Mantakassa: an epidemic of spastic paraparesis associated with chronic cyanide intoxication in a cassava staple area of Mozambique. 2. Nutritional factors and hydrocyanic acid content of cassava products*

Bulletin ofthe World Health Organization, 62 (3): 485-492 (1984) © World Health Organization 1984 Mantakassa: an epidemic of spastic paraparesis associated with chronic cyanide intoxication in a cassava staple area of Mozambique. 2. Nutritional factors and hydrocyanic acid content of cassava products* MINISTRY OF HEALTH, MOZAMBIQUE1 An outbreak of spastic paraparesis which mostly affected women and children occurred in a northern province of Mozambique in 1981. The epidemic was related to chronic cyanide intoxication associated with a diet consisting almost exclusively ofcassava. A prolonged drought in the area had exhausted allfood resources except cassava, especially the bitter varieties. A nutritional, toxicological and botanical investigation was carried out in two of the five districts affected. The main findings were that cyanide levels were unusually high in the cassava plant as a consequence of the drought with daily intakes estimated at 15-31.5 mg HCN. Detoxification of the bitter varieties by sun-drying was inadequate because of the general food shortage, and metabolic detoxification was probably reduced owing to the absence of sulfur-containing amino acids in the diet. The raw and dried uncooked cassava was eaten mostly by women and children. The nutritional status of the population, however, was not very poor and symptoms of advanced under- nutrition were rarely seen. An outbreak of spastic paraparesis occurred be- tween May and October 1981 in Nampula Province, northern Mozambique, and involved five districts: Memba, Erati, Nacala, Maiaia and Monapo. The dis- ease was called mantakassa, which means paralysis in the local language. The outbreak occurred in a drought-stricken rural area, where the staple food was cassava, and affected mostly women and chil- dren. The epidemiological, clinical and laboratory findings suggested that the disease was associated with cyanide intoxication and cassava consumption (see Part 1). Recently a neurological disease as- sociated with chronic cassava intoxication was de- scribed in Nigeria (1). Mantakassa also shows clinical and epidemiological similarities with lathyrism in India (2). This second part of the study describes the nu- tritional, toxicological, and botanical investigations * Part I appears on pages 477-484 in this issue of the Bulletin. Requests for reprints should be sent to: Secqao de Epidemiologia, Direccao Nacional de Medicina Preventiva, Ministerio da Sauide, Maputo, Mozambique. This text was prepared for the Ministry of Health, Mozam- bique, by E. Casadei, Head, Food and Water Hygiene Laboratory, National Directorate of Preventive Medicine, Ministry of Health, Maputo, Mozambique; P. Jansen, Botanist, Department of Tra- ditional Medicine, National Directorate of Preventive Medicine; A. Rodrigues, Nutritionist, Department of Nutrition, National Directorate of Preventive Medicine; A. Molin and H. Rosling, for- mer District Medical Officers, Nacala Hospital, Nacala, Mozam- bique. that were performed to test the possibility of a food- related cause. These investigations were carried out in September and October 1981 in the districts of Memba and Maiaia. The objective was to study the agricultural and nutritional situation in normal years, to compare this with the changes that occurred in 1981, and to investigate the relationship between cassava consumption and the high thiocyanate levels found in patients with the disease. The fieldwork was done in the short time available with limited resources. The sample size for the nu- trition investigation was small and the results could not, therefore, be subjected to statistical analysis. They do, however, indicate that nutritional and other factors could have played a part in the etiology of the disease. Ecological description of the area The area affected by the drought in 1981 can be divided into three distinct ecological and agricultural zones: coastal, intermediate, and interior. The usual rainy season extends from December to March with a mean annual rainfall of about 700 mm along the coast, increasing to 1000 mm in the interior (Fig. 1). Table 1 shows the distribution of crops in the three zones, based on a national agricultural survey carried out in Mozambique in 1961-66 (3). This distribution has not changed. 4422 -485- MINISTRY OF HEALTH, MOZAMBIQUE Fig. 1. Altitude and mean annual rainfall in the area affected by the drought. The dry and sandy coastal zone, 10-12 km wide, has a severely limited agricultural capacity. Cereals are rarely grown while cassava occupies 86'Vo of the cultivated area. The predominant cash crops are cashew and sisal. The interior zone, which is situated more than 60 km from the coast at an altitude of more than 200 m above sea level, has good agricultural con- ditions. Although much sorghum is grown, cassava is still the main staple food; cotton is the cash crop. The intermediate zone lies between 10 km and 80 km from the coast at an altitude of 50-200 m above sea level. The vegetation consists of secondary deciduous forest with deciduous parkland, open bushland, and spiny savanna (4). The number of cashew trees per peasant is among the highest in northern Mozam- bique; this zone also marks the eastern limit where cotton can be grown. Cassava accounts for 68% of the cultivated area of food crops, with smaller pro- portions for maize and sorghum. The main protein suppliers in the diet are beans and groundnuts. The present study refers mainly to the intermediate zone where the disease incidence was highest. History of cassava in Mozambique Cassava was probably introduced into Nampula Province, around 1760, from the French colonies in the Indian Ocean (5). It became established as the main staple crop along the coast in the period between then and the beginning of the present century. From Table 1. Distribution of crops in the family sector in the three agricultural zones % of area with crops Crops Interior Intermediate Coastal zone zone zone Cassava 59.5 68.5 86.0 Sorghum 18.6 6.6 2.6 Maize 3.3 5.8 1.8 Rice 3.3 0 2.6 Peanuts 10.5 11.0 2.1 Beans 4.2 8.0 3.8 Cotton (as % of the area with 27.2 22.2 0 traditional crops) Cashew (mean number of trees 21.9 51.0 45.3 per peasant) 1926 onwards, the Portuguese colonial power intro- duced a system of forced cultivation of cash crops, like cotton and cashew, which had a negative effect on the subsistence economy. It was during this period that cassava gained prominence as a staple food crop. Its cultivation calendar did not coincide with that for cotton, which resembled that for cereals. The peasants were also obliged to grow cassava for com- mercial purposes, to help feed the growing number of plantation and urban workers (6). In recent years, cash crop production has diminished owing to a com- bination of factors such as the abolition of forced labour after the country's independence and a break- down of the commercial network. As a consequence of the slave trade, cassava became the staple food in many parts of Africa. It yielded more energy per hectare and per man-hour than cereals; and the fact that its edible parts were not needed for sowing and that it was remarkably resist- ant to drought and pests contributed to its spread. Today, it provides around 4007o of the food energy requirements in Africa. MATERIALS AND METHODS From August to October 1981 information was obtained on the agricultural and nutritional situ- ation from a variety of sources - local authorities, farmers, teachers, traders, and the general popu- lation. Systematic investigations were made of the nu- tritional situation and samples of cassava products were collected from the following affected areas: 486 HYDROCYANIC ACID CONTENT OF CASSAVA - a communal village, Acordo de Lusaka, 20 km south of Memba; - the administrative division of Calawane, 40 km northwest of Memba; and - the administrative division of Muentaze, 20 km south of the port town of Nacala. The communal village was chosen for study be- cause it had a high incidence of the disease with 28 cases in a population of 815. Seventeen of the affected families were interviewed, as well as sex- and age- matched controls in 17 neighbouring families. Eight affected families in Calawane, which had an inci- dence of 10.7 per 1000 inhabitants, and 6 affected and 4 neighbouring families in Muentaze were also inter- viewed. The nutrition investigation included the collection of information on current agricultural practices, nu- tritional habits and customs before and during the drought, and socioeconomic conditions. In addition, a review of family food consumption during a 24- hour period and an assessment of nutritional status using anthropometry were carried out in the com- munal village. Weight-for-height and weight-for-age parameters were used to indicate the degree of "wasting" (7) or recent loss of body weight due to reduced food intake; the cut-off points were 7507o and 90%o for weight-for-height and 600o and 80%o for weight-for-age of the Harvard Standards, which cor- respond to those currently in use by the Ministry of Health in Mozambique. Various samples of fresh and processed tubers and leaves of cassava were collected for cyanide analysis using uniform collection methods. Fresh samples were placed in a plastic bag immediately after collec- tion, labelled and frozen on the same day, then trans- ported to the laboratory in cold boxes and frozen again immediately on arrival. Dry samples were placed in paper bags until they were analysed, with- out any special storage or transport conditions. The samples were collected from many different parts of the affected area and were analysed for cyanide con- tent at the National Food and Water Hygiene Labora- tory of the Ministry of Health in Maputo, using the alkaline titration method (8) and the direct potentio- metric method with a cyanide specific electrode (AMEL 201 I). Samples of domesticated and wild plants consunied by the inhabitants were also collected, identified, and recorded. RESULTS Cassava cultivation and diet in normal years Cassava constitutes about 800Vo of the staple food in the affected area. Twelve cultivated varieties were identified, two of which were considered to be bitter. It was estimated that the latter accounted for about 30%o of the cassava cultivated, but this was higher in some areas. Gurue, the more common of the bitter varieties, had been rejected when first introduced in the 1950s, but in recent years was more readily accepted because of its high yield and resistance to drought and attacks from wild animals. The entire cassava crop is harvested and replanted in August. The fresh cassava is cut into 20-cm-long pieces, sun-dried for 2-3 weeks, and then stored for up to 12 months. The population believes that this long drying period is necessary to make the cassava safe for consumption. Fig. 2 shows that families always reserved some of the previous year's dried cas- sava to cover the current harvest and drying period. In these areas the population uses only the sun- drying method of detoxification for bitter cassava; this is eaten uncooked or boiled or, most commonly, after pounding into flour which is mixed with boiling water to make a thick paste called chima. The fresh sweet varieties are eaten raw or boiled; they are also dried and processed as for the bitter varieties. Cooked cassava is usually eaten with a relish made from the leaves of the sweet cassava, pumpkin or sweet potato plants, beans, groundnuts, dried fish, tomatoes or coconut depending on availability. People normally eat two meals per day; one normally contains a cereal when this is available in the period after the cereal harvest. Meat and game are rare in the diet; pork is not eaten for religious reasons. Various fruits are eaten but the cashew apple is used to prepare an alco- holic drink. Cashew nuts are roasted and seldom used in cooked dishes. Effect of the drought on the harvest The rainy season normally extends from December to March, but in 1979-80 the rains failed and in 1980-81 the area had the lowest rainfall ever re- membered or recorded. Fig. 3 shows the low flow of 8 Sovgh.m Cassawv Toxic o \ ~~~~ ~ ~~~~~Toxict; 'X,.......":"& ............K i97r 1@|0*@ T IM IN MONTH$ Fig. 2. Availability of staple foods (maize, sorghum, and cassava) from 1979 to 1981 in the area affected by drought. 487 MINISTRY OF HEALTH, MOZAMBIQUE Flow Mean ofprevious E13 years 200 ' 2 |1 980-81 100 r 0 N DJ F M A MJ J A S Fig. 3. Flow of the Mecuburi river in the affected area in 1980-81, compared with mean values during the pre- vious 13 years (abscissa shows months from October (0) 1980 till September (S) 1981). the Mecuburi, the principal river in the most affected area in 1981, which was due to the severity of the drought. Fig. 2 shows the period of availability after harvest- ing of the three main staple food crops (maize, sor- ghum and cassava) from January 1979 to October 1981. Two important points should be noted: firstly, in 1981 cassava was harvested as early as June, while normally it is harvested in August; and second, in the middle of 1981, only the freshly harvested, toxic cas- sava was available, whereas in 1979 and 1980 some of the previous year's harvest lasted until September. The production of cereals and legumes was very low or completely lost owing to the drought. Sweet varieties of cassava also suffered severe losses but the bitter varieties produced a reasonable harvest. As a result of the poor harvests, edible wild plants, 110 of which were collected for identification, became an important constituent of the diet. By the end of June 1981 all normal means of pro- curing food were exhausted and it became necessary to harvest the cassava. People complained that the surviving sweet cassava had turned bitter, and that the bitter varieties were more bitter than usual. The fresh tubers were cut into 1 x 1 cm pieces and/or pounded and dried for 1-2 days only. As other foods were scarce, the population began to consume this in- adequately dried cassava. Because the cashew crop also had failed and work was difficult to obtain, ready cash was in short supply. Thus, even when food was occasionally avail- able, some families did not have the means to buy it. Diet during the drought Table 2 presents the results of the 24-hour recall on food consumption, which was carried out in the communal village. The low economic status of the affected families is demonstrated by their failure to join the food Consumer Cooperative which required Table 2. Results of a 24-hour recall of food consumption by affected and control families in Acordo de Lusaka, Memba District, in October 1981 No. of families with positive response Affected Control families families (n= 17) (n= 17) Ate "chima" (made from dried bitter cassava) a Ate dried, uncooked cassavaa Ate cereals Ate beans Ate relish (containing cassava leaves)0 Ate fish or meat Ate wild plants (in the last month) Member of the food Consumer Cooperative Depended on gifts of food 17 16 17 6 2 13 0 12 6 6 0 6 6 1 a Median estimated drying time of cassava was 7 days for both the affected families and the control families. a monetary contribution, and by the fact that these people depended on gifts of food from family and friends. Some affected families also ate fresh bitter cassava and augmented the quantity of flour with pounded, dried cassava peelings. This latter practice was introduced only in 1981. Most families ate twice a day but some, including children, sometimes ate only once a day. In Calawane and Muentaze, the dietary situation was similar, the only exceptions being one family that ate maize and two others that ate dried fish and beans, in Muentaze. Almost all the affected families recalled having acute symptoms 4-6 hours after ingestion of meals containing cassava from the 1981 harvest. However, at the time of the investigation the incidence of acute intoxication had decreased. Nutritional status Table 3 shows the results of the investigation of nu- tritional status of families in Acordo de Lusaka. There is little difference between the status of the affected families and the controls. Toxicology Table 4 shows the hydrocyanic acid (HCN) content of the cassava samples using the alkaline titration method. Comparison of this method with the direct 488 1 HYDROCYANIC ACID CONTENT OF CASSAVA Table 3. Nutritional status in Acordo de Lusaka, Memba District, in October 1981 Percentage No. of undernourished persons Parameter' or children Moderate Severe Affected families 65 W/H 49 12 Control families 37 W/H 38 11 Village children: 0-1 yearold 29 W/A 45 14 1-6 years old 121 W/H 32 2 a W/H = weight for height; W/A = weight for age. Table 4. Hydrocyanic acid (HCN) content of cassava samples from the affected area HCN (mg/kg) Type of No. of cassava sample samples Mean Standard Range value deviation Fresh leaves: bitter varieties 6 377 110 224-469 sweet varieties 14 347 178 71 -495 Fresh roots: bitter varieties 15 327 153 186-446 sweet varieties 20 138 55 54-235 Dried roots: bitter varieties 12 95 47 64- 178 sweet varieties 3 46 9 35-54 Flour 5 40 9 26-43 Cooked cassava 10 38 22 19-87 potentiometric method gave a maximum variation of 10%Vo. HCN values were highest for the fresh leaves, with no significant difference between bitter and sweet varieties; the mean HCN value was higher in fresh roots of the bitter than sweet varieties (P < 0.01) but there was considerable overlap. Progressively lower mean values were found in the dried samples, the reduction in cyanide content being 70% for the bitter varieties and 67% for the sweet varieties. The greatest range of variation in HCN values occurred in the fresh bitter roots. The relationship between the cyanide content, the length of drying time, and humidity showed that the cyanide levels appeared to decrease rapidly with dry- ing (within one week) and then remained stationary. Even with prolonged drying of two months and over, the HCN levels were still high: 85 mg/kg for bitter cassava and 35 mg/kg for sweet cassava. DISCUSSION Historical factors The problem encountered in 1981 had its real begin- nings in the colonial period when cassava was first introduced into the area. Virtual dependence on this one crop was created at the expense of subsistence production of cereals, legumes, etc. This, in turn, created a dependence on commercially available food which consequently generated the need to earn money through sale of cash crops or manual labour for wages. In recent years the commercially available food has diminished for several reasons such as a decline in cash crop production because of unfavour- able weather conditions, the abolition of forced cash crop production, and a breakdown in the commercial network. Owing to the shortage of cash and of food on sale, the dietary supply of protein-rich foods con- taining the sulfur necessary for metabolic detoxifi- cation of cyanide was probably reduced. Pressure of demand from the coastal and urban centres for food from the interior continued during the drought, thereby encouraging the rural popu- lation to sell the dried cassava and other food that they had. Agricultural potential The non-coastal areas most affected by the 1981 drought and outbreak of mantakassa had always had a limited agricultural potential and a relatively high population density, but this did not prevent them from exporting food to other areas. The coastal zone, in comparison, could not produce sufficient food and imported it from the interior, although fishing pro- vided a useful supplement. The coastal zone was least affected by the outbreak, possibly because of the con- tinued supply of fish which provided both the sulfur- containing amino acids needed for metabolic detoxifi- cation of cyanide, and the means for exchanging food with the interior. The urban populations, on the other hand, ate mainly dried cassava from the interior and were therefore prone to the risk of intoxication. Detoxification Unlike other cassava-producing areas where the cassava is left in the ground and harvested as re- quired, the crop in Nampula is harvested over a short period which is followed by extensive drying and storage periods. The practice in Nampula probably 489 MINISTRY OF HEAL I H, MOZAMBIQUE evolved both to fit in with the dates for cotton culti- vation during the year and to allow sufficient time for the cassava to dry out adequately, especially in the case of the bitter varieties. As sun-drying is the only method of detoxification used, the drying period is particularly important. In a review of traditional detoxification methods, Coursey (9) states that "sun-drying can only be ap- plied to cassava containing small amounts of glyco- side". It is therefore surprising that in an area where bitter cassava had for a long time been cultivated, other more efficient detoxification methods (such as soaking) were not practised. In other areas of Mozam- bique, however, methods such as soaking and grating are used (10). Soaking has been shown to reduce the glycoside content by about 9007o, while sun-drying and boiling were less effective (11). Also, a study among six ethnic groups in Zaire (12) discussed the importance of detoxification methods in relation to cyanide intake; a direct relationship was found be- tween the incidence of goitre attributed to a high cyanide intake linked with a low iodine supply, and the methods of preparation used. Soaking, again, gave more favourable results. Diet and nutritional status Before the drought, the diet in the affected area was probably well balanced in terms of energy and nu- trient requirements. The drought then caused a severe reduction and imbalance in the diet, which mostly affected the women of reproductive age and children who probably consumed the raw and dried uncooked cassava, which are higher in cyanogenic glycosides, in the course of collection and preparation. The children were affected because they often accompany women in their tasks. In contrast, the men normally ate well- cooked cassava. Lactating women who had a high incidence of the disease may also have eaten more raw cassava which was believed to stimulate lactation. Studies in a cassava-eating population in Zaire (13) have shown that the thiocyanate level in breast milk, although low, was not negligible but breast-fed babies did not present with the disease. The protein in the breast milk could have exerted a protective effect; in older children, however, the introduction of other foods possibly raised their cyanide intake. Vanderpas et al. (14) reported that when cassava juice was intro- duced as a first weaning food, the serum thiocyanate levels rose. Table 2 shows that a larger proportion of the affected families ate dried uncooked cassava which could have been an important factor in raising the cyanide intake. The similarity in nutritional status between the affected and control groups confirms the general lack of food in the area. Protein-rich foods, particularly, were insufficient and could have resulted in diminished metabolic detoxification. In spite of the general food shortage, the usual symptoms of ad- vanced undernutrition such as growth failure, oedema, wasted muscles, and apathy were rarely seen. However, many cases of kwashiorkor did appear in February 1982 after the rains failed again at the end of 1981. The high serum thiocyanate levels in both patients and controls must have been due to chronic cyanide intoxication. Factors in the diet such as high cyan- ogenic glycoside levels in the cassava plant and in- adequate detoxification of the cassava, as well as in- sufficient metabolic detoxification due to a dietary deficiency of sulfur-containing amino acids, may have triggered the disease. As none of the wild plants identified is known to be neurotoxic, it is unlikely that they played a role in the etiology of the disease. They did provide some food during the critical period, but the drought consider- ably reduced this source too. Toxic levels in cassava products The cyanide levels found in the cassava samples in this investigation are high when compared with those found by other authors (15, 16). In this study, 55%0 of the sweet fresh roots were extremely toxic and the re- mainder moderately so, which may have been due to the drought conditions increasing the cyanogenic glycoside content (17). The mean values were higher in the roots of the bitter varieties. In general, the dried samples had lower cyanide levels, but even with two months of drying some values were still high. This finding confirms the opinion of Coursey (9), mentioned above. To relate the values found in the samples to the amount of cyanide ingested, the dietary habits of the population must be considered. The highest values were found in fresh leaves but these are normally eaten cooked which probably detoxifies much of the cyanide. If it is assumed that 9007o of the diet was com- posed of cassava, then with a daily adult energy intake of 7.53 megajoules (MJ) (1800 kcallh) the quantity of cassava consumed would have been in the range of 500-700 g per day. Most of the dried bitter cassava samples had a cyanide content of 80-100 mg/kg, while for the cooked samples the range was 30-45 mg/kg. A calculation based on a diet containing dried bitter cassava gives an approximate daily intake of 15-31.5 mg HCN. These levels of intake may be com- pared with the adult lethal dose of cyanide of 50 mg and could explain the frequent acute reactions after meals. Over a period of time, the consumption of such quantities could result in chronic cyanide intoxi- cation, which was found in both patients and con- trols. Given the dependence of this population on cas- sava, and the proneness of the area to drought, people 490 HYDROCYANIC ACID CONTENT OF CASSAVA possibly consume sufficient cassava to result in chronic cyanide intoxication even under more normal conditions. They could therefore be at risk from other diseases reported in association with chronic cyanide intoxication such as chronic neurological disease (1), endemic goitre and cretinism (12), and diabetes (18). Cassava and economic development As one of the cheapest sources of food energy avail- able, cassava plays a very important role in countries such as Mozambique, where peasant populations are subjected to unreliable climatic conditions and need security in their food supplies; owing to its hardiness and natural resistance, it often provides this security. The solution to the problem of cyanide intoxication, however, is not the prohibition of cassava but a care- ful study of the agricultural situation so that other drought-resistant crops including cereals and leg- umes, and varieties of cassava with lower cyanogenic glycoside levels may be introduced. At the same time, efforts to promote economic and social development in the areas prone to such disasters offer the only long- term solution. CONCLUSIONS The following evidence links the chronic cyanide intoxication found in patients to cassava consump- tion: - raised levels of cyanogenic glycoside levels in the cassava plant, including sweet varieties, which were due to the drought; - higher levels than usual of cyanide in processed cassava products because of inadequate detoxifi- cation; - the almost total dependence on cassava in the diet during the drought period, and reduced meta- bolic detoxification of cyanide because of the lack of protein-rich foods. ACKNOWLEDGEMENTS The help of the following in this collective work is acknowledged: community leaders, administrative and health workers at the district level, the provincial health administration (especially Mr Ilidio Manjate and Mr Constantino Cuamba of the Nutrition Section and, in Maputo, Dr Jorge Cabral and Dr Ana Novoa of the National Directorate of Preventive Medicine), staff of the National Food and Water Hygiene Laboratory, and members of the Department of History at Eduardo Mondlane University. RESUME LE MANTAKASSA: EPIDEMIE DE PARAPARESIE SPASMODIQUE ASSOCIE A UNE INTOXICATION CYANUREE CHRONIQUE, DANS UNE REGION DU MOZAMBIQUE OU LE MANIOC EST L'ALIMENT DE BASE. 2. FACTEURS NUTRITIONNELS ET TENEUR EN ACIDE CYANHYDRIQUE DES PRODUITS DU MANIOC Une epidemie de paraparesie spasmodique ou mantakassa (de son nom local) a eclate en 1981 dans cinq districts d'une province septentrionale du Mozambique. C'est dans une region agricole assez pauvre et touchee par la secheresse que l'incidence la plus forte a e observee. Les principales victimes etaient les femmes et les enfants. Etant donne la rarete des denrees alimentaires, la population s'est presque exclusivement nourrie de manioc, de varietes ameres en particulier. Parmi les douze varietes de manioc cultivees dans la region, deux seulement sont ameres a proprement parler; mais par suite de la secheresse toutes les varietes se sont revelees plus ameres qu'a l'ordinaire. Le manioc se detoxique normalement par sechage au soleil et stockage pendant une duree qui va jusqu'a un an. Mais en 1981, a cause de la penurie alimentaire, le manioc a et recolte de bonne heure, concasse et sehe au soleil un ou deux jours seulement avant d'etre consomme. L'existence d'un lien possible entre la consommation de manioc et une intoxication cyanuree a et exploree dans les zones les plus touchees au moyen d'investigations portant sur la nutrition, la toxicologie et la botanique. Sur le plan de la nutrition, les familles touchees et les temoins ont et etudies sous l'angle des facteurs agricoles, economiques et sociaux et des habitudes alimentaires avant et pendant la secheresse, notamment la nature de la consommation jour- naliere. L'analyse d'echantillons de produits du manioc a revel, tant dans les produits seches que dans les produits frais, une teneur anormalement elevee en heteroside, qui libere par hydrolyse de l'acide cyanhydrique. Le taux de cyanure residuel dans les echantillons deshydrates atteignait respectivement 85 et 35 mg/kg dans les varietes ameres et douces. La ration de cyanure a et estimee a 15-31,5 mg par per- sonne et par jour. De plus, il s'est produit une penurie d'ali- ments riches en proteines contenant des acides amines soufres qui auraient pu renforcer la detoxification metabo- lique du cyanure. Les sympt6mes d'une intoxication aigue ont ete ressentis 4 a 6 heures apres les repas comportant du manioc de la recolte de 1981. Plus de 100 varietes de plantes sauvages consommees ont et recoltees et identifiees; aucune 491 492 MINISTRY OF HEALTH, MOZAMBIQUE n'a de proprietes neurotoxiques connues. Les familles eco- nomiquement defavorisees et qui ont consomme du manioc seche cru et mange moins de viande ou de poisson etaient relativement plus nombreuses parmi les familles atteintes que parmi les temoins. Les femmes et les enfants ont pro- bablement consomme davantage de manioc s&eh non cuit, car ce sont eux qui sont charges de la recolte et de la prepa- ration. Aucune difference importante n'a et notee entre les groupes atteints et les groupes temoins en ce qui concerne 1'etat nutritionnel; des signes de denutrition grave ont rarement e observes. Le lien entre l'intoxication cyanur6e chronique et la consommation de manioc ressort de diverses observations, parmi lesquelles il convient de citer l'absorp- tion anormalement elevee de cyanure pendant la secheresse, par suite de la teneur accrue de la plante en heteroside liberant de l'acide cyanhydrique, et le manque d'acides amines soufres dans le regime alimentaire. REFERENCES 1. OSUNTOKUN, B. 0. Cassava, diet, chronic cyanide intoxication and neuropathy in Nigerian Africans. World review of nutrition and dietetics, 36: 141-173 (1981). 2. HARTMAN, C. P. ET AL. A study on Lathyrus sativus. Indian Journal of nutrition and dietetics, 11: 178-191 (1974). 3. CARVALHO, M. DE [Traditional agriculture in Mozam- bique. The geographical distribution of the cultivations and their impact on the environment.] Lourenco Marques, Missao de Inquerito Agricola de Moqam- bique, 1969, pp. 1-67 (in Portuguese). 4. GOMES, P. J. & GRANDVAUX BARBOSA, L. A. [The vegetation.] In: [Sketch of an ecological andagricul- tural survey in Mozambique. ] Lourenco Marques, Centro de Investigacao Cientifica Algodoeira. Mem6- rias e Trabalhos, No. 23, 1955, pp. 67-224 (in Portu- guese). 5. SERRA, C. [Cassava: what it is and how it appeared in Africa and in Mozambique.] Tempo, (619): 51 (1982) (in Portuguese). 6. CENTRO DE ESTUDOS AFRICANOS. [The transformation of agriculture at the family level in the province of Nampula.] C.E.A. Relat6rio, No. 80/3, Maputo, Eduardo Mondlane University, 1980 (in Portuguese). 7. WATERLOW, J. C. Classification and definition of protein-energy malnutrition. In: Beaton, G. H. & Bengoa, J. M., ed., Nutrition in preventive medicine, Geneva, World Health Organization, 1976, pp. 530- 555. 8. HORWITZ, W. ed., Official methods of analysis of the Association of Official Chemists. Washington, Associ- ation of Official Analytical Chemists, 1980, p. 433. 9. COURSEY, D. G. Cassava as food: toxicology and tech- nology. In: Chronic cassava toxicity, Ottawa, Inter- national Development Research Centre, 1973, pp. 27-36. 10. MOTA, T. P. [The cassava flour of Mozambique.I Agronomia Mo!ambicana, 8: 47 (1974) (in Portuguese). 11. BOURDOUX, P. ET AL. Cassava products: HCN content and detoxification processes. In: Nutritional factors involved in the goitrogenic action of cassava, Ottawa, International Development Research Centre, 1982, pp. 51-58. 12. SIMONs-GtRARD ET AL. Foods consumed and endemic goitre in the Ubangi. In: The role of cassava in the etiology of endemic goitre and cretinism. Ottawa, International Development Research Centre, 1980, pp. 69-80. 13. DELANGE, P. ET AL. Effects of thiocyanate during preg- nancy and lactation on thyroid function in infants. In: Role of cassava in the etiology of endemic goitre and cretinism, Ottawa, International Development Re- search Centre, 1980, pp. 121-126. 14. VANDERPAS, J. ET AL. Breast feeding, thiocyanate metabolism and thyroid function in young infants in severe endemic goitre. In: Nutritional factors in- volved in the goitrogenic action of cassava, Ottawa, International Development Research Centre, 1982, pp. 59-64. 15. BOLHUIS, G. C. The toxicity of cassava roots. Journal of agricultural science, 2: 176-185 (1954). 16. OKE, 0. 1. The role of HCN in nutrition. World review of nutrition and dietetics, 11: 170-198 (1969). 17. DE BRUIJN, G. H. The cyanogenic character of cassava. In: Chronic cassava toxicity. Ottawa, International Development Research Centre, 1973, pp. 43-48. 18. MCMILLAN, D. E. & GEEVARGHESE, J. Dietary cyanide and tropical malnutrition diabetes. Diabetes care, 2: 202-208 (1979).