: Marituba Bran 15; and of other selected viruses Guama Be An 277, Magurai Bear 7272 and Guaroa (16).
The test most frequently employed to examine the statistical significance of observed differences between the coca and control group was the Chi-square test for matched pairs (17).
In the first part of this section the three hypotheses listed in the introduction are tested by systematic comparisons of the results of routine measurements which (1) are more or less direct indicators of the nutritional state; (2) reflect personal hygiene; and (3) indicate differences in the work performance of the coca and control group.
The following measurements of the study are related to the nutritional state of an individual and will be compared for differences between the coca chewers and their matched control partners. Figure 1 shows the frequency distribution curves of the weight/height ratios of the coca and control group. By comparing individual members of each matched pair the mean difference was not significant statistically for the available sample size.
Skinfold thickness is associated with subcutaneous fat thickness and can be considered as another relative index of the nutritional state, especially when the groups of persons to be compared are similar with respect to their physical activity, as are the predominantly peasant residents of Cachicoto. A pair by pair comparison between coca leaf chewers and controls reveals that the mean difference of the former was 1.17 millimeters less than that of the controls (P = .10).
Three serochemical tests were applied routinely in the Peruvian studies: the determination of total serum proteins (TPS), cholesterol, and an analysis of the major protein fractions by electrophoresis. The results of each of these tests are related to nutritional variables and can therefore serve as indicators of differences in the nutritional state of coca chewers and control subjects. Differences in total proteins average 0.53 gm% and are on the borderline of statistical significance (P = .07). Further analysis for differences by comparing the individual protein fractions separated by electrophoresis reveals that the lower total serum proteins of coca chewers have resulted mainly from corresponding differences in the serum albumin levels. Figure 2 shows the distribution curves of serum albumin for the two groups. The mean difference between chewers and controls was 0.36 gm% and was statistically significant (P = .05). A highly positive correlation between serum albumin levels and the nutritional state has been found by various investigators (18-20).
As can be seen from figure 3, serum cholesterol values are generally lower among coca chewers. The mean difference is 20.3 mg% and is statistically significant (P = .01).
The second of the three hypothetical effects of coca to be investigated is related to standards of personal hygiene. Of the various routine measurements included in the comprehensive epidemiologic studies the following were thought to be relatively uncomplicated indicators of personal hygiene: (1) the prevalence of scabies and of pyoderma as determined in standardized physical examinations; (2) the prevalence of intestinal parasites as determined in stool examination of single specimens using concentration techniques; (3) the frequency of rat bites as elicited from interviews.
Coca chewers with condition |
Controls with condition |
|||||
---|---|---|---|---|---|---|
Skin disease or condition |
No. |
Per cent |
No. |
Per cent |
Difference in per cent |
P * |
Pyoderma (folliculitis, impetigo, furuncles)
|
8 | 16.3 | 1 | 2.0 |
+14.3
|
0.03 |
Scabies
|
6 | 12.2 | 1 | 2.0 |
+10.2
|
0.10 |
Scars from leishmaniasis
|
4 | 8.2 | 2 | 4.1 |
+4.1
|
>0.30
|
Pinta, typical depigmentation
|
1 | 2.0 | 2 | 4.1 |
-2.1
|
>0.30
|
Based on Chi-square test for matched pairs.
Table 2 shows the prevalence, not only of scabies and pyoderma as two of the selected indices of personal hygiene, but also of leishmaniasis ("uta") and pinta lesions, as two of the regionally important diseases which have characteristic skin manifestations. Of the four conditions listed, only the difference in the prevalence of pyoderma was significant statistically (P = .03) for the available sample size of 49 matched pairs.
Table 3 compares the prevalence of infections with intestinal parasites between coca chewers and matched controls. As can be seen, under the prevailing environmental conditions which characterize Cachicoto, it appeared that the habit of coca chewing per se had contributed little to the high prevalence of infections with nematodes and protozoa, disregarding the possibility of differences in the intensity of infections. The difference in the prevalence of infections with ascaris was on the borderline of statistical significance (P = .08), but not for hookworm infections where a difference as large as or larger than the one observed could have arisen by chance alone in 20 per cent of samples of the same size. Nevertheless, there are two notable exceptions which deserve consideration. The first is the paradoxical reduction in the percentage of coca chewers infested with amoebae, including E. histolytica, E. coli, E. nana and I. butschlii (P = 0.07), and the second is the presence of infections with the potentially pathogenic Ciliatea Balantidium coli among the chewers. Whether or not coca leaves may have amebicidal properties is an interesting speculation which deserves further study.
Stool positive |
||||
---|---|---|---|---|
Coca chewers |
Controls |
|||
Parasite |
No. |
Per cent |
No. |
Per cent |
Ascaris lumbricoides
|
36 | 87.8 | 29 | 70.7 |
Trichuris trichiura
|
39 | 95.1 | 39 | 95.1 |
Hookworm
|
30 | 73.2 | 24 | 58.5 |
Entamoeba histolytica
|
||||
small form
|
4 | 9.8 | 10 | 24.4 |
large form
|
2 | 4.9 | 2 | 4.9 |
Entamoeba coli
|
25 | 61.0 | 30 | 78.0 |
Endolimax nana
|
3 | 7.3 | 4 | 9.8 |
Idoamoeba butschlii
|
1 | 2.4 | 6 | 14.6 |
Giardia lambllia
|
1 | 2.4 | 0 | 0 |
Balantidium coli
|
4 | 9.8 | 0 | 0 |
A comparison was made of the 3 most frequently reported animal bites experienced by the residents of Cachicoto, as elicited from personal interviews. In addition to rat bites, selected as an indicator of sanitary house conditions, snake and bat bites, because of their local importance in the area, were included. Of the observed differences only the percentage of rat bites approached the statistical significance level of 5 per cent.
The third hypothesis to be tested concerns itself with the inferior work performance of the coca chewers. Contrary to the view held by coca chewers who think of their work under the influence of coca as superior, medical experts and the Commission of Enquiry on the Coca Leaf believe work performance to be inferior to a comparable group of non-chewers. Whereas an appraisal of the work performance based on direct measurements made by the research team is not possible, other systematically acquired data of the study permit a comparison which reflects some aspects of the working capacity of the coca chewers and controls.
In routine interviews each individual was asked whether he had lost work days because of illness during the month immediately preceding the physical examination and, if the answer was yes, the number of days thus lost. The results are shown in table 4. There were not only more chewers than non-chewers who had been ill, but those with illness also reported more lost work days than the control subjects who had been sick. An index of incapacitating illness combining both the number of sick persons and the duration of illness is shown in the right column of table 4, which lists almost twice as many person-days of illness per month for the coca chewers over controls.
Persons ill |
Work days lost |
||||
---|---|---|---|---|---|
No. |
Per cent of total |
No. |
Average duration of illness in days |
Illness in person-days per month |
|
Coca chewers
|
19 | 38.0 | 139 | 7.3 | 2.78 |
Controls
|
11 | 22.0 | 84 | 7.6 | 1.68 |
Comparisons between coca chewers and matched controls for differences in distributions of physiological variables and in prevalence of conditions which indicate variations of the nutritional state, personal hygiene and the work capacity appear to support the hypotheses stated. In the following section, health conditions and diseases with a more complex pathogenesis will be studied in which the hypothetical effects of coca are known or suspected disease determinants.
Table 5 lists lifetime histories of "severe illnesses" based on statements made in routine health interviews. Of all diseases mentioned only the difference in the percentage of persons who had suffered from "severe anaemia" was large and statistically significant. This finding is not only confirmed but amplified when the results of blood examinations for haemoglobin and haematocrit are compared. Figure 4 gives the frequency distributions of haemoglobin and figure 5 presents the curves of the hematocrit values of coca chewers and matched controls. A large number of the coqueros had hemoglobin levels of less than 10 grams per cent.
Clinical studies, including microscopic examinations of bone marrow from sternal punctures, revealed that anemia of the coca chewers was hypochromic with all signs of iron deficience. Because of the high prevalence of hookworm infections in the community, a comparison for differences in their severity was made as the first step in the analysis for etiologic clues. The results are shown in table 6, which includes all pairs of coca chewers and controls from whom stool specimens were available for examination. The three categories listed to describe the hookworm load are based on the following criteria: "few" eggs were reported when between 1 to 10 hookworm eggs were detected on a single slide; "moderate" infections were recorded when the number of eggs detected ranged between more than 10 per slide but less than one per low power field; and "numerous" eggs were registered when the number of the eggs exceeded one per low-power field. The results indicate that coca chewers had slightly heavier infections than controls. In both groups hemoglobin mean levels decreased as the hookworm loads became heavier. Nevertheless, regardless of the intensity of hookworm infection in each category, coca chewers had significantly lower hemoglobin levels than the controls; the difference between the two groups widened as the hookworm load became heavier.
Because of the known interaction between hookworm infection and nutrition in the pathogenesis of hookworm disease (21), the hemoglobin and hematocrit values of coca chewers and controls were also compared for differences in corresponding classes based on the weight-height index. The results are shown in table 7. They indicate that in all weight-height categories coca chewers had lower hemoglobin and hematocrit levels than the control subjects.
Coca chewers with condition |
Controls with condition |
|||||
---|---|---|---|---|---|---|
Condition |
No. |
Per cent |
No. |
Per cent |
Difference |
P * |
Lifetime history of :
|
||||||
Smallpox
|
4 | 7.8 | 2 | 3.9 |
+3.9
|
0.3 |
Malaria
|
11 | 21.6 | 14 | 27.4 |
-5.8
|
0.3 |
Yellow fever
|
2 | 3.9 | 1 | 2.0 |
+1.9
|
>0.3
|
Severe anemia
|
12 | 23.5 | 2 | 3.9 |
+19.6
|
<0.01
|
Severe gastro-intestinal diseases
|
3 | 5.9 | 2 | 3.9 |
+2.0
|
>0.3
|
Leishmaniasis
|
2 | 3.9 | 2 | 3.9 | 0 |
>0.3
|
Tuberculosis
|
0 | 0 | 2 | 3.9 |
- 3.9
|
0.3 |
Values for P listed are based on Chi-square tests for matched pairs.
Among the residents of Cachicoto, and from observations of other investigators (1), coca leaf chewing and alcoholism are frequently found together, but the cause or effect of this association is still unknown. Because liver damage has been found associated with alcoholism (22) and in chronic experimental intoxication with cocaine (23) it was of interest to compare the prevalence of hepatomegaly between coca chewers and controls. Table 8 lists the percentage of hepatomegaly and splenomegaly for the two groups. The prevalence of hepatomegaly among the users of coca is almost twice as high as among the non-chewers.
Infections with Mycobacterium tuberculosis can be expected to have occurred more frequently among persons with inferior hygienic standards. As was shown previously in table 5 there was no evidence from lifetime histories that tuberculosis had occurred more frequently among coca chewers than among their controls. Further information on the subject is available from the results of tuberculin skin tests which were routinely applied to all residents of Cachicoto who participated in the study. Forty-one per cent of the coca chewers and 43.2 per cent of the controls were classified as reactors to 5 tuberculin units of PPD-S.
Certain locally endemic arthropod-borne infections can serve as indices for conditions which are hypothetically unrelated to coca chewing because exposure of the residents of Cachicoto to flying insect vectors is universal. It was therefore interesting to compare coca chewers and controls for dissimilarities in the prevalence of antibodies to locally important arthropod-borne virus infections. Differences in antibody titres in haemagglutination inhibition tests with 3 locally important arboviruses (Mayaro, Ilheus, Yellow Fever) between the coca and control group are trivial (2). Nevertheless, true differences in the prevalence of naturally acquired infections with Arbo-B viruses could have been obscured by compensating differences in the proportion of chewers and non-chewers who had received yellow fever vaccinations. Therefore vaccination histories were compared, including smallpox vaccinations for completeness. Similar percentages of both groups had received vaccinations against smallpox and yellow fever. The prevalence of arthropod-borne diseases was also found similar when the life histories of coca chewers and controls were compared for differences in the per cent who had had malaria, leishmaniasis and yellow fever.
Coca chewers |
Controls |
||||||
---|---|---|---|---|---|---|---|
Hookworm eggs |
No. |
Per cent of total |
Mean hgb. in gm per cent |
No. |
Per cent of total |
Mean hgb. in gm per cent |
Difference |
Negative
|
11 | 26.8 | 10.6 | 17 | 41.5 | 12.4 | 1.8 |
Few
|
13 | 31.7 | 11.1 | 14 | 34.1 | 12.6 | 1.5 |
Moderate
|
15 | 36.6 | 9.6 | 7 | 17.1 | 11.2 | 1.6 |
Numerous
|
2 | 4.9 | 7.7 | 3 | 7.3 | 10.8 | 3.1 |
TOTAL
|
41 | 100.0 | 10.4 | 41 | 100.0 | 12.2 | 1.8 |
Coca chewers |
Control subjects |
|||
---|---|---|---|---|
Condition |
No. |
Per cent |
No. |
Per cent |
Hepatomegaly
|
11 | 22.0 | 6 | 12.0 |
Splenomegaly
|
1 | 2.0 | 0 | 0 |
Another highly prevalent infection among the residents of Cachicoto is the treponematosis pinta. Results of the Fluorescent Treponemal Antibody Test (FTA-200) and the Venereal Disease Research Laboratory (VDRL) slide flocculation test reveal that the percentages of reactions in both tests are similar.
The main results of the study and their statistical significance are summarized in table 9.
The results of the study support each of the three hypotheses tested: that coca chewing is associated with an inferior nutritional state; that chewers have a higher prevalence of conditions resulting from poor personal hygiene; and that the work performance of coca chewers is inferior to that of a comparable control group because of a relatively higher percentage with intercurrent incapacitating illnesses. Nevertheless, in the interpretation of the results of the study the complexity and the interactions of environmental and host factors which have to be considered become quite obvious. The directions and sequences of causes and effects cannot be identified clearly, because the conditions recognized by the study as probable disease determinants are arranged in a vicious circle. The desire to inhibit the feeling of fatigue and hunger (1), the easy availability of coca, and the culture-associated glorifications of its use may be regarded as basic factors of the chewing habit. Repeated use of coca brings with it the desire to maintain euphoria. Habitual coca chewing can lead to or maintain a state of malnutrition which, in turn, increases the desire for more coca (1). In contrast to the high Andes, the original home of the Quechuas, where hookworm infections are absent (2), residents of Cachicoto are universally exposed to infections with this nematode (2). Thus, a probably unsubstantial disadvantage of coca chewers in the high Andes becomes overt in this tropical environment where the relative risk of developing anaemia (haemoglobin <10.0 gm%) was found to be 5 times that of the control subjects. The reasons for the difference in the pathogenicity of hookworm infections in the two groups of hosts can be seen in the higher prevalence of nutritional deficiencies among coca chewers (24). Besides the nutritional inadequacies recognizable from routine measurements employed in the study, it can also be assumed that the degree of iron deficiency among coca chewers must have been considerable (21).
Characteristic |
Number of pairs * |
Mean difference between coca chewers and controls |
d/SE diff |
P |
---|---|---|---|---|
Ratio: body weight (lbs)/height (cm)
|
47 |
-0.03
|
1.70 | 0.10 |
Skinfold thickness
|
47 |
-1.17 mm
|
1.65 | 0.10 |
Total serum protein
|
34 |
-0.53 gm %
|
1.96 |
<0.10 >.05
|
Serum albumin
|
34 |
-0.36 gm %
|
2.1 | 0.05 |
Serum cholesterol
|
34 |
-20.3 mg %
|
2.75 | 0.01 |
Haemoglobin
|
48 |
-2.0 gm %
|
4.34 |
<0.01
|
Haematocrit
|
46 |
-5.5 % of packed cells
|
3.12 |
<0.01
|
* Includes only the originally matched coca chewers and controls when test results from both partners in a pair were available.
In addition to the large differences in the nutritional state and, to a lesser degree, in the hookworm load of coca chewers and controls, one might also consider a more direct local effect of coca through the actions of the alkaloid cocaine on the duodenum of the host or on the worms.
The study has shown that under the peculiar constellation of environmental and host factors which characterize Cachicoto, habitual coca chewing is associated with poor health.
The authors acknowledge, with gratitude, the competent help of Miss Olga Degregori and Miss Aida Vainstein, the advice of Dr. Roberto Cornejo Ubillus, and the contributions of Mr. Filipe Sipan in the field studies. The authors also express their appreciation for the direction given by Dr. Javier Arias Stella, Minister of Public Health, and Dr. Carlos Quiros, former Director-General of the Ministry of Public Health and Social Assistance of Peru.
United Nations, Report of the Commission of Enquiry on the Coca Leaf, Economic and Social Council, 1950, Special Supplement No. 1, Lake Success, New York
002A. A. Buck, T. T. Sasaki, and R. I. Anderson. Health and disease in four Peruvian villages: Contrasts in epidemiology. Baltimore, The Johns Hopkins Press, 1968.
003Marc-Adélard Tremblay, The key informant technique: A non-ethnographic application. Amer. Anthropol., 1957, 59 : 688-701.
004M. M. Strumia, A. B. Sample and E. D. Hart, An improved microhaematocrit method. Amer. J. Clin. Path., 1957, 24 : 1016-1024.
005W. Blagg, E. L. Schloegel, N. S. Mansour and G.I. Khalaf. A new concentration technique for the demonstration of protozoa and helminth eggs in feces. Amer. J. Trop. Med. Hyg., 1955, 4 : 23-28.
006R. J. Henry, C. Sobel and S. Berkman. Interferences with biuret methods for serum proteins. Use of Benedict's qualitative glucose reagent as a biuret reagent. Anal. Chem., 1957 , 29 : 1491-1495.
007George V. Mann, A method for measurement of cholesterol in blood serum. Clin. Chem. Acta, 1961, 7: 275-284.
008B. W. Grunbaum, J. Zec and E. L. Durrum, Application of an improved microelectrophoresis technique and immunoelectrophoresis on serum proteins on cellulose acetate. J. Microchem., 1963, 7: 41-53.
009B. W. Grunbaum, M. F. Lyons, M. V. Carroll and J. Zec, Quantitative analysis of normal human serum proteins on permanently transparentized cellulose acetate membranes. J. Microchem., 1963, 7: 54-56.
010Training Manual, 1967. Serologic Tests for Syphilis; Venereal Disease Research Laboratory, N.C.D.C., Chamblee, Georgia.
011R. I. Anderson, Serologic diagnosis of Schistosoma mansoni infections . Amer. J. Trop. Med. Hyg., 1963, 9: 600-603.
012R. I. Anderson, E. H. Sadun and M. J. Schoenbechler, Cholesterol-lecithin slide (TsSF) and charcoal card (TsCC) flocculation tests using an acid soluble fraction of Trichinella spiralis larvae. J. Parasit., 1963, 49, 642-647.
013W. E. Deacon, E. M. Freeman and A. Harris, Fluorescent treponemal antibody test, modification based on quantitation (FTA-200). Proc. Soc. Exp. Biol. Med., 1960, 103 : 827-829.
014D. H. Clarke and J. Casals, Techniques for hemagglutination and hemagglutination-inhibition with arthropod-borne viruses. Amer. J. Trop. Med. Hyg., 1958, 7: 561-573.
015H. M. Hamman, D. H. Clarke and W. H. Price, Antigenic variation of West Nile virus in relation to geography. Amer. J. Epid., 1965 , 82 : 40-55.
016D. H. Clarke, Further studies on antigenic relationships among the viruses of the group B. tick borne complex. Bull. W.H.O., 1964, 31: 45-56.
017W. B. Cochran, The comparison of percentages in matched samples. Biometrika, 1950, 37: 256-266.
018C. C. Curtain, D.C. Gadjusek, C. Kidson, D. Gorman, L. Champness and R. Rodrigue, A study of the serum proteins of the peoples of Papua and New Guinea. Amer. J. Trop. Med. Hyg., 1965, 14: 678-690.
019F. Wuhrmann and C. Wunderly, Die Bluteiweisskoerper des Menschen. Basel (Switzerland). Benno Schwabe and Co., 1952.
020A. Hottinger, O. Gsell and E. Hehlinger, Hungerkrankheit, Hungeroedem, Hungertuberkulose. Basel (Switzerland), Benno Schwabe and Co., 1948.
021M. Roche and M. Layrisse, The nature and causes of "hookworm anemia". Amer. J. Trop. Med. Hyg., 1966, 15 : 1031-1102.
022M. Terris, Epidemiology of cirrhosis of the liver: National mortality data. Amer. J. Pub. Hlth., 1967, 57: 2076-2088.
023C. Gutierrez-Noriega and Vicente Zapata Ortiz, Estudios sobre la coca y la cocaina en el Peru. Ministerio de Educacion Publica, Lima, 1947.
024G. F. Otto, Hookworm. In Maxcy-Rosenau. Preventive medicine and public health, 9th ed. P. E. Sartwell, ed., New York, Appleton-Century Crofts, 1965.