KONZO AND CASSAVA TOXICITY

A STUDY OF ASSOCIATED NUTRITIONAL FACTORS IN THE POPOKABAKA

DISTRICT, DEMOCRATIC REPUBLIC OF CONGO

Delphin DIASOLUA NGUDI

Academic Year 2004-2005

KONZO AND CASSAVA TOXICITY: A STUDY OF ASSOCIATED

NUTRITIONAL FACTORS IN THE POPOKABAKA DISTRICT,

DEMOCRATIC REPUBLIC OF CONGO

KONZO-ZIEKTE EN DE TOXICITEIT VAN MANIOK: EEN STUDIE VAN DE VOEDINGSFACTOREN IN POPOKABAKA DISTRICT, DEMOCRATISCHE REPUBLIEK KONGO

Door

Delphin DIASOLUA NGUDI, M.Sc.

Thesis submitted in fulfilment of the requirements for the degree

of Doctor (Ph.D.) in Applied Biological Sciences

Proefschrift voorgedragen tot het bekomen van de graad van

Doctor in the Toegepaste Biologische Wetenschappen

Op gezag van: Rector: Prof. Dr. Apr. A. DE LEENHEER

Decaan: Promotoren: Prof. Dr. Ir. Herman VAN LANGENHOVE Prof. Dr. Patrick KOLSTEREN Prof. Dr. Ir. Fernand LAMBEIN

CITATION

“A person who has food has many problems.

A person who has no food has only one problem”

Chinese saying

DEDICATION

To Thérèse Luntala and our sons Berdit, Gaël and Beni

In memory of my Grand Oncle André Banketa and my father Bernard Ngudi-a-nkama

Promotoren: Prof. Dr. Patrick KOLSTEREN Vakgroep Voedselveiligheid en Voedselkwaliteit, Universiteit Gent Voedings unit, Prins Leopold Instituut voor Tropische Geneeskunde, Antwerpen

Prof. Dr. Ir. Fernand LAMBEIN Instituut Planten Biotechnologie voor Ontwikkelingenlanden, Universiteit Gent

Decaan: Prof. Dr. Ir. Herman VAN LANGENHOVE

Academiejaar 2004 - 2005

Delphin DIASOLUA NGUDI

KONZO AND CASSAVA TOXICITY

A STUDY OF ASSOCIATED NUTRITIONAL FACTORS IN THE POPOKABAKA

DISTRICT, DEMOCRATIC REPUBLIC OF CONGO

Proefschrift

voorgedragen tot het bekomen van de graad van

Doctor in de Toegepaste Biologische Wetenschappen

Op gezag van de rector,

Prof. Dr. Apr. A. DE LEENHEER

Nederlandse vertaling titel:

KONZO-ZIEKTE EN DE TOXICITEIT VAN MANIOK: EEN STUDIE VAN DE

VOEDINGSFACTOREN IN POPOKABAKA DISTRICT, DEMOCRATISCHE

REPUBLIEK KONGO

Diasolua Ngudi, D. (2005). Konzo and cassava toxicity: a study of associated nutritional factors in the Popokabaka District, Democratic Republic of Congo. Ph D. thesis Universiteit Gent, Belgium, 160 p

ISBN 90-5989-073-6

The author and the promoter give the authorisation to consult and to copy parts of this work for personal use only. Every other use is subject to the copyright laws. Permission to reproduce any material contained in this work should be obtained from the author.

Acknowledgments

I would like to express my gratitude to all the many people who and institutions that have contributed to this research. I am particularly grateful to my promoter Prof Dr Patrick Kolsteren for his encouragement, his guidance and constructive criticisms. My special thanks to Prof Dr Ir Fernand Lambein, the co-promoter of this thesis for accepting me to work under his supervision. I am indebted to him for improving my skills, including my laboratory techniques and for sharing his knowledge and experience on nutritional and neuro- toxicological disease. I also thank Prof. Em. Dr. Ir André Huyghebaert, my promoter until his retirement for handling the administrative issues of my enrolment to the university. I thank Dr J. Howard Bradbury from the Australian University and chairperson of the Cyanide Cassava Disease Network for his advices and for graciously providing me kits for analyses of cyanide and thiocyanate. I thank Dr Yu-Haey Kuo (Dianna) for the initiation to the use of HPLC instruments and for her meticulous attention in scrutiny of manuscripts. I thank Prof Dr Thorkild Tylleskär for inspiring and encouraging me to initiate this research. Dr Theophile Ntambwe, former Director of PRONANUT has to be thanked for his support and follow up of my scholarship file. I am thankful to Prof Dr JP Banea Mayambu, Director of PRONANUT for the encouragement, guidance and support. I am indebted to both of them for providing me field facilities. The staff and colleagues of PRONANUT are greatly acknowledged for their enthusiasm and encouragement. Prof Dr Simon Malele ma Ludani from the Australia’s University of Southern Queensland/ Dubai and Dr Fabienne Ladrière from Médecins du Monde are thanked for their assistance and for reading and improving this thesis. I am deeply grateful to the members of the examination committee: Prof Dr Ir Georges Hofman, Prof Dr Ir Colin Janssen, Prof Dr Geert Callewaert, Prof Dr Armand Christophe and Dr Ir Bruno De Meulenaer. Their criticism of this work was very constructive. The Belgian Cooperation and Development Ministry has exceptionally granted me a mixed scholarship through the Belgian Technical Cooperation (BTC) in April 2002. I thank Mr Marino Orban, Mr G. Kasende and Mrs Sarah Stijnen of BTC for their efficient management of my dossier. Congolese Ministry of Health through PRONANUT, OXFAM – Destelbergen (Katrien Goddemaer), GTZ (Nour Salua) and anonymous relatives and friends provided also financial support to conduct part of my program. Prof Em Marc Van Montagu, Prof Ann Depicker, Prof Dr Lieve Gheysen and staff of IPBO are thanked for the hospitality and their interest in my research.

Konzo and cassava toxicity II

Exceptional thanks to Ir. Anne-Marie De Winter, Regine Haspeslagh, Veerle Van Ongeval, Ir. Ann Peters, Christine Graveel, Jean-Pierre Dubois, Ing. Alfons Lenaert, Myriam De Vos and Ing. Yves De Jonge for their assistance, continuous interest and encouragement. I am indebted to the Visie newspaper and to the VRT-film crew Jo Frere and Jasmine De Bruycker for the interest in the topic of my research and for their objective reporting on the disease on the disease konzo and on our research. Piet Meyvaert, Paul Meirsman, Lieve Van Wijmeersch, Fernand Verhoeven, Erik Verhaegen, Willy Mpoyi, Louis Kitenge, Raf Nunga, Elie and Blaise Ndosimau, Sera and Emmanuel Kisuesue, Ir Bernard Lelou, Prof Dr Emmanuel Biey, Dr José Biey, Dr Thomas Mpiana, Dr Clément Mulenda Tshamala, Michel Fazili and other friends, not mentioned by name, are acknowledeged for their help and friendship. I would like to express my cheerfully thanks to my mother, to my sisters Lili, Micheline and Euphrasie,to my sister in law Makiese and to my brothers in law Malueki, Nsinga, Ntondo and Yende for taking care of our sons while we were out of the country and far from them. Many thanks to my uncle Prof Dr Mamingi Nlandu, to my brothers Domi, Sivis and Nsimba, to my cousin Dimbu, to my sisters in law Annie, Julie and Elisée and to all my nephews and nieces. To my sons Berdit, Gäel and Beni, I have to apologise for not giving them all the attention they deserved during the many busy years that research entailed. I thank Thérèse Luntala for often coping with my absence concerning things of life and for her advice, help and love. Last but not the least; I thank the authorities and the population of Popokabaka for their enthousiasm and collaboration during my field trips.

Gent, June 2005

Delphin Diasolua Ngudi

Konzo and cassava toxicity III

CONTENTS

Acknowledgments...... I

CONTENTS...... III

List of figures ...... V

List of tables...... VI

Abbreviations and acronyms...... VIII

Samenvatting ...... X

Résumé ...... XIII

Summary...... XVII

I Literature review...... 2

I.1 Introduction...... 2 I.1.1 Paraparesis and neurodegenerative diseases, what is the meaning? ...... 2 I.1.2 The “hidden endemias” ...... 3

I.2 Konzo...... 7 I.2.1 Background information on konzo...... 7 I.2.2 Clinical features and differential diagnosis...... 10 I.2.3 Epidemiology ...... 14 I.2.4 Infection or toxico-nutritional etiology? ...... 15

I.3 Dietary exposure to cyanide from cassava ...... 17 I.3.1 Cassava...... 17 I.3.2 Cyanide toxicity...... 27

I.4 Conclusion...... 39 I.4.1 Rationale of the research...... 40 I.4.2 Objectives...... 40

II Occurrence of konzo and dietary pattern...... 43

II.1.1 Introduction ...... 43

Konzo and cassava toxicity IV

II.1.2 Materials and Methods...... 43 II.1.3 Results ...... 47 II.1.4 Discussion and conclusion ...... 54

III Cassava food quality and safety...... 59

III.1 Food Safety and Amino Acid Balance in Processed Cassava "cossettes" ...... 59 III.1.1 Introduction ...... 59 III.1.2 Materials and methods ...... 62 III.1.3 Results and discussion...... 66

III.2 Residual cyanogens, free and total amino acid profiles of cooked cassava leaves "saka- saka”...... 79 III.2.1 Introduction ...... 79 III.2.2 Materials and methods ...... 80 III.2.3 Results and discussion...... 82 III.2.4 Conclusions ...... 92

IV Dietary cyanogen and sulphur metabolites excretion...... 95

IV.1.1 Introduction ...... 95 IV.1.2 Material and methods...... 96 IV.1.3 Results ...... 98 IV.1.4 Discussion ...... 100

V General discussion and conclusions...... 105

V.1 Occurrence of konzo...... 106

V.2 Cassava foods and sulphur metabolites...... 108

V.3 Conclusions and recommendations...... 111

References ...... 114

Curriculum Vitae ...... ii

Konzo and cassava toxicity V

List of figures

Figure I-1: Health zones of Kwango District in Bandundu Province, DRC...... 8

Figure I-2: From left to right: mild form, moderate form and severe form of konzo in young

subjects...... 12

Figure I-3 : Cassava roots harvested and cassava plant in the field...... 18

Figure I-4: Location of cassava production, 1996 (Scott et al, 2000)...... 23

Figure I-5: Summary of traditional cassava processing in Africa (from Banea-Mayambu,

1997c)...... 27

Figure I-6: Cyanogenesis from linamarin (McMahon et al, 1995)...... 29

Figure I-7: Basic processes involved in the metabolism of cyanide (ATSDR, 1997)...... 35

Figure I-8 : Cysteine catabolism...... 37

Figure II-1: Distribution of onset of konzo from 1980 to 2002...... 49

Figure III-1: Flow diagram of cassava cossettes processing ...... 60

Figure III-2: Free amino acids in cassava cossette samples...... 74

Figure III-3: Protein amino acids profile of the raw and cooked cassava leaves...... 87

Konzo and cassava toxicity VI

List of tables

Table I-1: Characteristic features of four tropical myeloneuropathies (Tylleskär et al, 1994c)5

Table I-2: WHO criteria for konzo versus newly suggested criteria (with permission and from

Tshala-Katumbay, 2001a) ...... 11

Table II-1: Socio-demographic variables and 24hr recall food consumption of participants

among the high prevalence of konzo health area (n = 224) and the low prevalence of

konzo health area (n =263)...... 46

Table II-2: Distribution of konzo cases per health area...... 49

Table II-3 : Degree of disability on walking and age distribution of konzo patients by gender

...... 50

Table II-4: 24-hour recall of household food intake frequencies (%) ...... 51

Table II-5: Seasonal food consumption availability (%) listed by the respondents ...... 53

Table III-1: Cyanogens content in cassava cossettes (mg HCN equivalent kg - 1 dry weight)67

Table III-2: Estimated daily cossettes and total cyanogens intake...... 69

Table III-3: Total protein amino acids content in cassava cossettes (mg g - 1dry weight)...... 72

Table III-4: Amino acid scoring pattern of different cossette samples ...... 73

Table III-5. Free protein amino acids content in cassava cossettes (mg g - 1 dry weight)...... 75

Table III-6. Essential Amino Acid (EAA) requirements and estimated daily intake ...... 77

Table III-7: Cyanogen content in raw and cooked cassava leaves (mg HCN equivalent kg-1

dry weight) ...... 84

Table III-8: Protein content and amino acid composition of raw and cooked pounded cassava

leaves (g kg-1 dry weight) ...... 86

Table III-9: Comparison of the essential amino acid contents of different raw and cooked

pounded cassava leaves samples and their amino acid score with the recommended FAO

reference...... 89

Konzo and cassava toxicity VII

Table III-10: Free amino acid and trigonelline content in raw and cooked cassava leaves (g

kg-1 dry weight) ...... 91

Table IV-1: Distribution of konzo- affected households in each health area with the number

of konzo patients given in brackets...... 98

Table IV-2: Total cyanogens in cassava flour, thiocyanate and taurine in urine samples

collected in three konzo prevalence areas of Popokabaka (DRC)...... 99

Konzo and cassava toxicity VIII

Abbreviations and acronyms

α-ABA α-Amino butyric acid AMPA α-Amino-3hydroxy-5-methyl-isoxazole-4-propionic acid ATC 2-Aminothiazoline-4-carboxylic acid ATSDR Agency for Toxic Substances and Disease Registry BOAA ß-Oxalylaminoalanine ß-ODAP ß-N-oxalyl-α,ß-diaminopropionic acid BTC/CTB Belgian Technical Cooperation/ Coopération Technique Belge CEPLANUT Centre National de Planification de Nutrition Humaine CI confidence interval CN- Cyanide CNS Central nervous system D. R. C. Democratic Republic of Congo EAA Essential amino acids FAO Food and Agriculture Organisation GABA γ-Amino butyric acid HCN Hydrogen cyanide HNL α-Hydroxynitrile lyase HPLC High Performance Liquid Chromatography HTLV-1 Human T cell lymphocyte virus Type I IITA International Institute for Tropical Agriculture KCN Potassium cyanide NaCN Sodium cyanide OCN- Cyanate OR Odds Ratio PITC Phenylisothiocyanate Ppm Part per million Prhz Popokabaka rural health zone PRONANUT Programme National de Nutrition R. D. C. République Démocratique du Congo SAA Sulphur containing amino acids SCN- Thiocyanate SD Standard deviation

Konzo and cassava toxicity IX

SPSS Statistical Package for Social Science TAN Tropical ataxic neuropathy TSP/HAM Tropical spastic paraparesis/ Human T cell lymphocyte virus Type I-associated myelopathy UDPG Uridine diphosphoglucose UK United Kingdom UNICEF United Nations for Infants and Children Emergency Funds UNU United Nations University USA United States of America WHO World Health Organisation

Konzo and cassava toxicity X

Samenvatting

Konzo is een neurologische aandoening die gekenmerkt wordt door een plots beginnende en blijvende verlamming van de benen. Dit komt vooral voor bij kinderen en vrouwen op vruchtbare leeftijd. Wortels van bittere maniok (Manihot esculenta Crantz) en daarvan afgeleide producten die een hoog gehalte aan cyanogenen bevatten, kunnen acute cyanide vergiftiging veroorzaken met symptomen als braken, duizeligheid, maagpijn, flauwte, hoofdpijn en diarree. Konzo werd toegeschreven aan een langdurige chronische inname van cyanide in bittere maniokwortels die onvoldoende geroot werden. Maniok is een belangrijke basis voor het dagelijkse voedsel van meer dan een half miljard mensen verspreid over de ganse wereld, nochtans bevat het cyanogeen glycosiden, voornamelijk linamarine dat na enzymatische omzetting tot cyanohydrine aanleiding kan geven tot het giftige blauwzuur

(HCN) na verdere enzymatische of spontane omzetting. Alhoewel er voldoende aanwijzingen zijn voor een verband tussen de ziekte konzo en de consumptie van bittere cassava, blijft het pathologisch mechanisme van de ziekte onduidelijk. Men vermoedt dat zowel HCN als zijn metabolieten (2-aminothiazolin-4-carbonzuur, cyanaat en isothiocyanaat) een rol spelen in de pathologie maar er is geen proefdier model om dit te bevestigen. Wel staat vast dat de enzymatische omzetting van cyanide naar thiocyanaat (80% van het cyanide wordt langs die weg gedetoxifieerd) gebruik maakt van zwavel afkomstig van de zwavelhoudende aminozuren methionine en cysteine.

In dit werk wordt het voorkomen van de konzo ziekte in Popokabaka, in de Bandundu provincie van de Democratische Republiek Kongo bestudeerd. Uit diezelfde regio kwam het eerste gepubliceerde verslag over konzo in 1938. Mogelijke associaties tussen de ziekte met gezinsgebonden factoren en met het eetpatroon van de bevolking die hoofdzakelijk maniok eet, werden onderzocht. Konzo komt nog steeds voor in die regio met een incidentie van 1,3

‰ in 2002. Meer vrouwen dan mannen waren aangetast en er werd geen informatie over de

Konzo and cassava toxicity XI levensverwachting van de patiënten gevonden. Zowel geslacht als burgerlijke stand vertoonden een statistische associatie met de kans op konzo. Maniok was het dominante bestanddeel van de voeding in praktisch elk gezin en werd minstens eenmaal per dag gegeten als luku, een dikke pasta van maniokmeel en water. Als bijgerechten werd voornamelijk maniok bladeren en lokaal geteelde bonen gegeten. Alhoewel die bijgerechten rijker zijn aan eiwit is de kwaliteit van dit eiwit laag door een gebrek aan zwavelhoudende aminozuren.

De ‘cossettes’, de gerote en gedroogde maniok wortels die het voornaamste voedingsbestanddeel vormen en de maniokbladeren die het voornaamste bijgerecht vormen in de regio werden onderzocht. De dagelijkse hoeveelheid werd bepaald, het gehalte aan en de dagelijkse inname van cyanogeen en de hoeveelheid zwavelhoudende aminozuren vereist voor de detoxificatie ervan werd berekend. De vrije en totale aminozuren in de maniok produkten werden bepaald om eventuele inherente toxinen op te sporen, om de kwaliteit van de proteïne te bepalen en om te kunnen vergelijken met de dagelijkse vereisten voor kinderen en volwassenen. Er werd berekend dat kinderen van 1 tot 9 jaar dagelijks 0,4 tot 1,1 mg HCN equivalenten innemen in 241 tot 389 g maniok produkten, hoofdzakelijk van ‘cossettes’.

Matig actieve volwassen mannen en vrouwen namen 0,6 tot 1,5 mg HCN equivalenten in per dag van 390 tot 532 g maniok produkten bereid uit ‘cossettes ‘, in de veronderstelling dat

60% van de dagelijkse energievereisten voldaan worden door maniokwortels die 1,6 tot 2,8 mg HCN equivalenten bevatten per kg droog gewicht. We vonden geen potentieel giftige niet- proteïne aminozuren in maniokwortels. Lysine en leucine zijn de limiterende aminozuren terwijl methionine slechts voor 13% bijdraagt in het lage gehalte van zwavelhoudende aminozuren van de ‘cossettes’. Er werd berekend dat, indien alleen maniok gegeten wordt, de dagelijkse behoefte aan methionine voor kinderen van 1 tot 9 jaar slechts voor 60% voldaan is, terwijl aan de dagelijkse behoefte van de volwassenen wel voldaan wordt indien 60% van de dagelijkse energiebehoeften komt uit maniokprodukten (het nationale gemiddelde). De

Konzo and cassava toxicity XII maniokbladeren die wel rijker zijn aan proteïne maar eveneens arm zijn aan methionine, kunnen dit gebrek bij kinderen in gebieden met konzo niet compenseren. De maniokbladeren als bijgerecht kunnen ook een bijkomende bron van cyanide zijn indien dit onvoldoende lang gekookt wordt wegens een gebrek aan brandhout, de enige beschikbare brandstof voor de bereiding. Zwavelhoudende aminozuren zijn noodzakelijk voor de detoxifiëring van cyaniden afkomstig van maniokwortels die onvoldoende geroot werden, maar ook van de bladeren die onvoldoende lang gekookt werden.

In een epidemiologische studie werden urinestalen en stalen van de maniokbloem van verschillende gezinnen onderzocht met de bedoeling de voedselveiligheid na te gaan, de inname van cyanogenen te berekenen en een mogelijk verband na te gaan met de zwavelmetabolieten taurine en thiocyanaat in de urine. Drie generaties na het eerste gedocumenteerd voorkomen van konzo in deze regio blijft er nog steeds een hoog risico voor de blootstelling aan cyanide van maniokbloem. In de helft van de gezinnen is het cyanidegehalte in maniokbloem hoger dan de door de Wereld Gezondheid Organisatie en de

FAO aanbevolen grens van 10µg HCN equivalent/g bloem. De helft van de urinestalen bevatten meer dan 300 µmol/l thiocyanaat. Dit wijst op een grote overmaat aan cyanide inname Alhoewel men geen significante correlatie vond, wijst de lage concentratie taurine, de eindmetaboliet van de zwavelhoudende aminozuren, op een hoog verbruik van zwavel voor de detoxifiëring van cyanide met vorming van thiocyanaat.

De bevolking van Popokabaka is nog steeds blootgesteld aan een te hoog cyanogeen gehalte uit de maniokvoeding en misschien ook aan cyanogenen uit de omgeving. Dit verhoogt het risico voor konzo in die regio waar die verlammende ziekte nog steeds voorkomt. Een beter gevarieerd en gebalanceerd dieet, dat rijker is aan methionine, is noodzakelijk voor een meer efficiënte detoxifiëring van cyanide in het menselijke lichaam.

Konzo and cassava toxicity XIII

Résumé

Konzo est un désordre neurologique caractérisé par un début soudain d’une paralysie permanente des membres inférieurs. Les enfants d’au moins trois ans et les femmes en âge de procréer sont les plus affectés. La consommation des tubercules de manioc et de leurs produits dérivés contenant une grande quantité de cyanogènes peut causer une intoxication pouvant se manifester par des vomissements, des nausées, des vertiges, des douleurs abdominales, la faiblesse, des maux de tête et la diarrhée. Konzo a été attribué à une haute consommation de cyanure des tubercules de manioc amer insuffisamment traité. Le manioc, aliment de base très important pour plus d’un demi milliard d’habitants de la planète, contient des glucosides cyanogéniques, principalement la linamarine qui après conversion enzymatique en cyanohydrines, peut libèrer spontanément ou enzymatiquement un prodruit toxique, l’acide cyanhydrique. Bien que le lien évident entre la maladie konzo et la consommation du manioc amer insuffisammnet traité soit établi, les mécanismes pathogéniques du konzo restent encore à élucider. L’acide cyanhydrique et ses métabolites

(acide 2-aminothiazoline-4-carboxylique, cyanate, thiocyanate) ont été suspectés de jouer un rôle dans la pathogénicité du konzo chez l’être humain mais il n’y a pas encore de modèle animal pour s’en assurer ou le confirmer. En effet, la conversion enzymatique du thiocyanate

à partir du cyanure (environ 80 % de cyanure est transformé par cette voie) nécessite le soufre provenant des acides aminés soufrés, la méthionine et la cystéine.

Notre travail rapporte l’apparition des cas de konzo à Popokabaka, province de Bandundu, R.

D. Congo, une des régions incluse dans la première publication sur le konzo en 1938. Nous décrivons l’association entre le konzo et les facteurs socio-économiques liés au ménage et les habitudes alimentaires des populations consommant le manioc. Le konzo continue à sévir dans cette région de Popokabaka avec une incidence de 1,3 ‰ en 2002. La paralysie affecte plus de femmes que d’hommes et nous n’avons trouvé aucune publication sur l’espérance de

Konzo and cassava toxicity XIV vie (ou sur la mortalité) des personnes atteintes de konzo. Le genre et l’état civil des chefs de ménage sont associés au degré de prévalence de konzo dans les différentes localités de la région. L’alimentation dans la contrée est largement dominée par le manioc. Presque tous les ménages consomment au moins une fois par jour du luku, pâte obtenue après cuisson et malaxage de la farine de manioc dans de l’eau bouillante. Les principaux aliments d’accompagnement du luku, les feuilles de manioc et le niébé, sont limités en acides aminés soufrés.

Nous avons analysé les cossettes de manioc (produit dérivé des tubercules de manioc roui et séché) et les feuilles de manioc, respectivement aliment de base et principal aliment d’accompagnement dans la région, pour estimer la quantité de cyanogènes résiduels consommée journellement et la quantité d’acides aminés soufrés disponibles pour la détoxification de ces cyanogènes. Nous avons déterminé les acides aminés libres et totaux dans les cossettes et les feuilles de manioc pour détecter la présence inhérente des acides aminés non protéiniques potentiellement toxiques, pour évaluer la qualité de leur protéine alimentaire et la comparer aux besoins recommandés en acides aminés des enfants et des adultes. Nous avons trouvé une consommation journalière estimée de 0,4 à 1,1 mg HCN

équivalent dans 241 à 389 g de luku et de 0,6 à 1,5 mg HCN équivalent dans 390 à 532 g de luku respectivement chez les enfants âgés de 1 à 9 ans et chez les adultes (masculin et féminin) avec une activité modérée lorsque 60 % de l’énergie alimentaire journalière requise provient du manioc de 1,6 et de 2,8 mg HCN équivalent par kg de produit sec. Aucun acide aminé non protéinique potentiellement toxique n’a été détecté dans les produits dérivés du manioc. La lysine et la leucine sont des acides aminés limitants et le contenu en méthionine est très bas contribuant pour environ 13 % du total des acides aminés soufrés dans les cossettes de manioc. Les enfants âgés de 1 à 9 ans ne peuvent pas s’attendre à satisfaire leurs besoins recommandés en méthionine tandis que les adultes peuvent les satisfaire à partir de la

Konzo and cassava toxicity XV quantité requise calculée pour satisfaire 60 % de l’énergie alimentaire journalière mais pas assez pour la détoxication de cyanure et le métabolisme normal. Les feuilles de manioc, quantitativement riches en protéines mais limitées en acides aminés soufrés, ne peuvent pas compenser la déficience alimentaire en acides aminés soufrés occasionnée par l’aliment de base dans les régions affectées par le konzo. Nous concluons aussi que les feuilles de manioc peuvent être une autre source non négligeable de cyanogène alimentaire dans cette région. En effet les feuilles de manioc requièrent pour leu détoxication une longue cuisson et avec, le manque d’électricité et de gaz, la rareté de bois de chauffage pour la préparation des aliments, il y a risque d’écourter le temps de cuisson et par conséquent de consommer des aliments insuffisamments cuits.

Les acides aminés soufrés sont essentiels pour la détoxication des cyanogènes résiduels contenus dans le manioc (tubercules ou feuilles) insuffisamment traité ou cuit.

Les échantillons de manioc prélevés auprès des ménages ainsi que ceux des urines obtenus des participants sélectionnés au hasard pour une étude épidémiologique que nous avions effectuée ont été examinés pour évaluer l’innocuité de la farine de manioc prête à la cuisson puis à la consommation, pour déterminer la charge en cyanogène et apprécier la relation potentielle entre les métabolites soufrés urinaires, la taurine et le thiocyanate. Il y a un risque

élevé d’exposition alimentaire aux cyanogènes dû à la consommation du manioc insuffisamment traité dans cette région où le konzo est rapporté depuis trois générations. La farine de manioc prélevée dans plus de la moitié des ménages contenait un taux de cyanogène au dessus du seuil de 10 ppm fixé par la FAO et l’OMS. Les urines de plus de la moitié des participants contenaient plus de 300 umol de thiocyanate par litre. Ce qui suggère qu’il y a une charge importante en cyanure. Par ailleurs, ces urines accusaient aussi une concentration très basse en taurine (produit du métabolisme des acides aminés soufrés), suggérant que le soufre est préférentiellement orienté dans la détoxication du cyanure par la formation du

Konzo and cassava toxicity XVI thiocyanate, bien que nous n‘avons presque pas trouvé de corrélation entre la taurine urinaire et le thiocyanate.

La population de Popokabaka reste toujours grandement exposé aux cyanogènes alimentaires du manioc et peut être aussi aux cyanogènes environnementaux. La prévention de risque élevé d’appariton de cas de konzo dans la région requiert une alimentation suffisante, variée et

équilibrée particulièrement riche en méthionine pour permettre une détoxication effective du cyanure par l’organisme en laissant assez d’acides aminés soufrés pour le reste des besoins métaboliques de l’organisme.

Konzo and cassava toxicity XVII

Summary

Konzo is a neurological disorder characterised by sudden onset of paralysis of the legs, which occurs particularly in children and women of childbearing age. Consumption of cassava

(Manihot esculenta Crantz) and its products that contain large amounts of cyanogens may cause acute cyanide poisoning with symptoms of vomiting, nausea, dizziness, stomach pains, weakness, headache and diarrhoea. Konzo has been attributed to the high dietary cyanide exposure from insufficiently processed roots of bitter cassava. Cassava which is an important staple food for more than half a billion inhabitants worldwide contains cyanogenic glycosides, mainly linamarin that after enzymatic conversion to cyanohydrins, may release spontaneously or enzymatically the toxic hydrogen cyanide (HCN). Although evidence linking the disease with consumption of bitter cassava has been established, the pathogenic mechanism of konzo remains unclear. HCN and its metabolites (2-aminothiazoline-4-carboxylic acid, cyanate and thiocyanate) have been suspected to play a role in the pathogenicity in humans but there is no animal model to ascertain or to confirm this. The enzymatic conversion of cyanide into thiocyanate (about 80 % of cyanide is transformed by this route) requires sulphur arising from the sulfur containing amino acids (SAA) methionine and cysteine.

In this work, we reporte the occurrence of konzo disease in Popokabaka, Province of

Bandundu, D. R. Congo, one of the areas included in the first published report on konzo in

1938. We described associated household factors involved in the disease and the dietary pattern of the cassava consuming populations. Konzo is still occurring in this area with an incidence rate of 1.3‰ in 2002. The disease affected a larger proportion of females than males but we found no reports on the life expectancy of konzo patients. Gender and marital status of the heads of household were associated with the degree of prevalence of konzo. The diet was largely dominated by cassava and almost all households consumed at least once daily the luku, a stiff porridge from the cassava flour. Major foods such as cassava leaves and

Konzo and cassava toxicity XVIII cowpeas consumed as side–dishes to the staple food luku are of poor quality in protein especially in SAA.

We analysed processed cassava roots ‘cossettes’, as the major staple food and cassava leaves, as the major side-dish to the staple food in the region to estimate the quantity of daily intake of cyanogen and for calculate the amount of SAA required for its detoxification. We determined free and total amino acids in the cassava products to investigate the presence of potentially toxic inherent nonprotein amino acids, to evaluate the dietary protein quality and to compare with the amino acid requirements of children and adults. We estimated that children (1-9yr) consumed daily about 0.4 to 1.1 mg of HCN equivalent in 241 to 389 g of cassava product from the ‘cossettes’ and moderately active female or male consumed 0.6 to

1.6 mg of HCN equivalent in 390 to 532 g cassava product from the ‘cossettes’ when 60% of the daily energy requirement is provided by cassava roots containing between 1.6 and 2.8 mg

HCN equivalent per kg dry weight. No potentially toxic nonprotein amino acids were detected in cassava products. Lysine and Leucine were the limiting amino acids and the methionine content was very low and contributed only about 13 % of the total SAA in the ‘cossettes’. We found that children of 1 to 9 years old cannot expect to meet methionine requirement whereas adults can meet SAA requirement from the calculated quantity required to satisfy 60 % of the daily energy from the staple food. Cassava leaves that were found to be quantitatively rich in protein but this protein is of poor quality with SAA as the most limiting amino acids, cannot compensate for the dietary deficiency in SAA in the staple food in konzo affected areas. We concluded also that cassava leaves could be an additional source of dietary cyanogen in the region. The leaves require prolonged cooking and with the unavailability of electricity or gas and scarcity of firewood they are consumed after a short cooking time.

SAA are essential for detoxification of the residual cyanogens in the insufficiently processed cassava roots and also in the improperly cooked cassava leaves.

Konzo and cassava toxicity XIX

In an epidemiological study, samples of cassava flour from households and samples of urine obtained from selected participants were examined to monitor the safety of the flour intended to be consumed, to check cyanogen overload and to assess a potential relation between urinary sulphur metabolites taurine and thiocyanate. There is a high risk of dietary cyanogen exposure from cassava flour in this region where konzo was first reported three generations ago. Cassava flour from more than half of the households had total cyanogen content above the WHO/FAO recommended safe limit (10 μg HCN equivalent/kg cassava flour). The urine samples from half of the participants contained more than 300 μmol/l of thiocyanate. This suggested a high cyanide overload. The low concentration in urinary taurine found suggested that more sulphur is directed to the detoxification of cyanide by formation of thiocyanate, although urinary taurine and thiocyanate were slightly or not correlated.

The populations of Popokabaka are still highly exposed to cyanogen dietary from cassava and perhaps to environmental cyanogens. The increased risk of konzo in this region where the paralytic disease is still occurring requires a more efficient post harvest processing and a better balanced diet, particularly richer in methionine, to allow efficient detoxification of cyanide in the body.

Konzo and cassava toxicity 1

CHAPTER I:

LITERATURE REVIEW

Konzo and cassava toxicity 2

I Literature review

I.1 Introduction

I.1.1 Paraparesis and neurodegenerative diseases, what is the

meaning?

Paraparesis is a common form of neurological disability in developing countries (Howlett,

1994). It is a slight paralysis or weakness of both legs resulting in mild to moderate loss of bilateral lower extremity motor function, which may be a manifestation of spinal cord diseases; peripheral nervous system diseases; muscular diseases; intracranial hypertension; parasagittal brain lesions; and other conditions. Paraparesis often progresses to paraplegia, paralysis of the legs and lower part of the trunk. Symptoms are mild and may include spastic paraparesis of the lower limbs, ataxia hypertonia (excessive muscle tone), mild peripheral neuropathy, and problems of urinary incontinence (Parker, 2004). Leprosy, poliomyelitis, tuberculosis and trauma are the main causes, but a heterogeneous group of diseases with paraparesis also exists whose occurrence is limited to the tropics and whose etiology is still unknown (Howlett, 1994).

In the 19th century, neurologists recognised that the muscle weakness could be due to primary disorders of muscle or secondary to loss of neuromuscular integrity, as it happens when peripheral nerves are cut or when motor neurones degenerate. Furthermore, it was observed that there are forms of motor neurone degeneration which selectively affect upper motor neurone (e.g.: primary lateral sclerosis, hereditary spastic paraplegia, tropical spastic paraparesis (TSP), neurolathyrism, konzo) or lower motor neurone (e.g.: spinal muscular atrophies, hereditary bulbar palsy), or combination of upper and lower motor neurone (e.g.: amyotrophic lateral sclerosis) (Donaghy, 1999; Swash et al, 1999; Talbot, 2002).

Konzo and cassava toxicity 3

Neurodegeneration corresponds to any pathological condition primarily affecting neurons. In practice, neurodegenerative disorders represent a large group of neurological disorders with heterogeneous clinical and pathological expressions affecting specific subsets of neurons in specific functional anatomic systems; they arise for unknown reasons and progress in a relentless manner (Przedborski et al, 2003). A number of mechanisms appear to contribute to the neurodegenerative process, including alterations in calcium homeostasis in the endoplasmic reticulum which contribute to neuronal excitotoxicity and apoptosis, and unregulated calpain (cysteine endopeptidase; EC 3.4.22.17) proteolysis, initiated by the dysregulation of calcium ion homeostasis. Mitochondrial disfunction may also be linked to neurodegenerative disease through free radical generation, impaired calcium buffering and the mitochondrial permeability transition. Apoptotic and necrotic cell death are both observed in neurodegenerative diseases. Another mechanism may be the disorganization of the cytoskeleton leading to neuronal degeneration (Sigma-RBI®, 2001).

Motor neurone disease is a term introduced by Brain in 1962, intending to unify under one umbrella various idiopathic degenerative motor system diseases (Swash et al, 1999). A large number of diseases of diverse aetiology may selectively affect the motor neurons in the central nervous system (CNS) and, among the hundreds of different neurodegenerative disorders. So far most attention has been given to only a handful of diseases including

Alzheimer disease, Parkinson disease, Huntington disease, and amyotrophic lateral sclerosis

(Tshala-Katumbay, 2001a; Talbot, 2002; Przedborski et al, 2003). Many of the less common or less publicized neurodegenerative disorders, though not less devastating, have remained essentially ignored or neglected (Przedborski et al, 2003).

I.1.2 The “hidden endemias”

Tropical myeloneuropathies (Table I-1) is a term proposed by Román et al (1985) for the

“hidden endemias” of the neurodegenerative diseases that predominantly affect the spinal

Konzo and cassava toxicity 4 cord and peripheral nerves with poorly known aetiology such as Tropical Spastic

Paraparesis/Human T cell Lymphocyte Virus Type I-Associated Myelopathy (TSP/HAM), tropical ataxic neuropathy (TAN), neurolathyrism and konzo, diseases in which a virus or a natural toxin causes selective upper motor neurone impairment (Tylleskär, 1994c; Verma and

Bradley, 2001). Upper motor neurone signs are the result of an interruption in the neural pathway above the anterior horn cell. Characteristic of an upper motor neurone disease are:

• Weakness – the extensors are weaker than the flexors in the arms, but the reverse is

true in the legs with weakness more pronounced in flexor muscles.

• Muscle wasting is absent or slight – muscle wasting is prominent in a lower motor

neurone lesion

• Hyper-reflexia and clonus in upper motor neurone disorder – reflexes are absent or

reduced in a lower motor neurone lesion

• Spasticity

• No fasciculations in upper motor neurone disorder – fasciculations occur in a lower

motor neurone lesion.

Tropical Spastic Paraparesis/Human T cell Lymphocyte Virus Type I-Associated Myelopathy

(TSP/HAM) is a neurological disease characterised by slowly progressive spastic paraparesis with insidious onset in adulthood. It has been found all around the world (except at the

Artics), mainly in tropical and subtropical regions. The diagnostic criterion of TSP/HAM is seropositivity to HTLV-I (Cassab and Penalva-de-Oliveira, 2000; Maloney et al 2000;

Zaninovic’, 2001).

TAN is a form of tropical myeloneuropathy which was first described in Nigeria but also occurs in other parts of the tropics. Its clinical presentation is characterized by gradual onset of ataxia due to posterior column loss. The clinical diagnosis requires at least two of the following features: bilateral optic atrophy, deafness, predominantly posterior column

Konzo and cassava toxicity 5 myelopathy and polyneuropathy. The occurrence of TAN is associated with chronic moderate dietary cyanide exposure arising from cassava (Howlett, 1994).

Table I-1: Characteristic features of four tropical myeloneuropathies (Tylleskär et al, 1994c)

Konzo Tropical ataxic Neurolathyrism HTLV-I neuropathy associated myelopathy Geographical area Africa Africa Asia/ Africa Worldwide Occurrence epidemic endemic epidemic and endemic and endemic endemic Highest prevalence 3 % 3 % 3 % 0.1 % Familial clustering yes yes yes yes Type of onset acute slow acute slow Course permanent progressive permanent progressive High incidence age < 40 > 40 < 40 > 40 group Main neurological findings: Gait abnormality Spastic Ataxia Spastic Spastic paraparesis paraparesis paraparesis Peripheral no yes no common neuropathy Sphincter no no rare yes involvement Optic atrophy rare yes no no Deafness no common no no Etiology Attributed Attributed to Caused by Caused by to weeks of prolonged, months of high chronic HTLV-I high varying cyanide grass pea infection cyanide exposure from (Lathyrus sativus) exposure cassava consumption from cassava

Konzo and cassava toxicity 6

Neurolathyrism is an upper motor neurone disease caused by excessive and prolonged consumption of grass pea (chickpea), Lathyrus sativus, which contains the glutamate analogue neurotoxin ß–N-oxalyl-α,ß-diaminopropionic acid (ß-ODAP) also known as ß-

Oxalylaminoalanine(BOAA). ß-ODAP is an excitotoxic amino acid that presumably acts on the neuronal glutamate receptors. Neurolathyrism is epidemic and endemic in geographic areas subject to famine and drought such as Afghanistan, Bangladesh, China, Ethiopia, Nepal and India. Neurolathyrism is characterized by spastic paraparesis of the legs with or without sphincter disturbances (Spencer, 1999; Getahun et al, 1999).

Konzo is a distinct disease entity with selective upper motor neuron damage which is characterised by a sudden onset of an irreversible, a non-progressive and symmetrical spastic paraparesis or, in severely affected subjects, tetraparesis (Howlett et al, 1990; Tylleskär,

1994b; Banea-Mayambu, 1997c; Tshala-Katumbay, 2001a; Mwanza et al, 2005). Konzo has been attributed to the high dietary cyanide exposure from insufficiently processed roots of bitter cassava (Manihot esculenta Crantz) and reported from remote rural areas of

Mozambique, Tanzania, Cameroon, Angola, the Central African Republic and the Democratic

Republic of Congo (DRC) (Trolli, 1938; Cliff et al, 1985; Howlett et al, 1990; Banea et al,

1992a; Tylleskär et al, 1992; Tshala-Katumbay, 2001b; Bonmarin et al, 2002; Ernesto et al,

2002a). Konzo has only been reported from cassava growing and consuming areas but affected populations constitute only a fraction of the total of over 500 million cassava- consuming populations of the tropics. Konzo has some similarities to neurolathyrism but there is no geographical overlap of the two diseases (Howlett et al, 1990; Tylleskär et al, 1994c;

Lambein et al, 2004). Prevalence rates for konzo vary between studies; rates between 1 and 30 per 1000 have been reported (Tylleskär et al, 1992). The total number of confirmed konzo cases in reported studies exceeds 4000 (Tylleskär, 1994a; Bradbury, 2004). The D. R. C. covers the largest reported number of konzo cases. The Health Ministry of D. R. C. estimated

Konzo and cassava toxicity 7 the number of konzo cases in D. R. C. to be around 100,000 (R. D. C., 2000). The age and the sex distribution of konzo show a distinct pattern. No child under the age of 2.5 years, of which most are breast-fed, has ever been found to contract konzo. Women of child bearing age and children 3-13 years of age run the highest risk of contracting konzo. No case of konzo has been reported from nearby urban populations (Banea-Mayambu et al, 1997a).

I.2 Konzo

I.2.1 Background information on konzo

Formerly called epidemic spastic paraparesis (Carton et al, 1986; Rosling, 1988), konzo is a neurological disorder that gives rise to crippling spastic paralysis of both legs (paraparesis) or of both legs and arms (tetraparesis) in severely affected subjects (Tshala-Katumbay, 2001a;

Mwanza et al, 2005). It is an upper motor neuron disease which was first described in the former Belgian Congo (present D. R. C.) by Dr Trolli in a published report that summarised regrouped observations in Kwango district about two affections of unknown origin; epidemic spastic paraparesis, “konzo” of the people in Kwango and a syndrome with oedema and dyschromic cutaneous lesions (Trolli, 1938).

In 1936, Dr Tessitore, a district medical officer in Kahemba area reported an outbreak of an affection that he called “amyotrophic lateral sclerosis” of which he described several cases. In

1937, Dr Mercken noted around Feshi, area neighbouring Kahemba some cases of affection which evoke symptoms of Heine-Medin disease. Some times later in 1937, Drs Doucet and

Orlovitch reported other cases in Moyen-Wamba (the present Popokabaka and Mwela

Lemba), another Feshi neighbouring area (Figure I-1). Some subjects reported having been affected earlier during outbreaks of the year 1928 or 1929 and 1931- 1932. The affection appeared periodically in those areas and was well known by the local population who named it “konzo”.

Konzo and cassava toxicity 8

Figure I-1: Health zones of Kwango District in Bandundu Province, DRC

Konzo and cassava toxicity 9

The word “konzo” originally khoondzo has its origin from kiyaka, the language spoken in

Kwango district and had three meanings (Trolli, 1938; Van der Beken, 1993). It means a fetish used with traps to catch wild animals by weakening their legs, the trap itself and, tied legs. This latter meaning is illustrated by a famous proverb in Kiyaka:

“mene, yakele khosi mutu, khoondzo watholula bidiimbu”,

This means “I was strong, but khoondzo (fetish) has weakened my legs”. As a consequence of this proverb, konzo has come to mean “trapped” or “weakened “or “tied” legs. It is in this sense that konzo is used to denote the paralytic disease and thus, a local belief that the disease is related to a bad destiny or sorcery (Van der Beken, 1993).

Lucasse (1952) who did not read the report of Trolli, described and suggested the first clinical description of konzo observed in some affected subjects 14 years ago in Kwango district, as follows:

Bilateral paresis of the lower limbs which is accompanied by spasms in the adductor and flexor muscles of the lower part of the body giving rise to vicious attitudes of lower limbs and sometimes of spine (lordosis).

After a period with no further reports on konzo, the National Planning Centre of Human

Nutrition of Zaire (present D. R. C.) CEPLANUT (currently PRONANUT) in 1982 reported hundreds of cases of spastic paraparesis from an outbreak that started in 1978 in the Central part of the Bandundu province, neighbouring Kwango district (CEPLANUT, 1982). Another outbreak occurred in 1983, in the northern part of the above area and up to now, outbreaks of konzo are still occurring in the Bandundu province and other parts of D. R. C. (Tshala-

Katumbay, 2001b; Bonmarin et al, 2002).

Other outbreaks of konzo have been described in several other parts of Africa, especially in

Mozambique, Tanzania, Central African Republic, and Cameroon (Cliff et al, 1985; Howlett et al, 1990; Tylleskär et al, 1994c; Ernesto et al, 2002a).

Konzo and cassava toxicity 10

Konzo is now accepted as the scientific name for a distinct human disease entity which is characterised by a sudden onset of a non-progressive and irreversible spastic paraparesis in a person formerly without other symptoms (Howlett et al, 1990; Tylleskär, 1994b).

I.2.2 Clinical features and differential diagnosis

I.2.2.1 Clinical features and classification

Konzo is a distinct type of upper motor neuron disease with a typical clinical picture of crippling spastic paraparesis (WHO, 1996). The clinical picture of konzo is identical in all studies (Trolli, 1938; Lucasse, 1952; Carton et al, 1986; Howlett et al, 1990; Tylleskär et al,

1995; Cliff and Nicala, 1997; Tshala-Katumbay, 2001b; Bonmarin et al 2002; Ernesto et al,

2002a). The disease typically occurs in an apparently healthy person and there is no prodromal phase or triggering illnesses. The onset is characterized by an abrupt paraparesis occurring the first days of the illness. A common history is that of a healthy person who goes to bed feeling well and wakes up during the night or early morning unable to stand or walk.

The paraparesis may also occur abruptly during or after manual work or a long walk. Initial symptoms are often described as heaviness, trembling or weakness of the legs associated with difficulty or inability to stand. Other complaints that may appear over time include weakness in the arms or hands, difficulty in articulating speech, and blurring of vision. Sensory symptoms of radicular low back pain, and paresthesia in the legs, can also be present but these usually clear in the first weeks or months. Incontinence is typically absent. The disease affects mainly children and women of childbearing age (Howlett, 1994; Tylleskär, 1994b; WHO,

1996).

Although the main clinical picture of konzo consists of the sudden onset of a non progressive and symmetrical spastic paraparesis of the legs in affected subjects, the diagnosis of konzo is based on the WHO criteria (WHO, 1996) given in the Table I-2.

Konzo and cassava toxicity 11

These criteria are easy to use in the field to screen the population. However Tshala-Katumbay

(2001a) suggested a new version with more operational criteria in comparison with the WHO criteria (Table I-2).

The degree of physical disability caused by konzo was classified by Lucasse (1952) and later amended by WHO (1996) as follows:

• Mild form: when the patient does not need to regularly use any walking aid

• Moderate form: when the patient regularly uses one or two stick(s) or crutches

• Severe form: when the patient is bedridden or unable to walk without living support.

Table I-2: WHO criteria for konzo versus newly suggested criteria (with permission and from

Tshala-Katumbay, 2001a)

Criteria WHO New version

1 Visible symmetric spastic Sudden onset of a non –progressive

abnormality of gait while walking bilateral and symmetric spastic

or running abnormality of gait while walking or

running

2 History of onset less than one week Bilaterally exaggerated knee or ankle

followed by a non-progressive jerks

course in a formerly healthy person

3 Bilaterally exaggerated knee or Absence of objective sensory and genito-

ankle jerks without signs of disease urinary symptoms

of the spine

4 Absence of grass pea (Lathyrus Living under conditions of sub-acute or

sativus) consumption chronic exposure to cyanogens and

undernutrition at the onset

Konzo and cassava toxicity 12

Figure I-2: From left to right: mild form, moderate form and severe form of konzo in young

subjects

This classification (Figure I-2) is easy to use even by paramedical agents unfamiliar with the symptoms but it is sometimes difficult to distinguish between slightly konzo affected persons

(mild form) with non-affected persons.

I.2.2.2 Differential diagnosis

Konzo with its upper motor neuron manifestations can be confused with other diseases. Using the WHO criteria, konzo by its spastic paraparesis can easily be distinguished from causes of flaccid paraplegia such as poliomyelitis, leprosy or trauma (Howlett, 1994). The commonest neurological diseases to be considered in its differential diagnoses include neurolathyrism,

TAN and TSP/HAM (Table I-1).

Konzo and cassava toxicity 13

Konzo is clinically very similar to neurolathyrism but differs from TAN and TSP/HAM

(Tylleskär et al, 1994c; Cliff and Nicala, 1997; Zannovic’, 2001; Lambein et al, 2004). The socio-economic conditions of konzo and neurolathyrism patients are very similar as well.

Both diseases can be considered a sign of poverty, monotonous diet and illiteracy (Getahun et al, 2002b). Neurolathyrism only differs from konzo with the somewhat higher age of onset, predominance of males among the affected, sphincter involvement in some cases and the absence of cranial nerve involvement (Tylleskär et al, 1994c). There is no geographical overlap between the consumption of the grass pea and cassava and therefore there is no geographical overlap of the two diseases. It would be difficult to make a differential diagnosis between neurolathyrism and konzo if both disorders occurred in the same population

(Tylleskär et al, 1994c; Lambein et al, 2004).

Konzo and TAN have been attributed to dietary cyanide exposure from consumption of insufficiently processed cassava roots, but rates of exposure differ in both diseases (Howlett et al, 1990). In contrast to konzo, TAN is a progressive disorder with slow onset that mainly affects older adults. Furthermore, konzo involves damage to upper motor neurone, whereas

TAN is mainly caused by damage of sensory neurons in the spinal cord resulting in ataxia.

TAN rarely progresses to inability to walk, whereas a high proportion of the konzo-affected subjects are unable to walk. About half of the TAN cases have optic atrophy which is rare among konzo cases. About one in five TAN cases has exaggerated reflexes, a sign that occurs in all konzo cases (Howlett, 1994; Tylleskär et al, 1994c).

TSP/HAM is clinically possible to distinguish from both konzo and neurolathyrism, although the clinical features of TSP/HAM include typical signs that are similar to both diseases such as muscle weakness in the legs, hyperflexia, clonus and extensor plantar responses.

TSP/HAM is characterised by a chronic progressive spastic paraparesis with sphincter disturbances, not to mild sensory loss, absence of spinal cord compression, urinary

Konzo and cassava toxicity 14 incontinence, impotence and seroposivity for HTLV-I antibodies (Tylleskär, 1994b; Cliff and

Nicala, 1997; Cassab and Penalva-de-Oliveira, 2000; Maloney et al 2000; Zaninovic’, 2001).

I.2.3 Epidemiology

The occurrence of konzo is limited to geographical pockets in rural Africa and the majority of cases occur in epidemic outbreaks during the dry season (Rosling, 1997; Bonmarin et al,

2002). Sporadic cases also occur, but they are also restricted to dry or war periods with monotonous cassava diet. Not a single case has been identified in an urban population

(Tylleskär et al, 1995). More than 4000 cases have been confirmed from reported studies. Of these, more than half of the cases are from D. R. C. but the reported number of cases is undoubtedly underestimated, as case detection is incomplete in the remote rural areas affected. The Health Ministry of D. R. C. has estimated the number of konzo cases in D. R. C. to be around 100,000 (R. D. C., 2000). Two extensive epidemic outbreaks, each numbering more than 1000 cases have been brought to the attention of the scientific community; the first reported outbreak occurred in Kwango district in the southern part of Bandundu province of

D. R. C. in 1936-37, and the second in Nampula Province of Mozambique in 1981. Other smaller outbreaks have been reported from very poor remote rural population of Central

African Republic, Mozambique, Tanzania, Cameroon, Angola, and Democratic Republic of

Congo (D. R. C.) (Howlett et al, 1990; Banea-Mayambu et al, 1992a; Tylleskär et al, 1994c;

Tshala-Katumbay, 2001b; Bonmarin et al, 2002; Ernesto et al, 2002a).

Konzo primarily affects children above the age of three and women in the fertile age group.

Adult males are less frequently affected. No breast-fed child (in affected populations, the breast feeding period extends beyond two years of age) has been found to have contracted konzo. A pronounced familial clustering of cases of konzo has been noted in all affected populations. Prevalence varies between studies and between most affected villages.

Prevalence rates of between 1 and 30 per 1000 have been recorded (Howlett, 1994; Cliff et al,

Konzo and cassava toxicity 15

1997b). The peculiar geographical, seasonal and age variation in occurrence as well as the abrupt onset have facilitated epidemiological studies for possible etiological factors in konzo

(Howlett, 1994; Tylleskär et al, 1995; WHO, 1996; Banea-Mayambu et al, 1997a; Rosling,

1997).

I.2.4 Infection or toxico-nutritional etiology?

“The etiology of konzo is unclear. Information given by local people is useless. Often, they say the disease is sent either by God or by the ndoki, the bad spirit of enemies” stated

Tessitore in 1930s (Trolli, 1938). Three generations later after the first report, the etiology of konzo has not been established with certainty. The infectious etiology was proposed (Trolli,

1938; Lucasse, 1952; Carton et al, 1986) but konzo patients do not show any signs of infections and are sero-negative to HTLV-I and other retroviruses (Tylleskär et al, 1996;

Tshala-Katumbay et al, 2001b). The facts that outbreaks are restricted to rural areas without any secondary cases along connecting roads or in neighbouring urban areas argues against infectious etiology. So far, all studies on konzo have failed to demonstrate an infectious etiology (WHO, 1996; Rosling, 1997).

A toxico-nutritional hypothesis was suggested in 1930s by Georgiades who observed similarities between konzo and lathyrism and recommended further studies on cassava concerning the processing and detoxification methods used before its consumption. Konzo might be caused by cyanide exposure resulting from consumption of insufficiently processed cassava roots and simultaneous low dietary intake of sulphur containing amino acids providing substrate to thiocyanate conversion (Trolli, 1938; Cliff et al, 1985). This hypothesis is supported by a consistent association between temporal and geographical occurrence of konzo and the chain of events that leads to high cyanide and low sulphur intake. This chain is:

• Intensive cultivation of bitter cassava varieties in poor rural areas,

Konzo and cassava toxicity 16

• A cassava dominated diet, which is brought about by having a farming system

dominated by bitter cassava

• Shortcuts in processing as indicated by high residual levels of cyanogens in cassava

products consumed,

• High cyanide intake indicated by high urinary and serum thiocyanate levels,

• Low intake of foods rich in sulphur containing amino acids indicated by low urinary

inorganic sulphate levels (Tylleskär, 1994d; Tylleskär et al, 1995; Cliff et al, 1997a;

Rosling, 1997; Banea-Mayambu et al, 1997c)

High consumption of cassava is by itself not sufficient to cause konzo and within the affected populations, cassava is consumed daily by everybody but only certain percentage of the population acquire konzo (Tylleskär, 1994b). The underlying cause inducing the high exposure to cyanide and unbalanced diet include drought, intensive trade of cassava by poor farmers and collapse of the socio-economic fabric due to political conflicts and civil war

(Banea-Mayambu et al, 1997a; Banea-Mayambu et al, 1997b; Cliff et al, 1997b; Tshala-

Katumbay et al, 2001b). So far, the resulting diet and toxic exposure associated to konzo is similar in all areas and konzo has not been reported from any area lacking this unbalanced diet (Rosling, 1997).

Although the etiology and the exact cellular malfunctions induced by overconsumption of cassava products remain unclear, cyanide (CN-), 2-aminothiazoline-4-carboxylic acid (ATC), cyanate (OCN-) and thiocyanate (SCN-) have been suspected playing a role in the pathogenic mechanism of konzo (Spencer, 1999; Tor-Agbidye et al, 1999):

• CN- has been suggested as a causal factor in some series of neurological disorders

because of its potential inhibitor effect on mitochondrial energy transformation

secondarily inducing neuronal dysfunction. CN- is unlikely to be responsible since the

Konzo and cassava toxicity 17

outcome of sub-lethal cyanide intoxication is Parkinsonism, with changes in basal

ganglia, cerebellum and cerebral cortex;

• ATC, a minor cysteine-dependent metabolite of cyanide, has not yet been investigated

for its systemic toxicity. However, its intracerebroventricular injection in rats induces

seizures and hippocampal damage, neither of which are known to occur in konzo;

• OCN-, a normal human metabolite that is produced by the spontaneous degradation of

urea (carbomoylation), is known to cause neurodegenerative disease in humans and

animal (WHO, 2004); but these neurological conditions appear to be more closely

related to TAN rather than to konzo (Tor-Agbidye et al, 1999);

• SCN- is generally considered a major, innocuous detoxification product of cyanide.

SCN- is constantly elevated in subjects with konzo (Tylleskär, 1994, Banea-Mayambu,

1997c; Tshala-Katumbay, 2001b). Since experimental evidence shows that SCN-

increases glutamate binding to the α-amino-3-hydroxy-5-methyl-isoxazole-4-

propionic acid (AMPA) receptor and potentiates AMPA-mediated responses. This

might secondarily induce excitotoxic effects and hence neuronal disfunction or cell

death. The potential role of thiocyanate in konzo merits attention (Tor-Agbidye et al,

1999).

I.3 Dietary exposure to cyanide from cassava

I.3.1 Cassava

I.3.1.1 Classification and botany

Cassava (Manihot esculenta Crantz) is the English name given to the manioc plant, a hardy perennial shrub belonging to the family Euphorbiaceae and ranging in height between 1 – 3 m depending on growing conditions (Brough, 1991). There are many cultivars or varieties

Konzo and cassava toxicity 18 under cultivation. Cassava varieties are usually classified into sweet and bitter cultivars but no morphological or other taxonomic characteristics seem to be associated with this classification (Nweke and Bokanga, 1994). The genus Manihot incorporates over 200 species of which Manihot esculenta is the most important, from the nutritional and economic points of view (Nartey, 1978).

The shrub may have multibranched or unbranched stems, grey, green or brown in colour, with large palmate leaves (Figure I-3). The primary leaves are unlobed, whereas the secondary leaves are 3-lobed, and subsequent leaves develop, the lobes increase in number, reaching the number of lobes characteristic of the cultivar. Mature stem cuttings which are universally used as propagules and planted erect or at a particular angle for vegetative regeneration of plants, give rise to roots at the cut end via callus tissue formation, and adventitious roots form at nodes in the soil.

Figure I-3 : Cassava roots harvested and cassava plant in the field

Konzo and cassava toxicity 19

The roots are initially fibrous, but gradually undergo enlargement. At maturity, they become fusiform, long and slender, occasionally globose appendices of the stem in the upper rhizosphere, seldom penetrating deeply into the soil. The mature root tuber (Figure I-3) possesses three distinct regions, namely, the phelloderm or peel, the cortex or flesh, and the central vascular core. The peel is generally 1-4mm thick and composed of outer epidermis, a sub-epidermis and an inner layer readily separable from the bulk of the tuber. The cortex consists of a mass of parenchyma cells and constitutes the region of carbohydrate storage.

Generally, the cortex lacks xylem vessels, and is therefore without fibre, but older tubers develop hardened xylem vessels, giving rise to stringy tubers, undesirable for food. Root pigmentation may vary with respect to variety, from light yellow, brown to pink and intermediate shades, whereas cortex pigmentation varies from white, yellow to pink. The tubers are the most valuable part of the plant, although in some countries the leaves are used as green vegetable (Nartey, 1978; Brough, 1991).

I.3.1.2 Introduction and distribution of cassava in Africa

Cassava is believed to have been introduced originally in the Gulf of Benin in 1562 and along the Congo River in 1611 from where it spread to the west coast of Africa (Nartey, 1978;

Carter et al, 1997). Later introductions in the islands of Reunion, Madagascar and Zanzibar led its spread in East Africa. Finally, it spread inland in all directions to encompass the region of Lake Tanganyika (Nartey, 1978; Brough, 1991).

The Portuguese first brought cassava to Africa in the form of flour or “farinha”. The

Tupinamba Indians of eastern Brazil had taught them techniques of cassava preparation and production and, they had developed a liking for the various processed forms. Cassava flour was used as a provision for ships plying between Africa, Europe and Brazil. The first mention of cassava cultivation in Africa dates back to 1568. At first, it was cultivated with the sole purpose of supplying slave ships, until 1600. In the late 19th century cassava had been

Konzo and cassava toxicity 20 successfully incorporated into many farming systems of Central Africa instead of millet, yam and plantain, the former staple food in most areas along the Congo River (Carter et al, 1997).

The ultimate wide distribution of cassava in the whole of tropical Africa was motivated by the ability of the crop to withstand locust attacks, and to tolerate drought, poor soils and weeds.

These characteristics, together with the fact that the crop can be left without harvesting over several years, made it a useful security against periods of famine (Nartey, 1978).

In the late 19th and 20th centuries, colonial administrators encouraged diffusion and increased cultivation of cassava. The encouragement by the colonial governments may often have taken place in a manner insensitive to the applicability of cassava to local farming systems and food habits. Moreover colonial governments displayed an ambivalent attitude towards cassava.

Whilst it was introduced as an anti-famine and anti-locust crop, cassava was also thought to promote laziness, soil depletion and malnutrition. Post-independence diffusion of the crop in

Africa has primarily been the result of local processes of migration and agricultural change.

There is ample evidence of the willingness of African farmers to experiment with the search for new crops and varieties. Cassava’s special characteristics make it well adapted to farmers’ risk aversion strategies and allow it to be grown under a great diversity of circumstances and changing economic conditions. The consumption of cassava leaves, in frequent rather than sporadic form, was probably an African invention (Carter et al, 1997).

Currently, cassava is grown on wide scale between latitudes 30° north and south, the so called

“cassava belt”, an ecological zone which coincides with many of the less developed countries where cassava is adapted to the prevailing conditions (Nartey, 1978).

I.3.1.3 Importance and advantages

Cassava (Manihot esculenta Crantz) is a shrub widely grown for its tuberous roots in tropical regions of Africa, Asia and Latin America. Sweet and bitter cultivars are produced as food, feed and for industrial uses.

Konzo and cassava toxicity 21

Cassava roots form a staple food for an estimated 500 million people in the tropics and the leaves are commonly consumed as a vegetable in several areas. Cassava ranks fourth on the list of major food crops in developing countries after rice, wheat and maize (FAO, 1990).

Cassava is of great importance for food security in Africa in general, and D. R. C in particular. Cassava possesses a number of useful agricultural traits. The crop is a relatively efficient producer even under adverse environmental conditions such as erratic low rainfall and low soil fertility. Cassava productivity in terms of calories per unit land area is significantly higher than that of other staple food crops. The edible portion in percent of dry weight of the root crop is high. Cassava is resistant to locust damage and most pests. Growth of cassava requires a low input in the timing of labour. Except being sensitive to drought shortly after planting, cassava requires no special planting or harvesting dates. The roots can be stored in the ground without harvesting for a lengthy period of time, up to three years or more after the formation of the edible roots is complete. Hence, cassava cultivation serves as something like a household food bank that can be drawn upon when adverse agro-climatic conditions or civil unrest limit the availability of and access to other food (Koch et al, 1994;

Scott et al, 2000).

In sub-Saharan Africa, cassava provides daily food products for nearly half of the continent’s population. DRC is the country with the highest per capita consumption in the world, about

60 % of total daily energy intake is provided by cassava (FAO, 1990). In addition, cassava leaves contribute 20 % of the protein in Congolese diets. The Congolese staple food production takes mainly place on traditional farms. The Congolese smallholder farming is characterized by reliance on family labour, on a small stock of physical capital and on a large area of land. Women play a predominant role in farming, processing, and marketing. In rural areas, most traditional subsistence food crops have become important as cash crops and urban food demand is the driving force behind this evolution. Farms without an adequate access to

Konzo and cassava toxicity 22 markets are generally characterized by low levels of cash income and surplus production

(Goossens, 1996). Cassava production is a commercial activity, and not merely a subsistence agricultural activity. Cassava is an important source of cash income for poor farmers as well as for prosperous ones. Both rich and poor farmers often sell a higher proportion of cassava than from any other crop or income earning activity. The proportion of cassava production marketed is a good indicator of the level of diversification in the crop production activity of an area (Nweke and Bokanga, 1994).

I.3.1.4 Production and consumption levels

“Cassava is apparently emerging from its obscurity in the tropics and is marching northward and southward to fill new roles in temperate climates”, is the assessment of Franklin D.

Martin quoted by Nartey (1978).

In the last two decades, cassava production grew at a more modest pace, 1.8 percent annually.

Production of cassava grew by 27 percent between 1983 and 1996 to 164 million metrics ton.

On a per capita basis, production of cassava in developing countries remained virtually constant at 37 kg per capita, supported mainly by the per hectare production growth in Sub-

Saharan Africa. Production tends to be highly skewed toward particular regions. Slightly more than half the global production of cassava takes place in sub- Saharan Africa, followed by Southeast Asia with 23 percent and Latin America with 20 percent (Figure I-4).

Most statistics do not usually distinguish between sweet and bitter varieties; in some, sweet varieties are not included as they are commonly grown as a secondary crop for home consumption. Brazil is the largest producer of cassava in the world, but most of the crop is consumed locally and exports are only a small portion of the total output. The same pattern applies to other important producers, such as Nigeria, Indonesia, D. R. C., India and

Colombia. Cassava does not form an important part of the staple diet in Thailand, and that

Konzo and cassava toxicity 23 country is the world’s largest exporter of cassava products. In contrast, Africa does not export much cassava because production is almost entirely consumed as food.

Figure I-4: Location of cassava production, 1996 (Scott et al, 2000)

India, 3.5% China, 2.2%

Southeast Asia, 23.0%

Sub-Saharan Africa, 51.6% Other South Asia, 0.2%

Latin America, 19.5%

In the last few years most of the important producers have greatly increased their production.

Surplus production of cassava products enters international trade in different forms, such as chips, broken dried roots, meal, flour and tapioca starch. Dried cassava roots and meal are used as raw material for compound animal feed, while cassava starch is used for industrial purposes, particularly the paper and textile industry; grocery tapioca is used solely for human consumption. The principal markets for cassava products are the European Community, USA,

UK and Japan (FAO, 2004).

Konzo and cassava toxicity 24

Between 1983 and 1996, increase of consumption of cassava as food has been particularly rapid in Sub- Saharan Africa at 3.1 percent per year. The region has experienced low and negative economic growth and booming populations (Scott et al, 2000).

Cassava is the staple food in most of D. R. C., and especially in Bandundu province where a great number of konzo cases have been reported (Trolli, 1938; Banea-Mayambu et al, 1992a;

Tylleskär et al, 1994c; Tshala-Katumbay, 2001b; Bonmarin et al, 2002).

Between 1987 and 1990, the annual production of fresh cassava roots in Bandundu was 4 million metric tons while the national production was estimated at 17 million tons. The annual production per household was 7.9 tons and the quantities of cassava produced are ten to twenty times larger than for other crops in Bandundu. The cassava production per farm is relatively stable during the year: between 560 kg and 680 kg per month in Bandundu. Farm purchases of cassava are highest in October and November, when fields are prepared and available labour to harvest roots is limited. Sales are highest from September to February, when receipts from other crops are low. Nearly all households buy cassava from time to time but the quantities are generally small and are used for immediate consumption or to be resold.

Only seven percent of the rural households are not self supporting for cassava (Goossens,

1996)

I.3.1.5 Problems associated with cassava

I.3.1.5.1 Nutritional value

Cassava roots contain around 30 to 40 % of dry matter of which starch and sugars account for approximately 90 %. This renders cassava root, an excellent source of carbohydrate but has extremely low levels of protein and fat (Bradbury and Holloway, 1988)

Unlike the roots, which are essentially a source of carbohydrate, fresh cassava leaves are a good source of proteins and vitamins which can provide a valuable supplement to

Konzo and cassava toxicity 25 predominantly starchy diets (Hahn, 1989). The nutrient content of cassava leaves is comparable with other green leaves and other vegetables generally regarded as good protein sources (Mbemba and Remacle, 1992). Vitamin A, thiamine, riboflavin, niacin and vitamin C are of high concentration in the fresh leaves (Ravindran and Ravindran, 1988; Almazan and

Theberge, 1989)

I.3.1.5.2 Constraints in production

Besides the advantages on the production of cassava, there are many production constraints which can include diseases, pest, weeds, soil and agronomic factors, and socio-economic factors. The major diseases of cassava are leaf diseases, stem diseases and tuber rot.

Vertebrate pest, nematodes, mites and insects may attack or infect the roots and render them susceptible to rot-causing organisms. Cassava can be seriously affected by early weed infestation. Weed competition in cassava crops reduces canopy development, tuberization and tuber yield. The important soil and agronomic factors that affect cassava production are soil temperature and moisture (if above 30° C and if drought is frequent), soil erosion and low soil fertility (continuous cultivation of cassava without adequate erosion control measures, can result in severe and irreversible soil degradation), and poor cultural practices. The main socio- economic factors affecting cassava production relate to inadequate resource allocation and infrastructure (IITA and UNICEF, 1990).

I.3.1.5.3 Post-harvest deterioration

Cassava roots are extremely perishable. They can be kept in the ground prior to harvesting for up to about 2 years, but once they have been harvested (removed from the stem) they begin to deteriorate within 40 – 48 hours (IITA and UNICEF, 1990). The fresh tubers in general have high moisture content, usually between 50 and 70 %, and hence have a relatively low mechanical strength. They also have a very high respiratory rate, and the resultant heat

Konzo and cassava toxicity 26 production softens the texture, which leads to damage. Unlike the other tuber crops, cassava roots do not exhibit exogenous dormancy, have no function in propagation, and possess no bud primordial from which regrowth can occur. For these reasons, cassava roots are more perishable than other tuber crops. Mechanical damage during the harvesting and handling stages also renders cassava root unsuitable to long-term storage. Deterioration of cassava has an adverse effect on the processed product, and thus the crop must be stored properly (IITA and UNICEF, 1990; Ravi et al, 1996, Ravi and Aked, 1996).

I.3.1.5.4 Cassava in human nutrition

Cassava root is an important starchy crop grown in the tropics, which constitutes the staple food of about half of billion people and the leaves are consumed as vegetable. Roots and leaves of cassava contain high level of cyanogenic glycosides mainly linamarin and to a lesser extent lotaustralin which can be hazardous to the consumer and of which the potential toxicity is a public health concern (Essers, 1995; Padmaja, 1995). Roots from the sweet varieties may be eaten raw or cooked by boiling, steaming or roasting. Roots from the bitter varieties required processing to remove the cyanogenic compound before consumption. The processing methods generally adopted comprise combinations of activities such as peeling, boiling, steaming, slicing, grating, soaking or steeping, fermenting, pounding, roasting, drying and milling (Hahn, 1989; Padmaja, 1995). They can be transformed in the form of flour or gari

(granules) and then consumed as fufu (stiff porridge), chikwangue (wrapped steamed boiled paste much stiffer than fufu), tapioca, dried gari or stiff paste of gari (Figure I-5).

Consumption of cassava leaves is of great significance in the nutrition of a population subsisting primarily on cassava. The most widely practiced processing for use of leaves as food involves crushing, parboiling in water, washing and cooking. Sometimes additional ingredients such as pepper, palm oil, fish, peanut or other aromatic ingredients are added. The leaves, in some parts, are sun dried then cut or pounded and finally cooked. Cooked cassava

Konzo and cassava toxicity 27 leaves are served and consumed as the main side-dishes to the processed cassava roots like fufu, chikwangue or boiled cassava root (Hahn, 1989).

Figure I-5: Summary of traditional cassava processing in Africa (from Banea-Mayambu,

1997c)

Fresh cassava roots

Soaking Heaping Cutting Grating

Fermentation Solid state - Moist in water fermentation fermentation

Pounding Drying Drying Drying Pressing

Boiling Pounding Pounding Pounding Roasting

Paste Flour Flour Flour Granules

I.3.2 Cyanide toxicity

I.3.2.1 Introduction

Cyanide (CN-) most commonly occurs as hydrogen cyanide (HCN) and its salts sodium cyanide (NaCN) and potassium cyanide (KCN). Cyanides comprise a wide range of compounds of varying degrees of chemical complexity, all of which contain a CN moiety, to which humans are exposed in gas, liquid, and solid form from a broad range of natural and anthropogenic sources. Cyanogenic glycosides, producing hydrogen cyanide upon hydrolysis, are found in a number of plant species. Cyanides are also produced by certain bacteria, fungi,

Konzo and cassava toxicity 28 and algae. Minute amounts of cyanide in the form of vitamin B12 (cyanocobalamine) are a necessary requirement in the human diet (ATSDR, 1989).

HCN, a colourless or pale liquid or gas with a faint bitter almond-like odour, has a molecular weight of 27.03 and a boiling point of 25.7° C. It is miscible with water and alcohol and slightly soluble in ether.

Cyanide is released to the environment from numerous sources. Metal finishing and organic chemical as well as iron and steel production, and automobile exhaust are major sources of cyanide releases in the atmosphere. Workers in a wide variety of occupations may be exposed to cyanides. The general population may be exposed to cyanides by inhalation of contaminated air, ingestion of contaminated drinking water, and/or consumption of a variety of foods (ATSDR, 1989).

Among the general population, subgroups with the highest potential for exposure to cyanide include active and passive smokers, individuals involved in large scale processing of food high in cyanogenic glycosides and individuals consuming foods high in cyanogenic glycosides (WHO, 2004).

I.3.2.2 Occurrence of cyanogenic glycosides

The cyanogenic glycosides are a group of nitrile-containing secondary plant compounds that yield cyanide (cyanogenesis) following enzymatic breakdown. They are amino acid-derived plant constituents and their functions remain to be determined in many plants; however, in some plants they have been implicated as herbivore deterrents and as transportable forms of reduced nitrogen (Kakes, 1994). Whereas most plants produce a small amount of cyanide associated with ethylene production, between 3000 – 12000 plant species produce sufficient quantities of cyanogenic compounds to be considered toxic (Poulton, 1990). The concentrations of cyanogenic glycosides can vary widely as a result of genetic and environmental factors, location, season, and soil types (Ermans et al, 1980). Several

Konzo and cassava toxicity 29 economically important plants are highly cyanogenic, including white clover, flax (containing linamarin), almonds, sorghum, wild lima bean, rubber tree, and cassava. The agronomically most important cyanogenic food crop however, is cassava (McMahon et al, 1995; White et al,

1998; Vetter, 2000).

I.3.2.3 Cyanogenesis in cassava

All cassava tissues, with the exception of seeds, contain cyanogenic glycosides mainly linamarin and lesser amounts of lotaustralin in about 10 to 1 ratio. An acyanogenic cassava cultivar has never been found (Bokanga, 1994). Cyanogenic glycosides are compartmentalised within the cell vacuole while the complementary hydrolytic enzymes are located within the cytosol of the same cells (White et al, 1994). The amino acids valine and isoleucine are the precursors used in the synthesis of linamarin and lotaustralin, respectively.

The initial step in the biosynthesis of linamarin is the N-hydroxylation of valine followed by the formation of 2-methyl-propanal oxime and its dehydration to yield 2-methylpropionitrile.

Figure I-6: Cyanogenesis from linamarin (McMahon et al, 1995)

Konzo and cassava toxicity 30

The addition of oxygen forms acetone cyanhydrin which is then glycosylated (by a soluble

UDPG-glucosyltransferase) to form linamarin (Conn, 1994). No HCN is released from intact cyanogenic plants, the substrates (cyanogenic glycosides) and the enzymes must be located in different compartments of the cell. Cyanogenesis is initiated in cassava when the plant tissue is damaged. The generation of cyanide from linamarin is a two-step process involving the initial deglycosylation of linamarin and the cleavage of acetone cyanhydrin to form acetone and cyanide (Figure I-6). These reactions are catalysed by linamarase (a ß-glucosidase) and by α-hydroxynitrile lyase (HNL). Since acetone cyanhydrin may enzymatically be broken down by HNL as well as a spontaneously decompose at pH slightly >4.0 or temperature >

30°C, it has been generally assumed that the linamarase is the rate-limiting step (McMahon et al, 1995; White et al, 1998; Vetter, 2000). In spite of the relative instability of acetone cyanohydrin, it can coexist with intact glycosides and HCN in differently processed cassava products. Therefore, cyanogens in cassava products can exist in three forms:

• Cyanogenic glycosides (linamarin and lotaustralin),

• Acetone cyanohydrin,

• Free HCN (Tewe, 1994).

Cassava tubers vary widely in their cyanogenic glycosides content, although most varieties contain 15 to 400 mg HCN equivalent/ kg fresh weight. Occasionally, varieties with high cyanide content (1300 to 2000 mg/ kg) are also encountered (Hahn, 1989).

Cassava leaves also contain high concentrations of cyanogenic glucosides and the values fall mostly in the range of 1000 to 2000 mg HCN equivalent/ kg on a dry matter basis (Bokanga,

1994). Very high values up to 4500 mg/ kg have been occasionally reported. The high content of cyanogenic glycosides in cassava is however a factor restricting its utilization as a food

(Padmaja, 1995).

Konzo and cassava toxicity 31

I.3.2.4 Effect of processing on cyanogenic glycosides in cassava

Cassava roots are processed by a number of methods that vary widely from region to region.

Generally, all those techniques are intended to reduce toxicity and improve palatability and storability (Tewe, 1994). Adequate processing of cassava is of prime importance in eliminating the toxic glycosides and converting cassava into a safe food. Cyanide–yielding substances of cassava are normally reduced to negligible levels by effective processing

(Rosling, 1988). Cassava roots provide an important source of dietary energy but they have some limitations. Firstly roots are readily perishable if they are not processed and they cannot be stored like cereals or other tubers (potatoes, yam). Secondly roots from the bitter varieties cannot be consumed raw and they are unsuitable for roasting and boiling as fresh roots because of the high levels of potentially toxic cyanogenic glycosides (linamarin and lotaustralin). Therefore, cassava roots from such varieties must be processed before consumption to reduce the content of toxic cyanogenic glycosides and their degradation products (acetone cyanhydrin and free cyanide) in the final food product. Because of their high water content, harvested roots rot if they are not processed shortly after harvesting. The processing considerably reduces the water content (about 50 - 70 % in the freshly harvested tuber) and thus facilitates transportation. The processing serves to make the starch of the cassava root suitable for consumption as a major food component in the form of boiled paste, flour or granules in the many different dishes prepared according to cultural preferences.

Cultural preferences vary a lot and the choice of processing method is thus aimed at obtaining a cassava food product which is safe to eat and has a desired taste, flavour and texture.

To be considered as safe for consumption, cyanogens should be removed by processing to a level below 10 mg equivalent HCN per kilogram (ppm) dry weight of cassava product, the recommended safe limit set by FAO/WHO (1991)

Konzo and cassava toxicity 32

Although the processing steps are different for each product, they permit the glucosidase to interact with the cyanogens and the release of cyanide. The processing leads to two end products mostly depending on the locally available processing resources (Hahn, 1989;

O’Brien et al 1995; Ravi and Abed, 1996). Dry chips or cossettes and flour are the main products where sunlight is abundant and wet paste is the main product where water supply is abundant. These products need additional home preparation. Cooked paste, steamed and toasted granules are relatively more advanced that enter the marketing system in ready-to- serve forms, although toasted granules may need minimal preparation by soaking in hot or cold water. These products are usually more convenient and attractive to urban consumers, and competitive with food grains in the market place. Cassava is more often processed into first group products for home use and into second group products for sale. However cassava is widely marketed in form of cossettes in D. R. C. and most of the cassava products pass through fermentation or soaking stage. The period of the fermentation usually lasts a number of days, and varies depending on the product, processing technique and on the market for which the product is intended for sale (Nweke and Bokanga, 1994). But where market access has been improved, the fermentation period tends to decline; e.g. in D. R. C., the completion of a new tarmac road to the capital city of Kinshasa resulted in an increased demand due to improved market access. This caused the farmers producing cassava cossettes with a reduced soaking period from three or four days to one or two days (Tylleskär et al, 1991).

Retting of cassava roots by steeping them in water causes higher losses in total cyanogens and, makes them soft and causes the cells to rupture, releasing linamarase. Cyanogenic glycoside removal can be enhanced by direct leaching into the soaking water (Muzanila et al,

2000). Wet fermentation has been reported to facilitate the breakdown of cyanogenic glycosides to low total cyanogen levels, up to 13,5 % reduction during the first of day soaking, and 65% reduction the second day of soaking (O’Brien et al, 1992) or even in one

Konzo and cassava toxicity 33 experiment up to 90 % reduction after 4 days of fermentation (Padmaja, 1995). The efficiency of fermentation or soaking of cassava root has been well documented as one of the best methods for cyanogen elimination (O’Brien et al, 1992; Padmaja, 1995; Ravi and Padmaja,

1997). Sun-drying of cassava is generally considered to be the least efficient of the various categories commonly practiced in Africa (Mlingi et al, 1995; Essers et al, 1996). The cyanogenic potential of heap fermented cassava roots was significantly lower than those from sun dried (Zvauya et al, 2002). The short-cut method of alternate pounding and drying of cassava roots resulted in a sharp decline in glycoside levels but high cyanohydrin levels may remain if the products are not sufficiently dried (Mlingi et al, 1995). Processing steps such as crushing and pounding may be incorporated prior to sun-drying to increase the efficiency of cyanogen removal. However, sun-drying alone as a processing method of highly cyanogenic cassava varieties remains inadequate if levels are to be reduced to the recommended FAO/

WHO safe limit set at 10 mg HCN equivalent /kg dry (Essers et al, 1996; Bainbridge et al,

1998).

Fresh cassava leaves contain very high levels of cyanogenic glycosides, usually 5 to 20 times more than the amount present in the edible parts of the roots. Effective detoxification processing is required prior to consumption. Bokanga (1994) found after pounding cassava leaves a reduction of the cyanogenic potential by 63 to 73 %. The rapid removal of cyanogens from cassava leaves can be attributed to the presence in the leaves of a high level of linamarase activity, to the extensive mechanical damage imparted to the leaves during pounding thereby facilitating the contact between linamarase and linamarin and promoting cyanogenesis. The removal of cyanide by altering the cooking time and initial water temperature in the preparation of pounded cassava leaves showed that starting with water at a temperature of 27°C, brought down the total cyanide level 3.8 times more effectively than starting with already boiling water and, increasing cooking time from half hour to one and

Konzo and cassava toxicity 34 half hour resulted in a 2.5-fold more effective reduction (Essers, 1989). Gradual heat applied during cooking of pounded cassava leaves accelerates the evaporation of hydrogen cyanide

(boiling point at 25.7°C) and starting cooking at 27°C will increase progressively the temperature to reach the maximum linamarase activity reported to be 55°C. Putting the leaves into boiling water will immediately reduce drastically the activity of linamarase and may therefore prevent liberation of cyanide from its glycosydic bond (Essers, 1989; Bokanga,

1994).

I.3.2.5 Metabolism of cyanogens in Humans

During cassava processing, cyanogenic glycosides (linamarin and lotaustralin) break down into glucose and acetone cyanohydrin through the activity of the endogenous enzyme linamarase. Acetone cyanohydrin gradually breaks down into HCN spontaneously

(temperature and pH dependant) or enzymatically (Figure I-6). These cyanogens (linamarin, acetone cyanohydrin and HCN) can be reduced to negligible levels by effective processing, but insufficiently processed cassava products contain varying amounts of cyanogens.

Following consumption, any of the three types of cyanogens may result in cyanide exposure

(Essers et al, 1992; Rosling, 1994). However the fate of the cyanogens will differ during digestion in the gut and metabolism in the body (Carlsson et al, 1995).

The ingested linamarin is thought to be hydrolysed to glucose and acetone cyanhydrin in the intestinal tract; hydrogen cyanide is then produced by a catalytic reaction in the intestine and rapidly absorbed from the intestine to the blood (Sreeja et al, 2003). A part of ingested linamarin has been found to pass through the human body unchanged, absorbed directly in the intestine and excreted intact in urine (Carlsson et al, 1995; Carlsson et al, 1999; Sreeja et al,

2003). Ingested residual cyanhydrins are assumed to break down to cyanide in the alkaline environment of the gut (Tylleskär et al, 1992). Cyanide is rapidly absorbed by the gastrointestinal tract and distributed throughout the body by the blood. The major portion of

Konzo and cassava toxicity 35 cyanide in blood is sequestered in the erythrocytes, and a relatively small proportion is transported via plasma to the target organs (liver, lungs, kidney, brain, central nervous system). Although cyanide can interact with substances such as methemoglobin in the bloodstream, the majority of cyanide metabolism occurs within the tissues. Cyanide is metabolized in mammalian systems by one major route and several minor routes (Figure I-7).

Figure I-7: Basic processes involved in the metabolism of cyanide (ATSDR, 1997)

Konzo and cassava toxicity 36

The major route of metabolism for cyanide is detoxification in the liver by the mitochondrial enzyme rhodanese (thiosulphate-sulphurtransferase, EC 2.8.1.1), which catalyses the transfer of the sulfane sulphur of thiosulphate to the cyanide ion to form the less toxic thiocyanate

(Figure I-7). About 80 % of cyanide is detoxified by this route. The rate-limiting step is the amount of thiosulphate which is produced by ß-mercaptopyruvate resulting from transamination of cysteine. While rhodanese is present in the mitochondria of all tissues, the species and tissue distributions of rhodanese are highly variable. In general the highest concentrations of rhodanese are found in the liver, kidney, brain and muscle, but the supply of thiosulphate is limited.

Cyanide is principally excreted as thiocyanate in the urine but the limiting factor in cyanide metabolism is the concentration of the sulfur containing substrates primarily thiosulphate, but also cystine and cysteine as product of methionine (essential amino acid) and cysteine catabolism (Figure I-8). There are several pathways for cysteine catabolism. The more important catabolic pathway is that via a cytochrome –P450-coupled enzyme, cysteine dioxygenase that oxidises the cysteine sulphydyl to sulfinate, producing the intermediate cysteinesulfinate. Cysteinesulfinate can serve as a biosynthetic intermediate undergoing decarboxylation and oxidation to produce taurine, the bile salt precursor. The enzyme cystathionase can also transfer the sulphur from one cysteine to another generating thiocysteine and pyruvate. Transamination of cysteine yields ß-mercaptopyruvate which then

2- reacts with sulphite (SO3 ) to produce thiosulphate and puryvate. Both thiocysteine and thiosulphate can be used by the enzyme rhodanese to incorporate sulphur into cyanide, thereby detoxifying the cyanide to thiocyanate (Hoffer, 2002; Komarnisky et al, 2003;

Stipanuk, 2004).

Konzo and cassava toxicity 37

Figure I-8 : Cysteine catabolism

Methionine glutamate α-ketoglutarate

Cysteinesulfinate Cysteine aminotransferase ß-Mercaptopyruvate

Cysteine

SO 2- + H 20 3 NH4 3-mercaptopyruvate Pyruvate sulphurtransferase

Hypotaurine Pyruvate + Thiosulphate Thiocysteine

+ cyanide Taurine rhodanese

Cysteine +orThiocyanate Thiocyanate

Konzo and cassava toxicity 38

The level of thiocyanate normally present in body fluids is low but increases with chronic exposure to cyanide and with smoking habits (Vesey et al. 1999, Kussendrager and Van

Hooijdonk, 2000). Thiocyanate remains the most useful chemical biomarker for dietary cyanogen intake because it is a very stable metabolite that can be determined with relatively cheap, specific and sensitive methods (Rosling, 1994, Ressler and Tatake, 2001). Urinary thiocyanate is commonly used to check cyanogen overload in a population using cassava roots and cassava products as staple food (Haque & Bradbury, 1999, Ernesto et al. 2002a).

Cyanide can also be metabolized by several minor routes, including the combination of cyanide and hydroxycobalamin (vitamin B12) to yield cyanocobalamin (vitamin B12) and the non-enzymatic combination of cyanide with cystine, forming 2-aminothiazoline-4-carboxylic acid (ATC) which is excreted via the urine. However in protein-deficient subjects, in whom sulfur amino acids are low, cyanide may conceivably be converted to cyanate (Tor-Agbidye et al, 1999).

I.3.2.6 Effects of cassava toxicity in humans

The toxicity of cassava arises from the release of cyanide during hydrolysis of cyanogenic glycosides by the glucosidases of intestinal microflora. Intact linamarin has also been reported to be absorbed through the intestinal mucosa. Cyanide can also be released in vivo by glucosidases of the liver and other tissues, causing in situ cytotoxicity (Padmaja, 1995).

Cyanide is a potent toxin that acts by inhibiting cellular respiration.

Cyanide toxicity occurs when the capacity for conversion of cyanide to thiocyanate is acutely exceeded. This leads to inhibition of cytochrome oxidase and prevents cell respiration and oxidative respiration. Acute toxicity results from the ingestion of lethal amounts of cyanide.

Doses of 50 to 100 mg are reported to be lethal to adults. Acute cassava poisoning, sometimes leading to the death of whole families, has been occasionally reported in humans after consumption of bitter cassava roots or inadequately processed cassava, usually at times when

Konzo and cassava toxicity 39 the normal eating habits are affected by famine. More common are incidences of chronic cyanide toxicity due to prolonged consumption of insufficiently processed cassava. Chronic toxicity of cassava has been implicated in several diseases such as tropical ataxic neuropathy, endemic goiter and konzo. Many of these conditions result from the consumption of poorly processed cassava.

I.4 Conclusion

Konzo is a paralytic disease rarely reported and little known even in the affected area. This symmetric paralysis of both legs affects mainly women at childbearing age and children above three years old, among the poor rural population of remote areas of Sub-Saharan Africa where cassava is the staple food. Affected persons live far from the big city where decisions are made and they are of no particular interest for political authorities. Literature on konzo is limited to some epidemiological consideration. Evidence linking the disease with high consumption of improperly processed cassava roots has been established. In addition, a low intake of sulphur amino acids (methionine and cysteine) needed for the metabolic detoxification of cyanide in the human body has also been thought to be an important co- factor in the development of konzo (Cliff et al, 1985).

Up to date, there is no medicine to cure this crippling non-progressive and irreversible disease. Nevertheless, some investigators (Ernesto et al, 2002b) have proposed the following strategies to prevent and to eliminate konzo:

• Introduction of other staples, vegetables, pulses and fruits to decrease the daily

cyanide intake and broaden the diet of the people,

• Improved processing of cassava roots to produce products that have less residual

cyanide,

• Introduction of low cyanide, high yielding, well-adapted, disease-resistant varieties of

cassava,

Konzo and cassava toxicity 40

• and improved early warning systems of a possible konzo epidemic.

I.4.1 Rationale of the research

Konzo is a neglected and an emerging neurological crippling disease that affects the poor segments of remote rural communities of sub-Saharan Africa. There is convincing evidence linking konzo with high cyanide exposure (Tylleskär, 1994a; Ludolph and Spencer, 1996):

First, heavy dietary reliance on bitter cassava is strongly associated with the development of konzo. Secondly, there is a consistent association between shortcut soaking cassava processing and outbreaks of konzo. Thirdly, the bitter cassava contains cyanogenic glycosides, mainly linamarin and to a lesser extent lotaustralin which upon hydrolysis release the mitochondrial toxin cyanide, a potent inhibitor of cytochrome C oxidase (complex IV of the mitochondrial respiratory chain). Also, in affected populations, the excretion of urinary thiocyanate and the ratio thiocyanate/ inorganic sulphur in the blood of affected populations are increased. (Tylleskär et al, 1991; Banea et al, 1992b; Tylleskär et al, 1992).

Definite confirmation of an etiologic role of cyanide in konzo by identification of the mechanism in an experimental animal model, or a quasi experimental preventive intervention is lacking. There is therefore a need to know if a wild plant (food), a vitamin deficiency, another toxin in cassava or some other factor may be contributing to or be an essential factor in the aetiology of konzo (Tylleskär, 1994a, Bonmarin et al, 2002).

I.4.2 Objectives

I.4.2.1 General objective

• Identify associated nutritional factors involved in konzo

I.4.2.2 Specific objectives

• To review the literature on konzo and its relation to cassava dietary exposure.

Konzo and cassava toxicity 41

• To determine the prevalence and associated dietary factors with Konzo.

• To assess dietary intake with special emphasis on intake of sulfur amino acids.

• To quantify the daily intake of cyanogen, and to estimate the amount of sulphur amino

acids required for their detoxification in konzo affected areas.

• To determine free amino acids in order to evaluate the presence of inherent potentially

toxic nonprotein amino acids in the cassava products.

• To develop total protein amino acids profiles of cassava products in order to evaluate the

dietary protein quality and to compare them with the amino acid requirements of children

and adults.

• To monitor the level of dietary exposure to cyanogens from cassava in the selected konzo

affected community.

• To assess a potential relationship between urinary thiocyanate as biomarker of daily

cyanogen exposure and taurine as modulator of neuroexcitation.

Konzo and cassava toxicity 42

CHAPTER II:

OCCURRENCE OF KONZO AND DIETARY PATTERN*

* This chapter will be submitted for publication in Tropical Medicine and International Health as: Delphin Diasolua Ngudi, Jean-Pierre Banea-Mayambu., Fernand Lambein and Patrick Kolsteren. Konzo and dietary pattern in cassava-consuming populations of Popokabaka, Democratic Republic of Congo

Konzo and cassava toxicity 43

II Occurrence of konzo and dietary pattern

II.1.1 Introduction

In Popokabaka district (D. R. C.) where cases of konzo were first reported three generations ago (Trolli, 1938), new cases were found and the diet has changed little. Cassava flour, the so called “luku” or “fufu” is consumed together with a sauce prepared from cassava leaves. Meat and fish are not eaten daily in the villages (Tshala-Katumbay, 2001a).

Protein-energy malnutrition which is a major health problem in the region has been attributed to the combined effect of infections and inadequate diet. An unbalanced diet is suggested to be the main risk factor for several diseases such as obesity, stroke, cancer (Thiele et al, 2004) and a factor aggravating growth retardation in children in Bandundu Province, D. R. C.

(Banea-Mayambu et al, 2000).

The present paper reports the prevalence of konzo, the household risk factors associated and the dietary pattern in cassava consuming populations. Dietary intake patterns and socio- economic variables are well known indicators for assessing nutritional status of a community

(Agrahar-Murugkar and Pal, 2004).

II.1.2 Materials and Methods

II.1.2.1 Study area

The study was conducted in Popokabaka rural health zone (Prhz) (5°38’35” – 5°43’0” latitude

South, 16°34’60” – 16°37’8” longitude East), district of Kwango, province of Bandundu (1 -

8° S, 16 – 20° E), D. R. C. in February 2003 during the rainy season. Prhz covers an area of

7,949 km2 with a population of 149,227 inhabitants in 2002 (density of 19 inhabitants/ km2).

Prhz is divided into 38 health areas. The vegetation is bushy savannah with few forest galleries where the climate is tropical with annual rainfall varying around 1200 mm and a

Konzo and cassava toxicity 44 long dry season of about 5 to 6 months (Kama, 1970). The poor soil makes the rural population focus on the cultivation of cassava, the most important cash crop and the main staple food for this region. Prhz faces various constraints for its development. Accessibility to

Prhz is difficult especially during the rainy season and limits the communication with

Kinshasa, the capital of D. R. C. where manufactured goods and other food products can be purchased or excess of crop production can be sold. Recrudescence of endemic pathologies

(malaria, trypanosomiasis, tuberculosis, leprosy, etc), poverty of the population and limited number of unsafe drinking water sources are main problems encountered by Prhz (Mwela,

2002).

II.1.2.2 Subject

Four health areas in Prhz with a total population of 12,416 inhabitants in 2,069 households

(national mean size of household members is 6, see R. D. C. 2001) were selected based on the accessibility and the reported prevalence of konzo in the areas. The Epitable calculator of

Epiinfo version 6 was used to calculate the sample size for a single proportion of a limited community study based on the size of the population, the desired precision (0.99), the expected prevalence of konzo (4 %), the designed effect (2) and on the confidence level (95

%). After introduction of those required data, a sample size of 2,685 inhabitants in 448 households was obtained. This sampled population was attained through heads or delegates of household* who were enrolled in this study in a random sampling after selection of the first participant (household) as starting point, near the main entrance road to the village or near the health centre (clinic). The sample size was adjusted upward to 487 participating households to avoid other factors that could decrease the yield of usable responses. No refusal was observed during the survey. Written authorisations were obtained from the administrative and health

* A household was defined as group of persons sharing the same meals since at least 3 months before the survey.

Konzo and cassava toxicity 45 authorities, oral consent and assistance were also obtained from the village leaders and from the subjects.

The heads or delegates of household were interviewed by trained enumerators using a closed and open-ended questionnaire designed to collect the following information: identification of the household (name of the respondent, village, name, sex, marital status, education and origin of head of household), socio- demographic characteristics: family size, family member affected by konzo (name, birthday, onset, sex, degree of walking), owning land farm or husbandry, list of all foods consumed during the previous day (morning, lunch and evening), the origin of the cassava consumed the previous day as staple food and the duration of retting

(soaking), the composition of the “luku” and list of foods consumed often during the rainy season and the dry season. The WHO criteria for konzo were applied to detect cases and to confirm the diagnosis (WHO, 1996).

II.1.2.3 Statistics

Descriptive statistics were used for socio-demographic and other household related variables.

On the basis of the prevalence found, the areas were pooled into two groups: the low prevalence area and the high prevalence area. Thus, the degree of konzo prevalence in the health area was used to measure the risk of konzo and as a dependent variable in multivariate analysis. Data were entered using EpiInfo (version 6.04) and analyses were performed with

SPSS (version 11.5 statistical packages for Windows). Excel 2003 for Windows was used to plot graphs.

Konzo and cassava toxicity 46 Table II-1: Socio-demographic variables and 24hr recall food consumption of participants

among the high prevalence of konzo health area (n = 224) and the low prevalence of

konzo health area (n =263)

Variables Health area High Low OR (CI 95 %) P Sex Female 81 70 1.36 (1.05 – 1.78) 0.024 Male 142 193

Age 0.87 (0.63 – 1.20) 0.438 Under 35 49 63 35 + 145 155

Marital status Unmarried 37 26 1.69 (1.06 – 2.71) 0.030 Married 183 236

Education Illiterate 109 99 1.28 (1.05 – 1.58) 0.021 Literate 114 161

Occupation Farmer 196 178 1.26 (1.15 – 1.38) < 0.005 Other 26 76

Native No 52 50 1.22 (0.86 – 1.72) 0.266 Yes 172 213

Origin of cassava consumed in the household Own farm 215 240 1.04 (1.00 – 1.09) 0.045 Other origin 7 19

Soaking time of the cassava consumed Less than 3 nights 62 202 0.34 (0.28 – 0.43) < 0.005 3 nights + 153 38

Konzo and cassava toxicity 47

Composition of luku consumed Only cassava 192 244 0.93 (0.87 – 0.98) 0.010 Cassava + cereal 28 15

Meat consumption Yes 84 79 1.25 (0.97 – 1.60) 0.084 No 140 184

Sesame consumption Yes 5 108 1.66 (1.50 – 1.84) < 0.005 No 219 155

Cereals consumption Yes 105 34 3.63 (2.57 – 5.11) < 0.005 No 119 229

Cowpea consumption Yes 76 70 1.27 (0.97 – 1.67) 0.092 No 148 193

Vegetables consumption Yes 161 186 1.02 (0.91 – 1.14) 0.84 No 63 77

Cassava tuber consumption Yes 39 23 1.99 (1.23 – 3.23) 0.006 No 185 240

Cassava leaves “Saka-Saka” consumption Yes 91 104 1.03 (0.83 -1.28) 0.853 No 133 159

Luku consumption Yes 223 260 1.01 (0.99 – 1.02) 0.628 No 1 3

II.1.3 Results

The majority (69 %) of the interviewed heads or delegates of household were male. The mean age of respondents was 43.3 years (SD 12.0) with a range of 61 years (maximum 78 and minimum 17). 79 % of the members of the households were native from the area. The illiteracy rate among the heads or delegates of household was 43 % and among the literate,

Konzo and cassava toxicity 48 more than 50 % did not finish the primary school (sixth grade), only 1.6 % and 0.4 % have reached respectively secondary school (at least ninth grade school) and high school or university, respectively. Most of the heads of household were married (86%) and small-scale farmers (78 %) as main occupation. Other main occupation of the participants encompassed mainly sawyer (5.5 %), teacher (4.5 %), worker (2.3 %), retailer (2.1 %), roadman (1.8 %) and hunter (0.4%). The mean family size was 6.2 ± 2.7 (range 14; max 15 and min 1). About

75 % of female respondents were illiterate and almost all of them were farmer except three with one retailer and two house wives with no other occupation. Table II-1 shows the distribution of some socio-demographic variables of participants among the high prevalence of konzo health area (n = 224) and the low prevalence health area (n =263). Degree of konzo prevalence was statistically significant and associated with female gender [OR (95 % CI) =

1.4 (1.1 -1.8), P = 0.024], unmarried status OR (95 % CI) = 1.7 (1.1 -2.7), P = 0.030], illiteracy [OR (95 % CI) = 1.2 (1.1 -1.6), P = 0.021], farmer as main occupation [OR (95 %

CI) = 1.3 (1.2 -1.4), P< 0.05] and slightly with consumption of cassava originated from own farm as opposed to cassava obtained elsewhere [OR (95 % CI) = 1.0 (1.0 – 1.1), P = 0.045].

The origin of 24 hour cassava flour consumed in 93.4 % of households was from their own farmstead and the retting time of the processed cassava was less than three nights in 57,4 % of those families. In the households from where cassava originated from their own farmstead,

90.6 % of them consumed luku that was composed only of cassava flour and 9.4 % mixed their cassava flour with maize flour.

Among the 3,015 individuals in the 487 households selected, 43 konzo patients were detected in 33 (6.8 %) of the households; thus, a prevalence of 1.4% and an incidence in 2002 of 1.3 ‰

(Figure II-1). The mean number of affected family members per household was 1.30 (SD

0.6) with a range of two (minimum 1 and maximum 3). The distribution of konzo cases per

Konzo and cassava toxicity 49 selected health areas is presented in Table II-2. 77 % of patients were female (male-to-female ratio 1:3.3) and 64 % of the patients were under 15 year of age.

Figure II-1: Distribution of onset of konzo from 1980 to 2002

12

10

8

6 Cases 4

2

0 1975 1980 1985 1990 1995 2000 2005 year of onset

Table II-2: Distribution of konzo cases per health area

Health area Number of household Number of konzo patients

per number of konzo

cases

1 2 3

High prevalence area 22 4 1 33

Masina 14 2 1 21

Mutsanga 8 2 0 12

Low prevalence area 3 2 1 10

Popo-Secteur 3 1 1 8

Imwela 0 1 0 2

Total 25 6 2 43

Konzo and cassava toxicity 50 The majority [25/ 33 (75.8 %)] of the affected households had a single affected family

member. The degree of disability on walking and the age distribution of konzo patients by

gender are shown in Table II-3. The mild form of the disease was the most common and

found in 74 % of patients followed by the crawler or severe form (16%) and the moderate

form (9 %). No male was found to be attained by the moderate form. The earliest year of

onset of paralysis reported is 1980 and the latest 2002. Figure II-1 shows the distribution of

konzo cases by year of onset. Around half of the cases occurred between 1998 and 2002.

Table II-3 : Degree of disability on walking and age distribution of konzo patients by gender

Variables Sex Total

Female Male

Degree of disability

a. Walk without stick 24 8 32

b. Walk with stick 4 - 4

c. Can’t walk (crawl) 5 2 7

Total 33 10 43

Age (Years)

a. Under 5 5 2 7

b. 5 – 9 10 3 13

c. 10 – 14 4 1 5

d. 15 – 19 2 2 4

e. 20 8 2 10

Total 29 10 39

From the food consumption patterns of the respondents, we observed that less than 60 % ate

three main meals a day as breakfast, lunch and dinner, and their dietary pattern was based

Konzo and cassava toxicity 51 basically on cassava stiff porridge, the so called fufu or luku. Table II-4 presents the frequencies of 24 hour foods consumed; 97.9 % had their evening meal whereas 58.7 % of households had lunch and 91.6 % had breakfast in the morning. Luku, the cassava flour stiff porridge was the main staple food consumed at least once during the day in 99.2 % of households. Rice was consumed as staple food in 0.6 %. Maize and other cassava roots products (boiled or raw cassava roots) were the other staple foods consumed mainly during lunch time as snacks.

Table II-4: 24-hour recall of household food intake frequencies (%)

Food Morning Lunch Evening Roots and tubers Luku (cassava flour stiff porridge) 85.1 31.2 95.4 Other cassava roots products 1.6 5.6 1.2 Sweet potato - 0.2 -- Yam - 4.3 Vegetables and fruits Vegetables Saka Saka (pounded cassava leaves) 25.5 5.9 20.7 Tomato 16.6 6.5 14.5 Mushrooms 10.6 2.8 11.9 Mbondi (Salacia pynaertii) 8.8 4.1 11.1 Mfumbwa (Gnetum africanum) 4.4 1.8 7.0 Spinach 2.0 0.4 2.2 Ngayi-Ngayi (Hibiscus sabandja) 1.0 - 3.2 Matembele (Sweet potato leaves) 0.6 0.4 1.4 Kikalakasa (Psophocarpus scandens leaves) 0.4 - 0.4 Amaranth 0.2 - 0.8 Other vegetables (unspecified) 1.6 0.8 2.4 Fruits Safou 1.0 1.0 0.6 Banana 0.6 2 - Pineapple - 0.4 -

Konzo and cassava toxicity 52 Legumes Cowpeas (Vigna ungiculata) 16.6 8.0 17.1 Beans (Phaseolus vulgaris) 1.0 0.2 0.8 Voandzou (Vigna subterranea) 0.2 2.0 0.6 Cereals Maize 4.4 13.3 3.8 Rice 0.6 1 0.2 Oleaginous grains (seeds and nuts) Sesame 13.9 4.3 12.6 Peanut 12.8 17.8 11.9 Soybean 1.2 0.2 0.2 Pumpkins seeds (Curcubitaceae sp.) 0.8 1.0 1.2 Flesh Red meat (cow, pork, goat, lamb, wild animals) 8.7 4 11.1 Fish 4.7 3.2 9.4 Chicken 1.2 0.8 1.6 Insect (Caterpillar and Larva) 1 - 1.6 Milk 0.6 - - Egg 0.4 - 0.4 Wild bird (unspecified) - - 0.2 Grasshopper - - 0.2 No Food 8.4 41.3 2.1

Saka - saka (pounded cassava leaves) was the main condiment consumed as side-dish with luku in 40 % of households followed by cowpeas (30 %) and sesame (23.2 %). Peanut and tomato were used as ingredients to prepare their sauce. Peanuts also were consumed during lunch time as snacks to accompany grains of maize or other cassava roots products.

Mushrooms and unconventional green leafy vegetables such as mbondi (Salacia pynaertii) and mfumbwa (Gnetum africanum), which villagers gather from the nearby bush or forests were popular. They were consumed as supplementary foods to the staple luku in 17.7 %, 18.1

Konzo and cassava toxicity 53 % and 11.3 % of households respectively. Only 15 % of households consumed garden vegetables.

Table II-5: Seasonal food consumption availability (%) listed by the respondents

Foods Rainy season Dry season Sweet potato 3.1 4.8 Yam 12.7 36.9 Saka Saka (pounded cassava leaves) 76 7 Tomato 49.5 4.6 Mushrooms 67 4.1 Mbondi (Salacia pynaertii) 36.1 13.3 Mfumbwa (Gnetum africanum) 21.1 21.1 Spinach 30.6 1.9 Ngayi-Ngayi (Hibiscus sabandja) 15.7 1.2 Matembele (Sweet potato leaves) 9.1 1.2 Amaranth 29.6 3.1 Maize 54.5 5.2 Rice 0.4 0 Banana 24.4 5.2 Pineapple 23.2 3.1 Cowpeas (Vigna ungiculata) 50.9 9.1 Beans (Phaseolus vulgaris) 18.6 8.7 Sesame 36.4 8.9 Peanut 54.5 11 Pumpkin seeds (Curcubitaceae sp.) 5 52.1 Red meat (cow, pork, goat, lamb) 3.3 68.7 Rats & wild animals 1 78 Fish 18.4 49.5 Chicken 0.4 0.6 Insect (Caterpillar and Larva) 18 17.4 Egg 0.6 0.2 Wild bird (unspecified) 0.4 12.7 Grasshopper 1.2 36.7

Konzo and cassava toxicity 54 Meat and fish were at lesser degree consumed to supplement the staple food. Usually they are added in the preparation of vegetables. Consumption of fruits is less; safou, pineapple, banana and orange were the only fruits listed and consumed by less than 1% of households the day before the survey.

The comparison between degree of prevalence of konzo in health area and 24 hour recall food

(groups) consumption is summarised in Table II-1. There is a statistically significant association between the prevalence of konzo with consumption of cereals ([OR (95 % CI) =

3.6 (2.6 -5.1)], [OR (95 % CI) = 1.2 (1.1 -1.4)], with consumption of tubers of cassava [OR

(95 % CI) = 2.0 (1.2 -3.2)] and with consumption of sesame [OR (95 % CI) = 1.7 (1.5 -1.8)].

No statistically significant association was found between the prevalence of konzo in health area with consumption of meat, with consumption of cowpea, with consumption of vegetables, with consumption of saka-saka and, with consumption of luku. The frequencies of foods consumed by season (Table II-5) show a decrease of consumption of almost all the foods listed by the participants from the rainy season to the dry season, except the consumption of meat and pumpkin seeds, which increases consumption.

II.1.4 Discussion and conclusion

Konzo is still occurring in this area three generations after the first report (Trolli, 1938). The prevalence of konzo (1.4 %) found in our study, is lower than the expected prevalence of 4 % but in the range of that reported in the literature (Tylleskär et al, 1991; Tshala-Katumbay et al,

2001b). The overall sex and age distribution of patients in our study was similar to most of the previous studies in the region and elsewhere (Howlett, 1994; Banea- Mayambu, 1997c,

Tshala-Katumbay et al, 2001a). Preponderance of female patients (female to male ratio 3.3: 1) in our study is similar to almost all other studies except the ones carried out in Tanzania and

Mozambique where male cases were preponderant (Howlett, 1994; Tshala-Katumbay et al,

2001b).

Konzo and cassava toxicity 55 Households whose head was illiterate carried an increased risk of konzo. More than half of the heads of household or representatives were illiterate or did not go beyond primary school level. Illiteracy was also found to be a factor that carries an increased risk of paralysis in the case of neurolathyrism, a spastic paraparesis with many similarities to konzo (Lambein et al,

2004). Education drives both individual and community development, and illiterates are likely to have low socio-economic status even in such remote rural areas. Literate people probably have better access to information than the illiterate on food processing especially home detoxification methods (Getahun et al, 2002b). Women, who belong to the most susceptible group to develop konzo and who also play the primary role in the household food security, has been found in this study to be less educated and at higher risk of konzo. Three quarters of them were illiterate and among those who were literate, half did not finish the primary school education. In D. R. C., more than half of the women of rural areas are illiterate (R. D. C.,

2001). Unmarried status of the head of household as a risk factor can be explained by the excessive workloads. The main occupation as a farmer for a head of household is associated with increased risk of konzo. This can be explained by the fact that people are relying on their own sole culture and consumption of cassava. Increased risk of konzo is found to be associated with consumption of cassava originating from the household’s own farm. This may be related to poverty, as the less poor have more access to marketed commodities, which results in a more varied diet.

The food consumption pattern of the selected households in both high and low konzo incidence areas is dominated by cassava diet. Fufu, the processed product of cassava roots is the staple food of almost all the households and saka-saka is the main side-dish consumed with fufu. Cassava roots are an excellent source of carbohydrate but contain extremely low levels of protein and fat (Bradbury and Holloway, 1988). The protein is of poor quality, leucine and lysine are limiting amino acids. The proportion of methionine is low and the

Konzo and cassava toxicity 56 chemical score of the protein is around 40 (chapter III-2). Cassava leaves have a high protein content but also this protein is of poor quality limited in lysine, histidine and sulphur containing amino acids (methionine and cysteine) (chapter III-1). High dietary exposure to cyanogen was found in this area (chapter IV).

Consumption of cowpea with fufu is also popular and provided a high quantity of protein in the diet but also in cowpeas the sulphur containing amino acids are low (Finetin, 2001).

Consumption of cereals and sesame is found in this study to be protective factors against konzo. Similar protection by methionine rich cereals was also found for neurolathyrism

(Getahun et al, 2003). Mixing cassava with cereals and mixing legumes such as cowpeas with cereals may thus increase the quality of the meal by optimising the balance of essential amino acids. Consumption of the unconventional green leafy vegetable mbondi (Salacia pynaertii) and pumpkin seeds that are rich in protein with a high content in methionine and cysteine

(Mbemba and Remacle, 1992; Finetin, 2001), should also be promoted. Consumption of fresh or boiled cassava tubers is a risk factor for konzo. Cassava roots contain high level of cyanogenic glucosides that should be removed by processing. Peeling and boiling cassava are not enough to lower the toxin compounds to a safe level. The peeled roots need to be soaked in slow running water (retting or “rouissage”) for at least three nights. Consumption of luku made from roots soaked less than three nights was associated with increased incidence of konzo.

Nutritional resources listed became scarce during times of dry season. Almost all crops are rain-fed and cannot survive during the dry season unless watering or irrigation which is not done in this area because of absence of inputs or of major river systems for irrigation, electricity for irrigation pumps and other factors that limit irrigation. Cassava roots that constitute the staple food resist the drought. Under drought conditions the linamarin content of cassava roots is known to increase due to increased water stress on the cassava plant

Konzo and cassava toxicity 57 (Bokanga et al, 1994). Moreover, the weather of the dry season is mostly cloudy and the mean temperature is 20 0C or less; which may be favourable for high exposure to cyanogen. The dry season is the period of intensive cassava trade resulting in frequently shortcuts in cassava processing with high residual cyanogens level in the product. As the duration of dry season in this area of particular dry tropical climate is estimated to 5 or 6 months, serious drought may increase the cyanide intake of individuals, if non- efficient processing techniques are used, to such a degree as to precipitate the occurrence of konzo. High prevalence of konzo has been reported in the dry season (Banea-Mayambu et al, 1997a; Cardoso et al, 2004).

In conclusion, konzo is still occurring in this area with an incidence in 2002 of 1.3 ‰, where women who play the principal role in the household food security are in majority illiterate.

Although konzo was reported in this area in 1938, in this study we found no cases with onset before 1980. We found no reports on the life expectancy of konzo patients. The diet is largely dominated by cassava and major foods consumed are of poor quality in protein especially in sulphur containing amino acids. Methionine and cysteine are required for the detoxification of cyanide in the body. The results obtained in this study, confirm that low intake of sulphur containing amino acids (methionine + cysteine) is associated with incidence of konzo, as well as with dietary cyanide exposure. Therefore the emphasis should be placed on increasing production and access to cereals, sesame and pumpkin seeds to increase the availability of sulphur amino acids in the diet. Vegetable gardens should be promoted to encourage the consumption of leafy vegetables in all seasons. Appropriate information, communication and training in cassava processing and promotion of a better balanced diet may prevent this irreversible crippling disease, konzo.

Konzo and cassava toxicity 58

CHAPTER III.

CASSAVA FOOD QUALITY AND SAFETY

Konzo and cassava toxicity 59

III Cassava food quality and safety

III.1 Food Safety and Amino Acid Balance in Processed Cassava

"cossettes"*

III.1.1 Introduction

Cassava (Manihot esculenta Crantz, Euphorbiaceae) is the major staple food consumed by the population of D. R. C. (Goossens, 1996). Processed cassava roots provide more than 60 % of the daily energy intake (FAO, 1990).

Sweet varieties of cassava roots may be consumed directly while bitter varieties with high content of cyanogenic glycosides are traditionally processed to reduce toxicity and to improve palatability and storability. Many varieties of processed cassava roots with different local names are known: cossettes, chikwangue, fufu, malemba, luku, ntuka, etc.

Cossettes, which is the most popular cassava product in D. R. C., is obtained by soaking or immersing fresh bitter cassava roots (whole or peeled) in a stream or stationary water (near a stream) for at least 3 days to allow them to ferment until they become soft. The fermented roots are then taken out, peeled and sundried on mats, racks, roofs of houses, etc. Depending on the weather, sundrying takes 2-5 days (Hahn, 1989). The dried fermented cassava root is the so- called "cossettes" (Figure III-1). This form of cassava product is preferred because it can be stored for a long period and can be traded over much longer distances (Goossens, 1996; Minten and Kyle, 1999).

* This sub- chapter has been published as: Delphin Diasolua Ngudi, Yu – Haey Kuo and Fernand Lambein (2002). Food safety and amino acid balance in processed cassava roots “cossettes”. Journal of Agricultural and Food Chemistry 50, 3042 – 3049.

Konzo and cassava toxicity 60

Figure III-1: Flow diagram of cassava cossettes processing

Steps of cossettes production Location

Harvesting of whole fresh bitter cassava Fields

roots

Peeling and chopping River, stream or village

or not

Soaking and natural fermentation in water River, stream or village

for 3-5 days (and peeling)

Sun-drying or fire-drying Village

for 3-5 days

Cossettes Village

Storage and packaging in Pounding or Village jute or propyl-ethylene milling sacks

Cassava flour

Trade Local markets or city markets

Konzo and cassava toxicity 61

When the roots are soaked and dried for a shorter period because of insufficient food supply or poor agro-ecological conditions, the remaining cyanogen content can be much higher than that after normal process. High intakes of dietary cyanogens from poorly processed cassava roots in a diet deficient in sulphur amino acids have been implicated in the causation of konzo (Tylleskär,

1994b).

D. R. C. is the most affected country where konzo has been reported from remote rural areas of

Bulungu, Kahemba, Masi-Manimba and Popokabaka in Bandundu province (Howlett, 1994;

Tylleskär et al, 1995).

Besides the high content of cyanogens, cassava roots are also known to be poor in protein content

(Hahn, 1989). Proteins are a necessary part of the daily diet because the human body does not store protein as it does with carbohydrate and fats. Furthermore, 9 of the 20 protein amino acids are either not synthesized at all by our body or can only be synthesized in insufficient amounts.

Humans must obtain them from dietary sources. These are known as the dietary essential amino acids that include histidine, isoleucine, leucine, tryptophan, lysine, methionine, phenylalanine, threonine and valine. Failure to receive an adequate dietary supply of essential amino acids leads to retarded growth and development in children and to disease and body deterioration in adults

(McMury and Castellion, 1996).

The objective of this study is to determine residual cyanogen in different samples of cossettes to check the safety, to quantify the daily intake of cyanogen and to estimate the amount of sulphur amino acids required for their detoxification. Free and total protein amino acids profiles of cossettes are determined to evaluate the dietary protein quality and to compare with the amino acid requirements of children and adults. Nonprotein amino acids have been reported to be present in many commonly eaten foods and these compounds have the ability to interfere with a wide range of fundamental biochemical processes and cause disease (Rozan et al, 2000;

Konzo and cassava toxicity 62

Rubenstein, 2000). Neurolathyrism, which shares clinical similarities with konzo, has been associated with the overconsumption of grass pea (Lathyrus sativus L., Fabaceae) which contains a neurotoxic nonprotein amino acid, BOAA or its synonym ODAP (Howlett, 1994; Tylleskär et al, 1994c; Getahun et al, 1999). Therefore the presence of any inherent potentially toxic nonprotein amino acid in cossette samples is also examined.

III.1.2 Materials and methods

III.1.2.1 Plant Materials

Cossettes were purchased in five different markets (Ngaba, Lemba, Livulu, Rond Point and

Matete) of Kinshasa, capital of D. R. C. The cossettes in those markets are supplied by Bandundu province where konzo has been reported and depending on the size, they are sold in bulk of about

10 pieces of roots. About 500g (2 or 3 pieces) of the cossettes from each market were finely ground with an electric small laboratory grinder "Culatti" with 200 µm sieve prior to sampling and analyses.

Cassava flour from Cameroon was purchased from an exotic food shop in Antwerp, Belgium for comparison. Cameroon is a part of central Africa where cassava is processed like in D. R. C.

III.1.2.2 Determination of Cyanogens

A simple picrate paper kit developed by Egan et al (1997) and improved by Bradbury et al (1999) was used for the determination of all forms of cyanogens in cassava products. Protocol B1 was followed for the determination of total cyanogens and acetone cyanohydrin + HCN/ CN-.

Konzo and cassava toxicity 63

III.1.2.2.1 Total Cyanogens

100 mg of sample was placed on top of 21 mm diameter Whatman 3 MM filter paper disc containing 1 M phosphate buffer at pH 8 and linamarase in a flat bottomed plastic bottle

(supplied in the kit). Millipore filtered deionised water (0.5 mL) was added and a yellow picrate paper attached to a plastic strip was immediately inserted into the vial that was closed immediately with a screw lid and allowed to stand at room temperature for 24 h. The plastic backing sheet was removed carefully from the picrate paper. This latter paper was immersed in

5.0 ml of deionised water for about 30 min. The absorbance of the solution was measured at 510 nm, using cuvettes of 1 cm light path against a blank, which contained a yellow solution produced by a picrate paper not exposed to HCN/ CN-.

The total cyanogens content (expressed in ppm) was calculated by the simple equation:

Total cyanogens content = 396 x Absorbance

Other samples were prepared as above but without cassava flour, using square linamarin papers equivalent to 50 and 400 ppm, to serve as controls.

III.1.2.2.2 Acetone Cyanohydrin + HCN/ CN-

This analysis was done following the above procedure. However 200 mg of guanidine hydrochloride was added after the addition of the phosphate buffer pH 8 filter paper disc. The incubation time was 3 h.

The amount of linamarin was calculated through the following equation:

Linamarin content = Total cyanogens - (acetone cyanohydrin + HCN/ CN-)

Konzo and cassava toxicity 64

III.1.2.3 Determination of Amino acids

An HPLC gradient system with precolumn phenylisothiocyanate (PITC) derivatisation (Khan et al, 1994) was used to analyse free amino acids. Total protein amino acids were determined after sample hydrolysis.

III.1.2.3.1 Extraction of free amino acids

50 μL of DL- Allylglycine (100 nmol/ ml) was added to finely ground sample (5 g) as internal standard. The samples were then extracted in 3 volumes of 70 % ethanol and stored overnight at

4o C. The extracts were centrifuged (34800g, 20 min) and the pellets were washed twice with 70

% ethanol. The supernatants were pooled and concentrated under vacuum and stored in a deep freezer at - 20 o C.

III.1.2.3.2 Sample hydrolysis

The flour sample was hydrolysed under vacuum in 6 M HCl following the AOAC 982.30 E procedure (18).

III.1.2.3.3 Amino acid analysis

Aliquots of extract or hydrolysate were concentrated and dried under vacuum (37 o C, 20 mm Hg) then a coupling reagent (methanol: water: triethylamine; 2:2:1; v/v) was added, mixed and dried immediately under vacuum during 10 min. After this, PITC reagent (methanol: triethylamine: water: PITC; 7:1:1:1; v/v) was added and allowed to stand at room temperature for 20 min before drying under vacuum. PITC derivatives were dissolved in buffer A (0.1 M ammonium acetate, pH 6.5) and filtered through a 0.22 μm Millipore membrane.

Konzo and cassava toxicity 65

20 μl of sample was injected into an HPLC (Waters model 991 equipped with photodiode array detector) using a gradient system of buffer A (100 - 0 %) and buffer B (0.1 M ammonium acetate containing acetonitrile and methanol, 44:46:10; v/v, pH 6.5) (0 - 100 % after 50 min). The operating temperature was 43 o C. A reverse phase column from Alltech (Alltima C 18, 5 μm,

250 x 4.6 mm) was used. The absorbance at 254 nm was recorded and used for calculations. A standard protein amino acid mixture (food hydrolysate A 9656, Sigma) was derivatised as above and used as standard for calculations. The results were analysed by Millennium software (Waters, version 1.10)

III.1.2.3.4 Tryptophan Determination

A rapid and simple acid ninhydrin method described by Gaitonde and Dovey (1970) and adapted for colorimetric determination of tryptophan by Sodek et al (1975) was used. Cossettes samples were partially defatted by suspension in 20 volumes of acetone and stirring occasionally for 30’.

After filtration, the powder was left to air-dry. Portions (500 mg) of defatted cassava cossettes were extracted in a centrifuge tube with 2.0 ml of 70 % ethanol for 30’ at room temperature. The mixture was occasionally stirred and homogenized with a glass rod. 5 mL of NaOH (0.5 %) were then added and extraction continued for another 1 h. After centrifugation, a clear supernatant was collected and 0.2 ml of it was taken for tryptophan assay. The acid ninhydrin method using reagent b (250 mg of ninhydrin dissolved in 10 ml of formic acid- hydrochloric acid; 3:2; v/v) was followed for the determination of tryptophan in the samples. Readings were made against a reagent blank in a spectrophotometer (Shidmazu, UV-1601) at 390 nm using cuvettes of 1 cm light path. Sample blanks contained a similar aliquot of extract together with reagent b without ninhydrin.

Konzo and cassava toxicity 66

After subtracting the absorbance value of the sample blank, the tryptophan content was read off a standard curve. Lysozyme (Grade I from egg white; Sigma Chemical Co.) was used to construct the standard curve. Tryptophan values obtained from this graph were then corrected for tyrosine interference according to Zahnley and Davis (1973).

III.1.2.4 Statistics

The results were computed and compared by analysis of variance using the software package

SPSS 9.0 for Windows. Significant differences amongst means were confirmed using the Tukey

Honestly Significant Differences set at 95 % confidence interval (P < 0.05). Data are expressed as means ± standard deviation.

III.1.3 Results and discussion

III.1.3.1 Cyanogens

The six samples of cassava cossettes had residual cyanogens below 10 mg HCN equivalent kg-1 as shown in Table III-1. This is the recommended safe limit by the Codex alimentarius (FAO/

WHO, 1991). The highest level was found in samples from Cameroon (9.37 mg HCN equivalent kg-1) and the lowest level in samples from Rond Point (1.45 mg HCN equivalent kg-1), showing a

6.5 fold variation with a significant difference between samples from Cameroon and all other samples (P < 0.05). No significant differences was found between samples from Matete and

Ngaba, and among samples from Rond Point, Livulu and Lemba but those last samples were significantly different from those from Matete and Ngaba (P < 0.05).

Enzymatic determination of the cyanogenic glycoside linamarin, the major source of cyanide in cassava, showed a variation of almost ten fold between 0.924 and 8.58 mg HCN equivalent kg-1.

Konzo and cassava toxicity 67

Table III-1: Cyanogens content in cassava cossettes (mg HCN equivalent kg - 1 dry weight)*

Cossettes Total Cyanogens Acetone Cyanohydrin + HCN/ Linamarin CN- Matete 2.772 b ± 0.396 0.264 a ±0.280 2.508 c ±0.457 (N = 3) Cameroon 9.372 c ± 0.229 0.792 a ± 0.280 8.580 d ± 0.229 (N = 3) Lemba 1.716 a ± 0.229 0.396 a ± 0.457 1.320 a, b ± 0.229 (N = 3) RondPoint 1.452 a ± 0.229 0.528 a ± 0.229 0.924 a ± 0.229 (N = 3) Livulu 1.584 a ± 0.396 0.132 a ± 0.229 1.452 a, b, c± 0.229 (N = 3) Ngaba 2.904 b ± 0.457 0.792 a ± 0.280 2.112 b, c ± 0.243 (N = 3)

Again the samples from Cameroon were significantly higher than all other samples (P < 0.05) while the linamarin content of Rond Point was significantly different from Ngaba and Matete, and also Lemba was different from Matete (P < 0.05). No significant difference was found for the content of acetone cyanohydrin + HCN/ CN - between samples (P > 0.05), this varied 6 fold between 0.13 and 0.79 mg HCN kg-1 in the cassava cossettes examined.

The fresh bitter cassava roots typically used in the region have total cyanogen levels of 100 to

500 mg HCN equivalent kg-1 root, even up to 1500 mg HCN kg-1 (Bradbury and Holloway, 1988;

O’Brien et al, 1992; Padmaja, 1996). Although the original content of the fresh roots from which the cossettes were prepared is not known, it is obvious that the processing and handling of the material resulted in a reduction of total cyanogen of at least 10-30 fold, up to 150-500 fold,

* Values are means ± standard deviation a, b, c same superscript within a column means no significant difference (P> 0.05)

Konzo and cassava toxicity 68 giving a final result within the recommended safe limit set at 10 mg HCN equivalent per kg of dry weight. The processing and handling included soaking, sundrying, storage and transportation to the open markets in Kinshasa where the cossettes are sold in jute sack or in bulk. The samples from Cameroon bought in Europe were packed in a plastic foil of low permeability.

The low levels of glycosides in the flour from cossettes can be explained by continued cell desintegration and enzymatic activity of the linamarase from the cytoplasm hydrolysing the cyanogens from the disrupted vacuoles during soaking and throughout the four days of drying before moisture fell to low levels in these big root pieces (Banea-Mayambu, 1997). When considering time and temperature factor, it can be assumed that even in short soaked cossettes cyanohydrins might be lost during storage and transportation over much longer distances from

Bandundu to the markets in Kinshasa than when consumed locally in Bandundu area. The low cyanogen exposure from cassava might explain the absence of cases of konzo in urban consumers, while the crippling disease konzo is prevalent in remote rural areas of Bandundu

Province (Minten and Kyle, 1999; Formunyam et al, 1985; Banea-Mayambu et al, 1998). Even the shortcut processed cassava products from Bandundu area sold in Kinshasa do not cause clinical symptoms of cyanide exposure (Banea-Mayambu et al, 1998).

Oke (1968) reported HCN contents of 1.0 mg/ 100 g in cossettes from D. R. C. and O'Brien et al

(1992) found a variation in cyanogens content of fermented cassava roots ranging between 0 to

11.3 mg kg-1 in villages of Cameroon. In populations with cassava roots as their main staple food, a basic daily energy need of 6276 kJ (1500 kcal) can be satisfied with 500 g dry weight cassava root products. Adult consumers would then be exposed to approximately less than 5 mg HCN equivalent per day comparing to the Codex alimentarius safe level of 10 mg HCN equivalent per kg dry weight (Rosling, 1988). If cossettes as staple food provide 60 % of dietary daily energy

Konzo and cassava toxicity 69

intake in D. R. C., calculated from the FAO/ WHO energy requirements (FAO/WHO/UNU,

1985), it means that for the samples from Matete about 0.7 mg HCN equivalent is present in 241

g of cassava cossettes to be consumed daily by children (1 to 3 years old) and about 1.5 mg HCN

equivalent is present in 532 g of cassava cossettes to be consumed daily by a moderately active

adult man to meet energy requirements of 3414 (816) and 7531 (1800) kJ (kcal) respectively

(Table III-2).

Table III-2: Estimated daily cossettes and total cyanogens intake

Subjects Daily 60 % daily Daily Daily cyanogens from

energy energy cossettes cossettes‡ (mg)

required* required in intake† (g) Matete Livulu

in kJ (kcal) kJ (kcal) samples samples

Children 1 – 3 yr 5690 3414 241 0.7 0.4

(1360) (816)

Children 7 – 9 yr 9162 5497 389 1.1 0.6

(2190) (1314)

Adult female 9204 5523 390 1.1 0.6

(moderately (2200) (1320) active)

Adult male 12552 7531 532 1.5 0.8

(moderately (3000) (1800) active)

* from FAO/ WHO/ UNU (1985) † 100 g cassava provides 338 kcal (FAO/WHO/UNU, 1985) ‡ Total cyanogens (Table III-1) x daily cossettes intake (Table III-2)

Konzo and cassava toxicity 70

III.1.3.2 Total protein amino acids

Table III-3 represents the total protein amino acids profiles in cassava cossette samples. The overall average of the total protein amino acids is 23.7 mg/ g dry weight cassava cossettes in which essential amino acids represent 54.6 % and the sulphur containing amino acids 9.7 %.

Alanine was the major protein amino acid in all samples except in the samples from Rond Point where glutamic acid was the most important. This finding is in agreement with some studies done with fresh cassava roots (Bradbury and Holloway, 1988; Firmin and Kamenan, 1996; Glew et al,

1997). This suggests that during the post-harvest processing practised, the loss of protein is negligeable while the loss of cyanogens is considerable.

The samples from Livulu had the highest total protein amino acid content (27.3 mg/ g of dry weight cassava cossettes) with 54.2 % of essential amino acids. The samples from Lemba contained the highest essential amino acids proportion (63.5 %) and those from Rond Point were the lowest (50.4 %). Leucine and lysine, the purely ketogenic amino acids, were the limiting amino acids in our samples (Table III-4). Leucine was the first limiting amino acid in the samples from Livulu and Ngaba with an amino acid score of 0.36 and 0.45 respectively. Lysine was the first limiting amino acid in the other samples with an amino acid score varying from 0.35 to 0.47.

Results of Firmin and Kamenan (1996) showed sulphur amino acids (methionine + cysteine) in fresh cassava roots and leucine in fermented pulp of cassava roots as first limiting amino acid, respectively. Yeoh and Truong (1996) found sulphur amino acids, leucine and lysine to be limiting amino acid in different cultivars of cassava roots studied. Bradbury and Holloway (1988) reported large differences in amino acid composition between different cultivars of cassava roots examined and there was no essential amino acid, which was clearly the first limiting amino acid.

Nevertheless, on the average histidine was the first and leucine the second limiting amino acid.

Konzo and cassava toxicity 71

Although sulphur amino acids were not the limiting amino acids in our samples, we should notice that the proportion of methionine represented in average only 13 % (4.7 - 16.2 %) and cysteine 87

% (83.8 - 95.3 %) of total sulphur amino acids. Methionine normally supplies part of the body's needs for cysteine. With cysteine-free diets, methionine can supply all of the body's needs for cysteine. Cysteine can spare methionine and a certain proportion of dietary methionine is converted to cysteine (Brody, 1994).

III.1.3.3 Free amino acids

The free amino acid pattern of cossettes samples is summarised in Table III-5. The concentrations of free amino acids were in general very low. Arginine and sulphur amino acids

(methionine and cysteine) were not found. Histidine was not found in in the samples from

Matete, Rond Point and Ngaba. No asparagine was detected in the samples fromMatete and

Livulu.

The samples from Livulu showed the highest amount in total free amino acids (6.2 mg/ g of dry weight cossettes) and those from Rond Point, the lowest (0.27 mg/ g of dry weight cossettes).

This represents about 23 fold variation among the few samples examined; duration and flow rate of water during soaking leading to leaching out can probably explain this finding. Threonine was quantitatively the most important free amino acid in five of the samples examined, while in the samples from Rond Point sample asparagine was the most abundant (Figure III-2). Alanine ranked the second place except in the samples from Livulu, Rond Point and Cameroon. No known potentially toxic nonprotein amino acids were detected in our samples.

Konzo and cassava toxicity 72

Table III-3: Total protein amino acids content in cassava cossettes (mg g - 1dry weight)*

Total Cossettes protein Matete Cameroon Lemba RondPoint Livulu Ngaba amino acids (N = 4) (N = 4) (N = 4) (N = 4) (N = 3) (N = 4) Asp 00.200a ± 0.022 00.234 a ± 0.044 00.351 a b ± 0.074 00.480 b c ± 0.120 00.403 b ± 0.049 00.564 c ± 0.041 Glu 1.074a ± 0.257 1.255 a ± 0.079 1.343 a ± 0.209 22.016 b ± 0.186 1.702 a b ± 0.589 22.270 b ± 0.291 Ser 00.700a ± 0.157 00.688 a ± 0.058 00.749 a ± 0.112 1.053 b ±0.023 00.970 a b ± 0.272 00.691 a ± 0.066 Gly 1.342 a ± 0.175 1.254 a ± 0.231` 1.423 a b ± 0.164 1.921 b ± 0.173 1.388 a ± 0.413 1.217 a ± 0.111 His 1.218 a b ± 0.177 1.744 b ± 0.140 1.082 a ± 0.038 1.203 a ± 0.113 1.111 a ± 0.580 1.060 a ± 0.121 Arg 00.796 a ± 0.073 00.715 a ± 0.024 00.744 a ± 0.085 1.029 b ± 0.071 1.034 b ± 0.089 00.796 a ± 0.092 Thr 1.787a b ± 0.638 22.239 a b c ± 0.71 3.177 b c ± 0.807 00.945 a ± 0.335 3.508 c ± 0.907 3.493 c ± 0.594 Ala 5.124 c ± 0.869 44.309 b c ± 0.675 3.349 b ± 0.334 00.962 a ± 0.059 5.329 c ± 0.968 5.628 c ± 0.316 Pro 1.494 a b ± 0.072 1.345 a ± 0.154 1.664a b ± 0.269 1.930 b ± 0.264 1.679 a b ± 0.269 1.498 a b ± 0.110 Tyr 1.410 a ± 0.064 1.272 a ± 0.190 1.035 a ± 0.506 1.343 a ± 0.186 1.035 a ± 0.506 1.191 a ± 0.346 Val 00.961 a ± 0.287 1.153 a ± 0.485 3.306 b ± 0.363 1.104 a ± 0.449 1.733 a ± 0.783 1.374 a ± 0.343 Met 00.162 a ± 0.028 00.108 a ± 0.029 00.399 a ± 0.079 00.372 a ± 0.110 00.302 a ± 0.368 00.398 a ± 0.243 Cys 1.828 a ± 0.070 22.183 a ± 0.231 22.157 a ± 0.168 1.975 a ± 0.440 1.853 a ± 0.250 22.061 a ± 0.999 Ile 00.740 a b ± 0.148 00.554 a ± 0.237 1.604 b± 0.556 00.609 a ± 0.303 00.612 a ±0.241 00.914 a b ± 0.539 Leu 00.660 a ± 0.108 00.648 a ± 0.261 1.717 b ± 0.692 00.699 a ± 0.527 00.652 a ± 0.112 00.760 a ± 0.321 Phe 00.900a ± 0.104 00.783 a ±0.234 00.842 a ± 0.063 00.708 a ± 0.202 22.314 b ± 0.568 00.523 a ± 0.363 Lys 00.497 a ± 0.123 00.527 a ± 0.092 00.531 a ± 0.256 00.530 a ± 0.108 00.751 a ± 0.124 00.691 a ± 0.164 Try 00.741 a ± 0.004 00.754 a ± 0.022 00.877 a ± 0.188 00.780 a ± 0.026 00.907 a ± 0.018 00.740 a ± 0.012

* Values are means ± standard deviation a, b, c same superscript within a row means no significant difference (P> 0.05)

Konzo and cassava toxicity 73

Table III-4: Amino acid scoring pattern of different cossette samples

Essential Amino Acid FAO/ WHO1* Amino acid Scores2†

(EAA) Children Matete Lemba Livulu Ngaba RondPoint Cameroon

(2-5 years)

Threonine 34 2.43 3.54 3.78 3.97 1.42 3.05

Cysteine + Methionine 25 3.68 3.88 3.16 3.80 4.80 4.24

Valine 35 1.27 3.58 1.81 1.52 1.61 1.66

Isoleucine 28 1.22 2.17 0.80 1.26 1.11 0.91

Leucine 66 0.46 0.99 0.36 0.45 0.54 0.45

Tyrosine + Phenylalanine 63 1.69 1.14 1.95 1.05 1.66 1.37

Histidine 19 2.97 2.16 2.14 2.16 3.24 4.25

Lysine 58 0.40 0.35 0.47 0.47 0.47 0.37

Tryptophan 11 3.11 3.02 3.02 2.60 3.62 3.17

First limiting Amino Acid Lysine Lysine Leucine Leucine Lysine Lysine

Second limiting Amino Acid Leucine Leucine Lysine Lysine Leucine Leucine

* Recommended amino acid scoring pattern from FAO/ WHO/ UNU (1985) † Amino acid score = mg of amino acid in 1g of test protein per mg of amino acid in 1 g of reference Protein (FAO/WHO, 1991)

Konzo and cassava toxicity 74

Figure III-2: Free amino acids in cassava cossette samples

70,0

60,0

50,0

Matete 40,0 Cameroon Lemba RondPoint

30,0 Livulu Ngaba

20,0 % of total free amino acids in different cossette samples 10,0

0,0 Asp Glu Ser Gly His Arg Thr Ala Pro Tyr Val Met Cys Ile Leu Phe Lys Asn Gln Trp Free amino acids

Konzo and cassava toxicity 75

Table III-5. Free protein amino acids content in cassava cossettes (mg g - 1 dry weight)*

Free protein Cossettes amino acids Matete Cameroon Lemba RondPoint Livulu Ngaba (N = 4) (N = 3) (N = 3) (N = 3) (N = 3) (N = 3) Aspartic acid 0.005a ± 0.003 0.012 b ± 0.000 0.030 c ± 0.001 ND 0.051 d ± 0.005 0.044 d ± 0.007 Glutamic acid 0.013a ± 0.000 0.079 d ± 0.003 0.048 c ± 0.001 0.012 a ± 0.001 0.082 d ± 0.006 0.025 b ± 0.001 Serine 0.002a ± 0.000 0.008 a b ± 0.005 0.015 b c ± 0.003 0.003a ±0.002 0.029 d ± 0.001 0.023 c d ± 0.007 Glycine 0.007 a b ± 0.000 0.016 b c ± 0.003` 0.025 c ± 0.003 0.006 a ± 0.003 0.051 d ± 0.002 0.005 a b ± 0.009 Histidine ND∗ 0.017 a ± 0.000 0.048 a ± 0.050 ND 0.245 b ± 0.015 ND Arginine ND ND ND ND ND ND Threonine 0.370 a ± 0.119 1.106 b ± 0.056 1.135 b ± 0.212 0.072 a ± 0.005 2.171 c ± 0.344 1.523 b ± 0.264 Alanine 0.046 a ± 0.029 0.151 a b ± 0.016 0.223 c ± 0.047 0.025 a ± 0.000 0.284 b ± 0.116 0.270 b ± 0.065 Proline 0.039 b ± 0.002 0.064 c ± 0.007 0.098 e ± 0.004 0.017 a ± 0.000 0.298 f ± 0.008 0.081 d ± 0.005 Tyrosine 0.014 b ± 0.002 0.031 c ± 0.002 0.077 d ± 0.001 0.004 a ± 0.000 0.145 e ± 0.005 0.087 d ± 0.007 Valine 0.025a b± 0.000 0.043a b ± 0.001 0.076 b c ± 0.043 0.010 a ± 0.001 0.124 c± 0.010 0.056 a b ± 0.000 Methionine ND ND ND ND ND ND Cysteine ND ND ND ND ND ND Isoleucine 0.012 b ± 0.000 0.016 b ± 0.001 0.070 d ± 0.005 0.003 a ± 0.002 0.107 e ±0.006 0.027 c ± 0.003 Leucine 0.023 a ± 0.001 0.044 a ± 0.001 0.139 a b ± 0.002 0.007 a ± 0.000 0.541 b ± 0.000 0.081 a b ± 0.003 Phenylalanine 0.015 a b ± 0.000 0.043 a b ±0.000 0.089 c ± 0.004 0.006 a ± 0.004 1.925 d ± 0.060 0.075 b c ± 0.003 Lysine 0.010 b ± 0.000 0.024 c ± 0.002 0.036 d ±0.003 0.005 a ± 0.003 0.112 e ± 0.003 0.018 c ± 0.002 Asparagine ND 0.003 a ± 0.000 0.002 a ± 0.003 0.096 a ± 0.083 ND 0.010 b ± 0.001 Glutamine 0.006 a b ± 0.000 0.024 c d ± 0.000 0.015 b c ± 0.003 0.002 a ± 0.001 0.032 d ± 0.002 0.160 e ± 0.010 Tryptophan 0.027 a b ± 0.000 0.221 c ± 0.011 0.038 a b ± 0.026 0.006 a ± 0.003 0.005 a ± 0.004 0.063 b ± 0.008

* Values are means ± standard deviation a, b, c , d, e same superscript within a row means no significant difference (P> 0.05) ∗ ND: not detected

Konzo and cassava toxicity 76

III.1.3.4 Essential Amino Acid (EAA) requirements and estimated daily

intake

Rose and Wixon (1955) demonstrated the influence of cysteine on the methionine requirement for an adult man by determining the conditions that supported a zero or slightly positive nitrogen

(N) balance. They observed that cysteine alone without methionine resulted in a negative N balance. A near zero N balance was observed with a diet containing 0.8 g of methionine, while N balance was negative with 0.7 g methionine diet. Higher levels of methionine resulted in a positive N balance. They concluded that oversupply of cysteine could give a positive N balance with lower intake of methionine, but even then the intake of methionine remains essential. This statement illustrates the limiting of the ability of cysteine to spare methionine. Although cysteine can fulfill a large fraction of our requirement for sulphur amino acids, according to Altman and

Dittmer (1974) in the combination cysteine + methionine, 30 - 50 % of total requirement for adults may be furnished by cysteine and 50 - 70 % furnished by methionine.

The expected daily methionine and sulphur amino acids intake provided by cassava cossettes consumption, which in the case of D. R. C. represents 60 % of daily energy intake, are compared with the suggested amino acid patterns requirement (Table III-6). It can be concluded that children of 1 to 9 years old cannot expect to meet methionine requirement whereas adults can meet sulphur amino acid requirement. Sulphur amino acids are required for cyanide detoxification in the human body (Bradbury and Holloway, 1988; Rosling, 1994). A daily supply of about 1.2 mg of dietary sulfur from S-containing amino acids is needed by the human body to detoxify 1.0 mg of HCN (Padmaja, 1996). When the body is regularly exposed to cassava cyanogens the increased synthesis of rhodanese, enzyme responsible for cyanide detoxification in

Konzo and cassava toxicity 77 the human body by forming thiocyanate, makes extra demands on the body's reserves of sulphur amino acids. If this demand is prolonged as in the regular consumption of cassava root insufficiently processed, and the diet is inadequate, the synthesis of many proteins vital for bodily

Table III-6. Essential Amino Acid (EAA) requirements and estimated daily intake

EAA suggested patterns of Estimated daily EAA (from cossettes) intake† (mg)

requirement*

(mg AA/ day) Matete samples Livulu samples

Child Adult Adult Child Child Adult Adult Child Child Adult Adult

> 1yr female male 1-3 yr 7-9 yr female male 1-3 yr 7-9yr female male

Thr 1000 305 500 431 695 697 951 845 1365 1368 1867

Cys + Met - 550 1100 - - 776 1059 - - 840 1146

Met 800 - - 39 63 - - 73 117 - -

Val 900 650 800 232 374 375 511 418 674 676 922

Ile 1000 450 700 178 288 289 394 147 238 239 326

Leu 1500 620 1100 159 257 257 351 157 254 254 347

Tyr+ Phe - 1120 1100 - - 901 1229 - - 1306 1782

Phe 800 - - 217 350 - - 558 900 - -

Lys 1600 500 800 120 193 194 264 181 292 293 400

Trp 250 157 250 179 288 289 394 219 353 354 483 functions may be impaired and lead to the development of protein deficiencies and other diseases

(Padmaja, 1996; Onwuka et al, 1992; Tor-Agbidye et al, 1998). Other food components of the diet should contribute to a better balanced amino acid composition of the diet, especially the level of sulphur amino acids. In the case of lathyrism, a neurodegenerative disease with similar clinical

* From Altman and Dittmer (1974) † Daily cossettes intake (Table III-2) x Amino acid (Table III-3)

Konzo and cassava toxicity 78 symptoms as konzo, Lambein et al (2001) have suggested that the ratio of cereals (rich in methionine) to Lathyrus seeds (rich in lysine and low in sulphur amino acid) may be a determining factor in the etiology. In the regions neighbouring the konzo-affected areas in

Bandundu where traditionally corn or millet flour is mixed with cassava as staple food, no cases of konzo have been reported. This may corroborate our views as to the importance of methionine for a healthy balanced diet.

Hence, the recommended daily methionine allowance should be reconsidered and given separately from total S-amino acid requirement.

III.1.3.5 Conclusion

The processed cassava roots available on the markets in Kinshasa have cyanogens content within the safe limit recommended by FAO/ WHO. Proper processing, time and storage conditions and traditional transport in jute sacks appear to contribute to reduce residual cyanogens in the cossettes whereas insufficient processing and transport in airtight wrapping which prevents the release of cyanide can probably explain the level of cyanogen found in the cossettes from

Cameroon samples.

No potentially toxic nonprotein amino acids were detected in this study.

The dietary requirements for sulphur amino acids need to be adjusted for the loss caused by cyanide detoxification. The total sulphur amino acids availability does not give a correct value for the requirement of the essential amino acid methionine. In the case when cassava is taken as staple food, the low methionine content may aggravate the risk for cyanide toxicity and konzo disease, even when the cysteine present covers the dietary requirement for sulphur amino acids.

Konzo and cassava toxicity 79

III.2 Residual cyanogens, free and total amino acid profiles of cooked

cassava leaves "saka- saka” *†

III.2.1 Introduction

In the konzo-affected areas of D. R. C., processed cassava roots are prepared as described before

(chapter III-1), while cassava leaves or "saka-saka" are prepared as follows: the hard petioles are removed, the tender leaves and the shoots are selected and may be blanched in warm/ boiled water for a few minutes or partially dried on a pan or a pot over fire and then squeezed to remove liquid before pounding. The spinach-like mass obtained after pounding with a traditional wooden mortar and pestle is then cooked with some water added. Usually palm oil and salt are added and sometimes also traditional spices and onion. (CEPLANUT, 1988; Almazan and Theberge, 1989;

Hahn, 1989).

Reports on the nutritional quality of cassava leaf protein as food are scanty and conflicting

(Bokanga, 1994). The majority of studies considered cassava leaf as animal feed and focused mainly on cyanogen removal. Residual cyanogens and the presence of inherent potentially toxic nonprotein amino acids were examined in this study before and after cooking pounded cassava leaves to check their safety. The aim of this paper is also to assess the amino acid profiles and the protein quality of cooked pounded cassava leaves as food, which is the most common daily side dish as sauce and as main source of protein in a diet consisting of processed cassava roots as the exclusive staple food in konzo affected areas of DRC, especially in Bandundu province.

* This sub-chapter has been published as: Delphin Diasolua Ngudi, Yu – Haey Kuo and Fernand Lambein (2003). Cassava cyanogens and free amino acids in raw and cooked leaves. Food and Chemical Toxicology 41, 1193 – 1197. † Delphin Diasolua Ngudi, Hu – Haey Kuo and Fernand Lambein (2003). Amino acid profiles and protein quality of cooked cassava leaves or “saka saka”. Journal of the Science of Food and Agriculture 83, 529 – 534.

Konzo and cassava toxicity 80

III.2.2 Materials and methods

III.2.2.1 Sample acquisition

Deep-frozen pounded raw cassava leaves from D. R. C., about 500 to 600 g packed in plastic foil, were purchased in five different exotic food shops in Ghent, Belgium (Dampoort, Foreign and

Ghana) and in Paris, France (Congo and Chateau).

III.2.2.2 Sample handling and culinary processing

Each packet of raw sample was divided into two parts. One part was kept as such for analysis and the other part was subjected to the following culinary treatment on a hot plate:

About 250 ml of water was added to 100 g of raw pounded cassava leaves and allowed to boil. 10 ml of palm oil and about 1 g of salt were added when boiling started and the dish was stirred with a wooden spoon for mixing of the ingredients. The cooked pounded cassava leaves or "saka- saka" were ready to eat after 30 minutes of boiling (CEPLANUT, 1988). The samples were analysed after cooling down to room temperature.

III.2.2.3 Determination of cyanogens

See section III.1.2.2

III.2.2.4 Determination of protein

The samples were extracted in 3 vol of physiological solution (NaCl 0.15 M; pH 5.96 at room temperature) and stored overnight at 4 oC. The extracts were centrifuged (34800g, 20 min) and the pellets were washed twice with physiological solution. The supernatants were pooled and used for protein analyses.

Konzo and cassava toxicity 81

The Bio-Rad® Protein assay kit, consisting of dye reagent concentrate and lyophilised bovine albumin as protein standard was used to determine protein content in our samples. A standard curve was made using several dilutions of protein standard containing 0.2 to 1.4 mg ml –1.

Analyses of protein were done as follows: 0.1 ml of sample was placed in a test tube and then 5 ml of diluted dye reagent was added and mixed several times. The absorbance was measured at

595 nm versus reagent blank within a period of 5 minutes to one hour after mixing. The absorbance was converted to protein content using the standard curve.

III.2.2.5 Determination of amino acids

See section III.1.2.3

III.2.2.6 Tryptophan determination

See sub-section III.1.2.3.4

III.2.2.7 Protein quality evaluation

The amino acid scoring pattern recommended by FAO/WHO/UNU (1985) was used for the evaluation of dietary protein quality as follows:

Amino acid score = mg of amino acid in 1 g of test protein 18 mg of amino acid in 1g of reference protein

The essential amino acid showing a score less than 1 was a limiting amino acid. The lowest amino acid score (the most limiting amino acid) indicates the quality of the protein.

Konzo and cassava toxicity 82

III.2.2.8 Statistics

The software package SPSS 10.0 for windows was used for the analysis of variance of the data.

The statistically significant differences among means were confirmed using the Tukey Honestly significant differences at 95 % confidence interval (P<0.05).

III.2.3 Results and discussion

III.2.3.1 Total cyanogens

The total cyanogens content of the raw and the cooked cassava leaves samples are summarised in

Table III-9. The initial levels of total cyanogens in the raw (pounded) cassava leaves samples ranged from 35.9 ± 0.4 to 107.5 ± 0.8 mg HCN equivalent kg–1, acetone cyanohydrin + HCN/

CN- from 5.7 ± 1.9 to 24.1 ± 4.5 mg HCN equivalent kg–1 and the linamarin from 30.2 ± 2.4 to

83.4 ±5.3 mg HCN equivalent kg–1 dry weight. Those values are up to 10- fold higher than what we detected in the processed cassava roots (Chapter III-1). A Significant reduction (P< 0.05) in total cyanogens was observed when the raw samples were cooked; 96 - 99 % of the total cyanogens were removed after cooking the cassava leaves. Bokanga (1994) observed that pounding alone reduced the cyanogenic potential by about 60 - 70 %.

After cooking, the total cyanogens varied from 0.30 ± 0.04 to 1.9 ± 0.2 mg HCN equivalent kg-1 dry weight and the acetone cyanohydrin + HCN/CN- were not detected. The residual cyanogens were below the recommended safe limit set at 10 mg HCN equivalent kg-1 by the Codex alimentarius (FAO/ WHO, 1991).

Although the original content and the varieties of the fresh leaves from which the samples were pounded are not known, this decrease can be explained by the following considerations. Besides the genetic differences of the plant varieties, the variation in cyanogen content between samples

Konzo and cassava toxicity 83 can also be explained by the maturity of the leaves. Padmaja (1989) reported lower contents of cyanide in the older leaves compared with young leaves. Cyanogen content may also depend on the heat treatment during preparation of the leaves before pounding. The leaves may be washed with tap water or blanched in warm water for a few minutes or partially dried over fire or grinding before pounding (Almazan and Theberge, 1989). The heat treatment can be favorable for cyanide reduction or can destroy the endogenous hydrolysing enzyme linamarase. Finally, the consistency of pounding can also play a role in the reduction of cyanogens during pounding of cassava leaves. Destruction of the cells leads to contact between the cyanogenic glucosides and the endogenous enzyme with subsequent release of HCN. Our finding is in agreement with the fact that the rapid removal of cyanogens from cassava can be attributed to the heat applied during boiling which accelerates the evaporation of HCN and cyanohydrin produced by the linamarin hydrolysis (Almazan and Theberge, 1989; Essers, 1989; Bokanga, 1994).

Konzo and cassava toxicity 84

Table III-7: Cyanogen content in raw and cooked cassava leaves (mg HCN equivalent kg-1 dry weight)

Dampoort Foreign Congo Chateau Ghana

Raw Cooked Raw Cooked Raw Cooked Raw Cooked Raw Cooked

(n=4) (n=4) (n=4) (n=4) (n=4) (n=4) (n=4) (n=4) (n=4) (n=4)

Total 35.9b±0.4 1.3a±0.3 107.5d±0.8 0.8a±1.1 87.9c±4.5 0.30a±0.04 83.7c±2.4 0.7a±0.1 86.1c±3.1 1.9a±0.2 cyanogens

5.7ab±1.9 NDb 24.2d±4.5 ND 15.3c±0.3 ND 10.5bc±2.5 ND 15.0c±3.9 ND

Acetone cyanhydrin

+HCN/CN-

b a d a c a c a c a Linamarin 30.2 ±2.4 1.3 ±0.3 83.4 ±5.3 0.8 ±1.1 72.7 ±4.2 0.30 ±0.04 73.2 ±4.9 00.7 ±0.1 71.1 ±0.8 1.9 ±0.2

Same letter within a row means no significant difference (P>0.05)

a Values are means ± standard deviation

b Not detected.

Konzo and cassava toxicity 85

III.2.3.2 Total protein and amino acid profiles

The total protein content and the amino acids composition (g kg-1 dry weight) of five different samples of cassava leaves before and after cooking are listed in Table III-7. The total protein content of the raw pounded cassava leaves samples ranged from 235.8 g in the Foreign sample to

351.8 g in the Dampoort sample. Those values are within the ranges reported in the literature

(Almazan and Theberge, 1989; Hahn, 1989; Bokanga, 1994; Yeoh and Chew, 1976; Ravindran and Ravindran, 1988). Ravindran and Ravindran (1988) observed a decrease of protein content in cassava leaves with ageing: from 381 g kg-1 in very young cassava leaves to 286 g kg-1 in young leaves and 197 g kg-1 in mature leaves. The total protein content in the cooked samples ranged from 111.8 g kg-1 dry weight in the Chateau sample to 144.6 g kg-1 dry weight in the Dampoort sample. The results showed a significant (P < 0.05) decrease by an average of 58% in protein content after cooking pounded cassava leaves. The large volume of water added and the prolonged cooking time (at least 30 min of boiling) necessary for this culinary treatment to remove the bitter taste, which might lead to losses of amino acids by diffusion and by thermal degradation, can explain this decrease (Clemente and al, 1998; De la Cruz, 1999). Cooking of green beans in a covered pot or pressure cooker was observed to cause important losses of amino acids compared to the raw sample (De la Cruz et al, 1999). Other authors also observed a significant reduction of amino acids in chickpea seeds after cooking with distilled water (Attia et al, 1994; Clemente et al, 1998). During preparation of food, the side chains of some protein- bound amino acids can react chemically with each other or with other molecules present in the food and those reactions can result in a reduction of nutritive value (Sherr et al, 1989).

Konzo and cassava toxicity 86

Table III-8: Protein content and amino acid composition of raw and cooked pounded cassava leaves (g kg-1 dry weight)*

Amino acids Foreign (n= 5) Chateau (n= 5) Congo (n= 4) Dampoort (n= 5) Ghana (n= 4) Raw Cooked Raw cooked Raw Cooked Raw cooked Raw cooked Aspartic acid 40.8c±0.1 12.1a±1.0 33.9b±0.4 17.0a±2.5 33.9b±0.4 16.3a±3.9 37.1bc±3.2 16.0a±2.6 38.6bc±2.8 13.6a±0.4 Glutamic Acid 36.8c±0.3 12.8a±2.4 36.2c±0.7 16.7ab±3.2 37.9c±4.1 16.9ab±4.0 44.4d±4.1 19.6b±3.0 41.1cd±0.2 15.5ab±0.6 Serine 11.8b±2.1 3.9a±1.0 12.9bc±2.3 5.3a±1.6 15.1bc±2.5 4.6a±0.7 16.8c±1.7 6.0a±0.3 15.0bc±0.3 4.7a ±1.1 Glycine 11.0c±0.2 5.2a±1.2 11.5cd±0.5 5.9a±0.8 13.7de±1.3 5.5a±1.0 18.0f±1.7 8.2b±1.1 13.9e±0.5 6.3ab±0.3 Histidine 5.4b±0.9 1.2a±0.2 6.7cd±0.5 1.3a±0.7 6.6bc±0.6 1.6a±0.2 7.9de±0.7 2.0a±0.2 8.1e±0.8 1.6a±0.3 Arginine 16.8b±0.6 8.5a±1.2 16.1b±0.5 9.3a±1.8 18.4bc±2.2 7.9a±1.0 24.0d±2.4 10.6a±1.5 21.4cd±0.8 8.8a±0.6 Threonine 8.8c±0.6 4.2a±0.6 9.6c±0.4 4.7ab±1.1 11.7d±1.1 4.4ab±0.1 17.5e±0.7 6.3b±0.6 12.8d±0.6 4.5ab±0.3 Alanine 21.7bc±1.9 13.6a±3.4 20.7bc±1.9 13.9a±1.9 25.1cd±3.0 14.4a±1.1 30.9d±4.2 15.9ab±3.7 25.4cd±1.7 15.8ab±3.2 Proline 11.3b±0.4 6.4a±1.3 11.6b±0.5 6.7a±1.0 14.4c ±1.8 6.3a±0.4 17.9d±2.1 8.1a±1.1 14.9c±0.3 6.3a±0.4 Tyrosine 9.4c±0.3 4.3a±0.5 10.1c±0.4 4.7a±0.8 12.1d±1.4 4.9a±0.1 15.4e±1.5 6.6b±0.5 12.2d±0.4 4.3a±0.3 Valine 10.6c±0.2 4.9ab±1.1 11.5cd±0.4 4.8ab±0.9 12.9de±1.5 4.6a±0.5 16.8f±1.4 6.6b±0.4 14.8e±1.0 4.7ab±0.2 Methionine 3.2c ±0.4 0.3a±0.1 3.1c±0.5 ND† 3.4c±0.2 0.8ab±0.3 4.7d±0.6 1.5b±0.4 2.8c±0.4 0.2a±0.2 Cysteine ND ND ND ND ND ND ND ND ND ND Isoleucine 7.5b±0.4 2.3a±0.9 7.8bc±0.3 2.6a±1.0 9.0bc±1.1 2.5a±0.8 11.8d±1.2 4.1a±0.8 9.8c±0.8 2.5a±1.0 Leucine 14.3c±0.3 6.3a±1.4 15.9cd±0.4 7.1ab±1.1 17.6de±1.8 5.9a±0.7 24.0f±2.3 9.7b±0.3 18.6e±0.9 7.0a±0.2 Phenylalanine 12.4b±1.8 5.4a±1.2 13.4b±0.3 5.9a±2.7 15.0b±2.0 4.3a±0.8 18.7c±1.1 6.9a±0.6 14.9b±0.8 6.0a±1.1 Tryptophan 3.5 c±0.0 1.4a±0.0 3.5 c±0.0 1.4 a±0.1 3.8 d±0.0 1.3 a±0.0 4.1 e±0.2 1.7 b±0.0 3.6 cd±0.1 1.6 b±0.0

Lysine 9.6c±0.9 4.1a±1.1 15.3d±0.7 3.7a±1.3 15.5d±1.1 5.6ab±0.3 20.5de±2.0 7.1bc±0.6 18.3e±2.2 5.1ab±0.7

Total Protein 235.8c±5.6 114.1 a±2.4 256.7 d±7.3 111.8 a±2.4 293.3e±2.7 113.1a±3.2 351.8f±7.3 144.6b±3.5 291.7e±5.3 112.2a±1.1

* Values are means ± standard deviation a,b,c,d,e,f same superscript within a row means no significant difference (P>0.05) † ND= not detected

Konzo and cassava toxicity 87

Aspartic acid, glutamic acid and alanine were the major amino acids found in all the samples

studied. They represented together an overall average of 101 g and 46 g kg-1 dry weight in the

raw samples and in the cooked samples respectively. Histidine, tryptophan and methionine were

the amino acid found in lowest concentration with together an overall average of 14 g and 3.6 g

kg-1 dry weight in the raw and in the cooked samples respectively. Other authors obtained similar

profiles in all the varieties of raw cassava leaf studied (Yeoh and Chew, 1976; Ravindran and

Ravindran, 1998). Cysteine was not detected in any sample.

Figure III-3: Protein amino acids profile of the raw and cooked cassava leaves

160

140 Raw

cooked 120

100

80

60 g amino acid per kg total protein kg total per acid g amino

40

20

0 Asp Glu Ser Gly His Arg Thr Ala Pro Tyr Val Met Ile Leu Phe Try Lys amino acids

Figure III-3 shows the profile of the individual amino acid per total protein, which is almost

similar for all the samples. No marked differences can be observed between the raw and the

cooked samples when considering individual amino acid, except alanine which increased

significantly after cooking and methionine showing the highest decrease after cooking.

Konzo and cassava toxicity 88

Methionine is highly required in the konzo-affected areas for dietary cyanide detoxification. The reactive thioether group in methionine that is involved in oxido-reduction reactions and the thermal breakdown of methionine can explain this finding (Clemente et al, 1998; De la Cruz et al,

1999). Excessive heat treatment causes considerable nutritional damage to methionine (Geervani and Theophilus, 1980; Shemer and Perkins, 1975).

A comparison of the total essential amino acid profiles with the FAO/WHO reference pattern

(Table III-8) showed that the raw cassava leaves samples contained 357 g to 401 g of total essential amino acids per kg of cassava leaves protein. This is higher than the 339 g of total essential amino acids in the recommended FAO/WHO-reference protein. The cooked samples contained less total essential amino acids than the FAO/WHO reference ranging from 299.3 g kg-

1 total protein content in the Foreign samples to 330.2 g kg-1 total protein content in the Ghana samples

Konzo and cassava toxicity 89

Table III-9: Comparison of the essential amino acid contents of different raw and cooked pounded cassava leaves samples and their

amino acid score with the recommended FAO reference

Essential AA FAO Foreign Chateau Congo Dampoort Ghana Ref.* Raw Cooked Raw cooked Raw Cooked Raw cooked Raw cooked Histidine 19 (1.0) 23 † (1.2) 11 (0.6) 26 (1.4) 12 (0.6) 23 (1.3) 14 (0.8) 22 (1.2) 14 (0.7) 28 (1.5) 14 (0.7) Threonine 34 (1.0) 37 (1.1) 37 (1.1) 37 (1.1) 42 (1.2) 40 (1.2) 40 (1.2) 50 (1.5) 44 (1.3) 44 (1.3) 40 (1.2) AAA‡ 63 (1.0) 92 (1.5) 85 (1.3) 91 (1.4) 95 (1.5) 92 (1.5) 81 (1.3) 97 (1.5) 94 (1.5) 93 (1.5) 91 (1.4) Valine 35 (1.0) 44 (1.3) 43 (1.2) 45 (1.3) 43 (1.2) 44(1.3) 41 (1.2) 48 (1.4) 46 (1.3) 51 (1.5) 42 (1.2) SAA 25 (1.0) 13(0.5) 0.3 (0.01) 12 (0.5) ND§ 12 (0.5) 7 (0.3) 13 (0.5) 10 (0.4) 10 (0.4) 0.2 (0.01) Isoleucine 28 (1.0) 31 (1.1) 20 (0.7) 30 (1.1) 23 (0.8) 31 (1.1) 22 (0.8) 33 (1.2) 28 (1.0) 34 (1.2) 22 (0.8) Leucine 66 (1.0) 61 (0.9) 55 (0.8) 62 (0.9) 63 (0.9) 60 (0.9) 52 (0.8) 68 (1.0) 67 (1.0) 64 (1.0) 62 (0.9) Tryptophan 11(1.0) 15(1.4) 12(1.1) 14 (1.3) 12(1.1) 13(1.2) 12(1.1) 12(1.1) 12(1.1) 12(1.1) 14(1.3) Lysine 58 (1.0) 41 (0.7) 36 (0.6) 60 (1.0) 33 (0.6) 53 (0.9) 50 (0.9) 58 (1.0) 49 (0.8) 63 (1.1) 45 (0.8) Total 339 357 299.3 377 323 368 319 401 364 399 330.2 1st limiting amino acid SAA SAA SAA SAA SAA SAA SAA SAA SAA SAA 2nd limiting amino acid Lysine Histidine Leucine Lysine Leucine Histidine Histidine Histidine Other limiting amino Leucine Lysine Histidine Lysine Leucine Lysine Isoleucine acids Isoleucine Isoleucine Isoleucine Lysine Leucine Leucine Lysine Leucine

* FAO-protein reference from FAO/ WHO/ UNU (1985) † Calculated from Table III-3: Essential amino acid per total protein, the results are expressed in g kg -1 of the protein, the amino acid scores are indicated between brackets ‡ AAA= aromatic amino acids (Phenylalanine +Tyrosine) § ND= not detected

Konzo and cassava toxicity 90

The scoring pattern in Table III-8 showed that sulphur-containing amino acids were the most limiting amino acid in all the samples (raw and cooked) with an amino acid score varying from

0.5 to less. Other investigators obtained the same results from their raw cassava leaves samples studied (Yeoh and Chew, 1976; Lancaster, 1983). Besides sulphur amino acids, lysine and leucine were limiting amino acids in the raw Foreign and Congo samples and leucine in the raw

Chateau samples. The maturity of the cassava leaves (young or mature) used and preliminary processing (blanching or partial grinding or not) before pounding were unknown and might have an effect on lysine and leucine. In all the cooked samples except the Chateau sample, histidine was the second limiting amino acid. Lysine, leucine and isoleucine were also second or third limiting amino acids in some of the cooked samples.

III.2.3.3 Free amino acids

The free amino acids and trigonelline (N-methyl-nicotinic acid) pattern of the raw and cooked cassava leaves samples are summarised in Table III-10. The total free amino acids detected and trigonelline varied from 10.8 g kg-1 to 38.2 g kg-1 in the raw samples and from 7.4 g kg-1 to 25.6 g kg-1 in the cooked samples. Thus, pounded cassava leaves showed a decrease in the total free amino acids content after cooking. The highest decrease was observed in the Ghana samples

(45.4 %) followed by the Foreign samples (38.3%), the Dampoort samples (31.3 %), the Congo samples (23.5%) and the Chateau samples (15 %). The concentration of total free amino acids including free protein amino acids and free nonprotein amino acids was at least 6-fold higher in the leaves than what we found in the cassava roots (Chapter III-1). Aspartic acid, glutamic acid and alanine are the major free protein amino acids found in the samples. Methionine and cysteine were not detected as free amino acid in any samples. γ-Amino butyric acid (GABA) and α-amino butyric acid (α- ABA) are the free nonprotein amino acids detected in all the samples.

Konzo and cassava toxicity 91

Table III-10: Free amino acid and trigonelline content in raw and cooked cassava leaves (g kg-1 dry weight)*† Amino acids Dampoort (n= 4) Foreign (n= 4) Congo (n= 4) Chateau (n= 4) Ghana (n= 4) Raw Cooked Raw cooked Raw Cooked Raw cooked Raw cooked Aspartic acid 1.16b±0.44 0.48a±0.07 4.57e±0.24 2.23 c±0.01 1.34b±0.23 0.94a,b±0.16 3.53d±0.38 3.16d±0.23 1.98c±0.15 1.09 b±0.08 Glutamic Acid 0.21a±0.02 0.38a±0.18 4.44e±0.22 2.1 0 b,c±0.02 2.51c,d±0.13 1.92 b±0.27 5.53f,g±0.21 5.10 f±0.30 5.72f,g±0.39 2.73d±0.24 Serine 0.18a±0.01 0.13a±0.01 1.77e±0.11 1.02c±0.01 0.50 b±0.01 0.43b±0.00 1.75e±.0.08 1.87e±0.19 3.02f±0.16 1.39d±0.02 Glycine 0.28d±0.02 0.13b,c±0.02 0.18c±0.02 0.10a,b±0.00 0.07a,b±0.02 0.05a,b±0.03 0.10a,b±0.00 0.09a,b±0.06 0.52e±0.01 0.26d±0.00 Histidine ND‡ ND 0.87e±0.11 0.51c±0.03 0.19 b±0.01 0.21 b±0.03 0.72 d±0.11 0.87e±0.07 1.07f±0.06 0.53c±0.01 Arginine 0.40a±0.03 0.17a±0.02 2.55 d±0.28 1.52c±0.06 0.25a±0.01 0.26a±0.01 1.14b±0.19 1.32b,c±0.16 2.75d±0.16 1.25b,c±0.00 Threonine ND ND 0.15a±0.09 0.02a±0.00 0.24a±0.00 0.25a±0.02 0.89b±0.19 1.08b±0.19 1.01b±0.03 0.26a±0.26 Alanine 1.64a,b±0.15 1.05a±0.12 1.65a,b±0.10 0.70a±0.01 2.09b±1.47 0.79a±0.09 1.73a,b±0.12 2.59 b,c±1.22 3.67c±0.14 1.54a,b±0.01 Proline 0.51d±0.06 0.41b,c±0.02 0.48c,d±0.04 0.24a±0.00 0.37 b,c±0.06 0.31a,b±0.01 0.72e±0.05 0.64e±0.04 2.46g±0.09 1.01f±0.00 Tyrosine 0.42 b±0.03 0.27a±0.01 0.85c,d±0.09 0.45b±0.00 0.42b±0.06 0.39a,b±0.00 0.97d±0.06 0.89d±0.02 1.46e±0.10 0.72c±0.03 Valine 0.86b±0.10 0.44a±0.01 1.48c±0.08 0.85b±0.04 0.77b±0.05 0.73b±0.01 1.88d±0.08 1.74d±0.10 2.81e±0.06 1.40c±0.03 Methionine ND ND ND ND ND ND ND ND ND ND Cysteine ND ND ND ND ND ND ND ND ND ND Isoleucine 0.44±0.04 0.26±0.00 0.69±0.03 0.35±0.00 0.39±0.38 0.21±0.00 0.53±0.02 0.45±0.02 1.42±0.04 0.68±0.00 Leucine 0.65±0.04 0.37±0.01 0.75±0.04 0.43±0.00 0.24±0.02 0.26±0.00 0.57±0.03 0.51±0.02 2.21±0.08 1.12±0.00 Phenylalanine 0.51±0.05 0.39±0.01 2.09±0.21 1.10±0.01 0.71±0.14 0.67±0.10 1.44±0.33 1.38±0.15 2.39±0.07 1.16±0.00 Tryptophan 0.19±0.022 0.21±0.02 1.46±0.08 0.84±0.02 0.59±0.03 0.57±0.07 1.37±0.25 1.35±0.13 1.67±0.07 1.07±0.01 Lysine 0.28±0.02 0.07±0.00 0.55±0.07 0.45±0.00 0.12±0.01 0.18±0.02 0.33±0.04 0.41±0.07 0.98±0.06 0.67±0.02 Trigonelline 0.81±0.10 0.94±0.13 0.29±0.02 0.41±0.03 0.37±0.03 0.35±0.00 1.33±0.13 0.44±0.05 0.32±0.02 0.49±0.02 Asparagine 0.18±0.02 0.16±0.00 1.49±0.10 2.25±0.02 0.32±0.01 0.27±0.03 3.51±0.20 1.14±0.11 1.39±0.08 1.81±0.00 Glutamine 0.23±0.02 0.39±0.02 0.78±0.05 0.71±0.00 0.22±0.01 0.11±0.01 1.75±0.11 0.44±0.03 0.67±0.03 0.54±0.00 GABA 1.69±0.17 1.07±0.08 0.18±0.05 0.40±0.09 0.05±0.00 0.05±0.00 0.03±0.07 0.01±0.00 0.52±0.03 0.91±0.18 α-ABA 0.19±0.02 0.09±0.01 0.10±0.01 0.20±0.00 0.06±0.00 0.07±0.00 0.28±0.02 0.09±0.00 0.14±0.00 0.20±0.00

* Values are means ± standard deviation, a,b,c,d,e Same superscript within a row means no significant difference (P>0.05) ND= Not detected; GABA = γ-Amino butyric acid; α- ABA= α-Amino butyric acid

Konzo and cassava toxicity 92

GABA is a major constituent in higher plants and its physiological function in the plant is suggested to be involved in pH regulation, nitrogen storage, plant development and defense, as well as a compatible osmolyte and an alternative pathway for glutamate utilisation (Shelp et al.,

1999). GABA is a major inhibitory neurotransmitter in mammalian brain and alterations in

GABAergic function have been postulated to underlie seizure pathogenesis (Goldsmith et al.,

1990). Trigonelline, which is not an amino acid but a multifunctional natural plant hormone, was also found in all the samples. The toxicological effects of trigonelline have not been studied but considering its multiple effects in the plant, there is a need to study its potential effect on human health (Rozan et al., 2000).

III.2.4 Conclusions

In konzo affected areas, cassava leaves can contribute to the total uptake of cyanide in the diet besides the cassava roots. There is no electricity or gas available and in general cooking is done in the evening when the mothers are tired after hard work and long walking from the field. The availability of firewood, time to cook and duration of cooking can contribute as factors to higher dietary exposure to cyanogens in those regions.

All the raw samples had high protein content and high essential amino acids compared to the recommended FAO/ WHO pattern but limiting in sulphur amino acids, in lysine and leucine.

Cooking lowered the protein content of the raw pounded cassava leaves studied from 285.9 g kg-1 dry weight to 119.2 g kg-1 dry weight on average, but it is still relatively high comparing to other vegetables. Quantitatively, the cooked cassava leaves can almost fulfil the recommended daily protein intake (FAO/WHO/UNU, 1985). This can be illustrated by the comparison between the average daily consumption of cassava leaves in DRC estimated at 500 g, thus an average of about

60 g of protein (from our samples), and the recommended safe level of daily protein intake in

Konzo and cassava toxicity 93 terms of protein qualities which is 48 g or 62 g for food protein quality of score 0.6 for adult woman and adult man, respectively (Lancaster, 1983; FAO/WHO/UNU, 1985). Unfortunately, our results showed that the cooked-pounded cassava leaves were deficient in at least 3 essential amino acids (sulphur amino acids, histidine and lysine) and thus of poor quality (Friedman,

1996). Therefore the consumption of cassava leaves as the main source of dietary protein cannot compensate the methionine deficiency in konzo-affected areas where the dietary requirement for methionine needs to be adjusted for the loss caused by cyanide detoxification (Diasolua Ngudi,

2002). This dietary methionine requirement may be further increased if the leaves are also not properly cooked because of high level of cyanide in the fresh leaves (Lancaster, 1983). Cereals and legumes should be promoted as sources of sulphur amino acids and lysine respectively to improve protein quality of the diet of the poor population at risk for konzo and thus to prevent konzo and malnutrition.

It has been suggested that the deficiency in sulphur amino acids in unbalanced diets could be a contributing factor in the etiology of neuro-toxico-nutritional diseases such as konzo and neurolathyrism (Lambein et al, 2001). Neither ODAP nor other known potentially toxic nonprotein amino acid was detected in our samples. One major peak found in our samples with elution time of 37.6 min in HPLC analysis and absorption maximum at 265.4 nm after PITC derivatisation was not identified.

Better information and education especially of those preparing the food could be a relatively cheap and sustainable intervention. Considering the level of socio-economic impact of such diseases, such intervention would save resources to those communities.

Konzo and cassava toxicity 94

CHAPTER IV

DIETARY CYANOGEN AND SULPHUR METABOLITES

EXCRETION*

* This chapter has been submitted for publication in Food and Chemical Toxicology as: Delphin Diasolua Ngudi, Yu – Haey Kuo, Fernand Lambein and Patrick Kolsteren. High risk of dietary cyanogen exposure in a population living in a konzo – affected area of Democratic Republic of Congo.

Konzo and cassava toxicity 95

IV Dietary cyanogen and sulphur metabolites excretion

IV.1.1 Introduction

Urinary amino acid excretion is an important tool for the diagnosis and clinical management of disturbances of amino acid metabolism. Common indications for urine amino acid testing include clinical presentations such as neurological deterioration, hyperammonemia, kidney stones, metabolic acidosis, failure to thrive, inborn errors of amino acid metabolism, etc

(Bezkorovainy & Rafelson, 1996, Venta, 2001). The alteration in urinary excretion is principally a reflection of changes that occur in plasma amino acid composition since the concentrations of the free amino acid in urine seem to be mainly related to protein intake

(Pavy et al, 1988, Brand et al. 1997).

Urinary thiocyanate is commonly used to check cyanogen overload in a population using cassava roots and cassava products as staple food (Haque & Bradbury, 1999, Ernesto et al.

2002a). The level of thiocyanate normally present in body fluids is low but increases with chronic exposure to cyanide and with smoking habits (Vesey et al. 1999, Kussendrager and

Van Hooijdonk, 2000). Thiocyanate remains the most useful chemical biomarker for dietary cyanogen intake because it is a very stable metabolite that can be determined with relatively cheap, specific and sensitive methods (Rosling, 1994, Ressler and Tatake, 2001).

Taurine (2-amino-ethyl sulphonic acid) is an ubiquitous free amino acid highly abundant in excitable tissues, including the heart and brain. In addition to functioning as a neuroprotectant, antioxidant, osmoregulator and Ca2+ modulator, taurine may function as an inhibitory neuromodulator and neurotransmitter in the central nervous system. It is an end product from the catabolism of sulphur amino acids methionine and cysteine, and it is excreted almost entirely in urine (Laube et al. 2002, Olive, 2002, Hou et al. 2003). The

Konzo and cassava toxicity 96 urinary levels of taurine have been proposed as a potential biochemical marker of total body protein status or of sulphur amino acids catabolism (Waterfield et al. 1995, Hou et al. 2003).

In this study, we compared the level of total cyanogen in the sampled cassava flour to the recommended FAO/ WHO safe limit. We measured urinary thiocyanate to check cyanogen overload in the selected community. The potential relationship between urinary taurine and urinary thiocyanate, biomarker of daily cyanogen, was assessed and cases of konzo were detected.

IV.1.2 Material and methods

IV.1.2.1 Subjects

Samples of cassava flour and urine were obtained and examined from about one tenth of the participants selected randomly in an epidemiological study (Chapter I) we carried out in

February 2003 in Popokabaka rural health zone (Prhz), province of Bandundu (1° - 8° South;

16° – 20° East), Democratic Republic of Congo (D. R. C.). Three health areas were chosen in cooperation with the chief medical doctor and the nurse supervisor of Prhz based on the number of reported konzo cases: Popo-secteur (low prevalence area), Mutsanga (medium or moderate prevalence area) and Masina (high prevalence area). After informed oral consent, forty two heads of household or their delegates (11 females and 31 males; age 46 ± 12 yr, range 20 – 76 yrs) among the participants of the above mentioned study were randomly selected to provide samples. After the interview, each selected participant received two empty plastic vials; one to fill up with cassava flour of the evening meal and the other with the first morning urine of the next day. Konzo affected-households were registered and the patients were checked for confirmation. Filled vials were collected without addition of any preservative early in the morning of next day. Twelve participants did not return the vial with cassava flour because they either had no evening meal or did not prepare cassava flour for the

Konzo and cassava toxicity 97 evening. The collected samples were transported from the field to the laboratory at 4° C.

Samples were stored at –20 °C until assayed.

IV.1.2.2 Analytical methods

IV.1.2.2.1 Total cyanogen in cassava flour

See chapter III 1.2.2

IV.1.2.2.2 Urine sample

Urinary thiocyanate

Protocol D1 of the picrate kit method developed by Haque & Bradbury (1999) was used to determine thiocyanate in the urine. The thiocyanate content in ppm was calculated by the equation:

thiocyanate content (ppm) = 78 x absorbance

The thiocyanate content in µmol/l was obtained by multiplying the thiocyanate content in ppm by 17.2. Blank and controls were prepared as described using water instead of urine and standard paper disc loaded with thiocyanate of 68.8 or 688 µmol/l solution (4 ppm or 40 ppm), respectively. The absorbance was measured at 510 nm, using a spectrophotometer

(Shidmazu, UV 1601).

Urinary taurine

Taurine was analysed by high performance liquid chromatography as described for the other amino acids in chapter I.

IV.1.2.3 Statistics

The data showed skewed distributions, therefore median and intervals are presented as the observed ranges of total cyanogens, urinary thiocyanate and taurine. The Spearman’s rho statistics were used for bivariate correlations to measure the association between two

Konzo and cassava toxicity 98 variables; correlation was significant at the 0.01 level or at the 0.05 level (two-tailed). Results were computed using Microsoft Windows Excel 2003 and statistical analyses of the data were carried out using the software package SPSS 11.5 for Windows.

IV.1.3 Results

Out of 42 participants, 21 % were living in a household affected by at least one konzo case

(Table IV-1). No konzo cases were reported in the low prevalence area while in other areas,

13 cases of konzo were reported, from which 9 patients were in the high prevalence area and

4 other patients in the moderate prevalence area. Confirmation of the diagnosis was done by applying the WHO criteria for konzo (WHO, 1996): a visible symmetric spastic abnormality when walking and/or running, a history of abrupt onset (< 1 week), a non- progressive course in a formerly healthy person, showing bilaterally exaggerated knee and/or ankle jerks without signs of spinal disease.

Table IV-1: Distribution of konzo- affected households in each health area with the number

of konzo patients given in brackets

Number of affected household

Health area (Number of konzo patient)

1 2 3 Total

Low prevalence 0 0 0 0

Moderate prevalence 2 (2) 1 (2) 0 3 (4)

High prevalence 4 (4) 1 (2) 1 (3) 6 (9)

Total 6 (6) 2 (4) 1(3) 9 (13)

Cyanogen content in cassava flour, thiocyanate and taurine content in the urine samples are summarised in Table IV-2. Cyanogens were not detected in 26.7 % of samples and 46.7 % of cassava flour samples had total cyanogens below 10 ppm (µg HCN equivalent/g cassava

Konzo and cassava toxicity 99 flour), the WHO/FAO recommended safe limit (FAO/ WHO, 1991). There is a large variation in cyanogen content, values ranged from 2.90 to 169.75 ppm with a median of 16.50 ppm.

One sample had cyanogens content above 100 ppm. This highest concentration was found in the low prevalence area where 60 % of cassava flour samples contained total cyanogens within the WHO/FAO recommended safe limit. In the moderate area, only 22.2 % of cassava samples had total cyanogens within the WHO/FAO recommended safe limit compare to 54.5

% found in the high prevalence area.

Table IV-2: Total cyanogens in cassava flour, thiocyanate and taurine in urine samples

collected in three konzo prevalence areas of Popokabaka (DRC).

Konzo prevalence area Low Moderate High Total Total Cyanogens n 10 9 11 30 (µg HCN equivalent/ g cassava flour) Median 19.00 22.44 10.69 16.50 Min 9.11 2.90 4.75 2.90 Max 169.75 62.83 54.12 169.75 Urinary thiocyanate (µmol/ l) n 12 14 16 42 Median 279.28 400.91 287.10 300.74 Min 6.26 41.59 21.02 6.26 Max 675.27 1037.06 1101.00 1101.00 Taurine (mmol/ mol creatinine) n 12 14 16 42 Median 6.47 4.29 13.57 8.84 Min 0.00 0.00 0.00 0.00 Max 23.90 97.59 41.55 97.59

The urinary thiocyanate content ranged from 6.26 to 1101 µmol/l urine. The lowest concentration (6.26 µmol thiocyanate/l urine) was found in the low prevalence area while the highest concentration (1101 µmol thiocyanate/l urine) was found in the high prevalence area and was excreted by a participant from a konzo-affected household. 69 % of the urine samples had thiocyanate content above 172 µmol/l urine (10 ppm) with 13.8 % of them above values

Konzo and cassava toxicity 100 of 900 µmol. All the konzo-affected household participants of the moderate area excreted high urine concentration of thiocyanate (more than 500 µmol thiocyanate/l urine), while in the high prevalence area only one participant had excessively high values.

The urine concentrations of taurine were low. Except the one sample with the highest concentration (97.59 mmol/mol creatinine), the taurine concentrations of all other samples ranged between 0 (or not detected) to 41.55 mmol/mol creatinine. More than half (61.9 %) of the urine samples were below the reference limits (13 to 534 mmol taurine/mol creatinine) calculated by Venta (2001). The highest concentration of taurine was found in a sample from the moderate prevalence area. Taurine was detected in only 83.3 % of urine samples among which 71.4 % excreted thiocyanate above 10 ppm or 172 μmol/ l urine. Urinary taurine was slightly or not correlated to urinary thiocyanate (R2= 0.017, P = 0.415).

IV.1.4 Discussion

The identification of cases of konzo in the moderate and high prevalence areas in the present study shows that this crippling neurodegenerative disease is still occurring in this part of

Bandundu province from where the first cases were reported three generations ago by Trolli

(1938). The study found that a high proportion of cassava flour samples contain total cyanogens above the recommended safe limit set at 10 µg HCN equivalent/g by the Codex alimentarius (FAO/ WHO, 1991). Even when compared to the higher acceptable limit (40 ppm) used in Indonesia, that is one of the highest cassava producer and consumer countries worldwide (Djazuli & Bradbury, 1999), 16.6 % of our cassava flour samples are still above this limit. The values in this study were higher than the ones we previously reported (Diasolua

Ngudi et al. 2002) on the processed cassava roots available on the markets of Kinshasa, the capital of D. R. C., where important quantities of cassava cossettes coming from the study areas are sold. Therefore, there is a risk of dietary exposure to cyanogen from consumption of

Konzo and cassava toxicity 101 cassava flour in the areas studied. Essers et al. (1998) stated that the cyanogen content in cassava flour in rural areas of Africa usually grossly exceeds the safety limit set by the Codex alimentarius but toxic effects are rare under normal conditions. Partly because the cyanogens are mainly bound in glucosides which are relatively stable in the human body, and the form in which the product is consumed (stiff paste) causes a slow release of the toxicant which can then be detoxified gradually and more effectively by the body’s defence mechanism. Konzo has been reported to occur most frequently when the mean cyanide content of cassava flour exceeds 100 ppm (Lawrence, 1999). Shortcuts in the processing of cassava roots have been reported to result in high residual levels of cyanogen substances and the consumption of such roots leads to dietary cyanogen exposure (Banea et al. 1992).

Cyanogen exposure from cassava roots is the essential risk factor for konzo and thiocyanate levels remains the best indicator of daily cyanide intake (Rosling, 1994, Banea-Mayambu et al. 2000). An additional source of dietary cyanogen exposure can come from consumption of cassava leaves which is the main source of protein in a diet consisting of processed cassava roots in those konzo-affected areas (Chapter III). The preparation of this vegetable requires prolonged boiling (at least 30 minutes) with additional water and firewood in order to reduce the cyanogen content. Consumption of cassava leaves inadequately prepared and long-term smoke inhalation from the firewood, the only fuel available, may be additional factors in the overall cyanogen exposure. Low availability of water and firewood might be considered as additional risk factor for cyanide exposure (Chapter III).

Half of the urine samples analysed contained levels of thiocyanate above 300 µmol/l while

Tshala-Katumbay et al. (2001b) reported that 75 % of the urine samples contained thiocyanate levels above 300 µmol/l in the same “high prevalence area” of this study.

Tylleskär et al. (1992) also reported a high thiocyanate excretion in this region. When comparing the study areas, we found higher cyanogen content in cassava flour from the low

Konzo and cassava toxicity 102 prevalence area than in cassava flour from the other two areas. However, this cyanide content in the flour reflects only the safety of the food and the potential human exposure to cyanide, as the excretion of thiocyanate is higher in the urine from moderate and high prevalence areas than in the urine from the low prevalence area. The most affected population might be better aware of a further konzo attack and take more precaution on improving the processing of cassava, the sole staple food consumed (Cardoso et al. 2004).

The proportion of thiocyanate formed from a cyanide load will decrease if the subject is malnourished. The conversion of cyanide to thiocyanate implies a reaction with sulphur originating from dietary sulphur amino acids in the presence of rhodanese (thiosulphate- sulphurtransferase, EC 2.8.1.1) as catalyst. The rate of detoxification is therefore limited by the supply of a sulphur donor. Addition of different condiments containing sulphur amino acids in the cassava – based diet might also reduce the risk for konzo.

The concentrations of taurine in our samples were lower than those reported as reference limits from urine samples of participants on a normal diet (Venta, 2001) and from female college students of Japan (Nakamura et al. 2002). More than half (61.9 %) of the urine samples were below the reference limits of 13 to 534 mmol/ mol creatinine set by Venta

(2001). The average excretion of taurine from our samples (0.11 ± 0.15 mmol/g creatinine) was 7 fold lower than the average excretion of taurine (0.78 ± 0.53 mmol/g creatinine) from urine of 58 female college students of Japan (Nakamura, 2002). This may reflect a low intake of dietary sulphur containing amino acids in the study areas. Taurine as well as thiocyanate is an end product of the catabolism of sulphur containing amino acids metabolism involving methionine and cysteine. Low sulphur amino acid intake can lead to low excretion of taurine and thiocyanate. However, in this study the two metabolites were found to be slightly or not correlated. Production of thiocyanate may affect the quantity of taurine excreted. In vivo production of taurine in rats has been observed to be reduced preferentially over sulphate

Konzo and cassava toxicity 103 production when the supply of sulphur containing amino acids is limited (Tomozawa et al.

1998). Taurine is abundant in the brain where it has multiple functions as an anti-oxidant and neuroprotectant. The physiological effects of taurine depletion from the brain are not well documented.

It has been mentioned before that konzo occurs when the diet contains predominantly insufficiently processed cassava (Tylleskär et al. 1992), which is also deficient in sulphur containing amino acids (chapter III). Intake of sulphur amino acids from other components of the diet needs to be known to allow correct evaluation.

In conclusion, this study reveals that konzo is still occurring in this area and that there is a risk of dietary cyanogen exposure from cassava flour. Cassava flour samples from more than half of the selected households contained total cyanogen above the WHO/ FAO recommended safe limit. The urine analyses suggest an overload of cyanogen. The urine samples from more than half of the participants excreted high amounts of thiocyanate and low amounts of taurine.

The low concentrations of taurine in the urine samples may suggest that more sulphur metabolites be directed to detoxification of cyanide by formation of thiocyanate and can also reflect the suboptimal intake of sulphur containing amino acids in the diet. No correlation was found between taurine and thiocyanate; this might be due to the small number of urine samples and the large variability in the data (taurine content of 61.9 % of the samples were below the limit reference). More samples are needed to better evaluate the relationship between urinary taurine and thiocyanate.

Food diversification and proper cassava processing combined with better and organised public information can contribute to decrease the high dietary cyanogen exposure and the risk for konzo. There is a need to adjust upwards the dietary requirements for sulphur amino acids to compensate for the demand for cyanide detoxification in cassava consuming areas.

Konzo and cassava toxicity 104

CHAPTER V

GENERAL DISCUSSION AND CONCLUSIONS

Konzo and cassava toxicity 105

V General discussion and conclusions

This thesis summarises research on toxicological and dietary factors involved in konzo. The investigations were observational and focussed on possible associations between the occurrence of the disease and the exposure to cyanogen and the composition of the diet. Field observations and measurements and laboratory analyses were done. Cross sectional and ecological design were used to construct an updated epidemiological picture of differential distribution of konzo among people with different risk profiles, to measure and to explore both exposure and outcome. The design focussed on the incidence and distribution of the disease in the community, and on the characteristics of the population groups rather than on the individual members. Aggregate data based on surveys of groups of people were used to assess diet-disease relationships. The unit in which the data were collected is the household.

Urine was collected on an individual basis. The household food consumption focussed on nutritional aspects of the diet. Although individual based studies would allow a more direct estimation of the risk of disease in relation to exposure, population based methods might show that populations that have a higher exposure to cyanogen also have a higher rate of konzo but it would not necessarily follow that konzo only occurred in areas with high dietary exposure of the population to cyanogen.

Therefore, a limited community design was used to describe konzo and to identify possible dietary and household associated factors. We subsequently sampled cassava foods and urine to allow the evaluation of potential etiologic exposure as well as the interrelationships among them.

Cases of konzo were unambiguously distinguished by their physical disability.

Misclassification was not an issue in our studies. High expert medical doctor and nurse, and the author of this thesis confirmed, after examination, all konzo cases included and further supervised the interviews conducted by trained enumerators. However since the studies rely

Konzo and cassava toxicity 106 on information obtained from the heads of household, recall bias cannot entirely be excluded.

Furthermore, in a typical rural setting like that of the study area where most of the heads of household have a low educational level and food intake is based mainly on cassava as staple food, the use of a food frequency questionnaire was not applicable. A qualitative 24 hour recall was used to assess food consumption. The lack of quantitative measurements is a limitation of the work. Estimation of food portion sizes and measurement of food intake per kg of individual body weight could allow a better toxicological evaluation. However, the study subjects share the same geographical environment, ethnicity and culture and we have validated the data through repeated observation, using a set of open questions format to investigate the 24 hour and seasonal food recall.

The studies in this thesis addressed new areas in konzo research ranging from the identification of household factors to dietary risk and protective factors. The identification of this wide array of associated factors led us to suggest possible measures to prevent the disease and to recommend new directions for further research. We can not claim geographical representation since we did not cover all konzo prone areas of the D. R. C. Underestimation of the magnitude of the konzo problem is possible. The numbers of konzo cases in the study area represent about 4.3 ‰ of the total number of cases in D. R. C., roughly estimated at

100,000 (R. D. C., 2000)

V.1 Occurrence of konzo

Although most of the cases reported in chapter 3 occurred in the 1990s. At present, konzo, a preventable disease, is still occurring in Popokabaka Rural Health Zone areas, one of the regions where cases were described in the first report on konzo in 1938 (Trolli, 1938). The clinical picture of the affected subjects in our studies was similar to those previously described in the same region (Trolli, 1938; Tylleskär et al 1992; Banea-Mayambu et al 2000;

Tshala-Katumbay et al, 2001; Bonmarin et al., 2002). The main symptom was non-

Konzo and cassava toxicity 107 progressive symmetrical spastic paraparesis (paralysis of both legs) with sudden onset. Konzo was already described in other parts of D. R. C. and in other sub-Saharan African countries such as Mozambique, Tanzania, Angola, the Republic of Central Africa and Cameroon.

However, konzo does not occur in all regions with cassava-based diets even in the same country. No konzo case has been reported in Latin America from where cassava originated and is consumed as staple food neither in Nigeria, the big producer and per capita consumer of cassava nor in Indonesia and Thailand where cassava is a popular food. Therefore, high consumption of cassava root by itself is not the only or exclusive cause for konzo (Tylleskär,

1994c). The occurrence of konzo in a community reflects a deterioration of socio-economic conditions and those who are affected by konzo are trapped in a spiral of poverty, and educational and political neglect. Poverty and lack of education affect people’s capacity to prevent the disease as well as their ability to live in areas having less exposure to this risk.

There is a relationship between poverty (socio-economic as well as educational) and vulnerability, and they are mutually reinforcing.

Age and sex distributions of konzo affected subjects in our study were similar to most of the previous studies in the same region and elsewhere (Howlett, 1994; Banea-Mayambu, 1997,

Tshala-Katumbay et al, 2001b). Children less than 14 years and women at childbearing age are particularly susceptible to konzo. In Chapter 4, our studies compared the expected daily intake of the sulphur amino acids methionine and cysteine provided by cassava cossettes consumption with the suggested essential amino acid requirement, and we concluded that children of 1 to 9 years old cannot expect to meet methionine requirement whereas adults can meet the minimal requirement for these sulphur amino acids. No child under two years was affected by konzo. In Bandundu province, breastfeeding is general at birth and this practice decreases slowly until two years old: 98.6 % of children aged from 12 to 15 month and 62.7

% of those aged from 20 to 23 month old are still breastfed (R. D. C., 2001). Mother milk is

Konzo and cassava toxicity 108 rich in amino acids containing sulphur. The dominance of female patients (female to male ratio 3.3: 1) in our studies is similar to almost all other studies except the ones carried out in

Tanzania and Mozambique where male cases were preponderant (Howlett, 1994; Tshala-

Katumbay et al, 2001b, Bonmarin et al, 2002). The absence of a suitable animal model for konzo makes it difficult to explain or to study the reasons for the high susceptibility of females at reproductive age for konzo. The female hormones, especially 17 β-oestrogen have been proposed as protective factor for neurolathyrism to explain the high susceptibility for young men, but can these hormones be an aggravating factor for konzo (Lambein et al,

2004)? Rural depopulation resulting in migration of active males to bigger cities looking for

“welfare” and the primary socio-economic role played by women in the household food security may expose women to higher workload and higher oxidative stress.

The socio-economic crisis that affected D. R. C. for several years has probably had a major impact in this particular region and led to the persistent occurrence of konzo during several generations. The socio-economic burden of this crippling disease on this impoverished region is heavy (Bonmarin et al, 2002). Additional stresses such as military activity, political conflicts and drought-provoked food shortage have been identified as factors leading to inadequate diets which triggered konzo epidemics (Essers, 1995).

V.2 Cassava foods and sulphur metabolites

Cassava flour mixed or stirred in boiling water to obtain a stiff porridge, the so called “fufu” or “luku” is the main daily staple food for almost all the households studied (Chapter 2).

Cassava flour is derived from the roots which in normal conditions are soaked (retted) for at least three nights, then sun dried for 3 to 5 days, then pounded and finally sieved (Chapter 3-

1). Cassava flour is an excellent source of carbohydrate (Bradbury and Holloway, 1988).

However, we find that the protein of cassava roots is of poor quality, leucine and lysine are limiting amino acids and also the proportion of methionine is low, giving a chemical score of

Konzo and cassava toxicity 109 the protein of around 40 (Chapter 3 - 1). Our studies (Chapter 4) also reveal that more than half of the households were soaking cassava roots for less than three nights. As a consequence, a high proportion of households had cassava flour containing total cyanogens above the recommended safe limit set at 10 µg HCN equivalent/g by the Codex alimentarius

(FAO/ WHO, 1991). High cyanogen exposure from frequent and exclusive consumption of insufficiently processed cassava roots is thought to be a major etiological factor in konzo.

In this region, saka-saka (pounded cassava leaves) constitutes the main condiment consumed as side-dish with luku (Chapter 2). Although quantitatively the main source of protein in the diet, this protein is also of poor quality with sulphur amino acids as the most limiting amino acids (Chapter 3-2). Cassava leaves were also found to be a potential additional source of dietary exposure to cyanogens, apart from the cassava roots. The cooking of saka-saka requires prolonged boiling (at least 30 minutes) with additional water and firewood in order to reduce the cyanogen content but the supply of both water and firewood is limited (Chapter 3-

2). There is no water source in the villages, neither is there electricity nor gas. Sources of water are located at least at 15 minutes walking (RDC, 2001) and cooking is done exclusively with firewood which nowadays is becoming scarce in this savannah area.

No potentially toxic nonprotein amino acid was detected in cassava roots as well as in cassava leaves (Chapter 3). Neurolathyrism, which shares clinical similarities with konzo, has been associated with β-ODAP, a neuro-excitatory nonprotein amino acid present in the grass pea

(Lathyrus sativus) (Getahun et al, 1999).

High thiocyanate content was found in more than half of the urine samples analysed. This suggested a high exposure to cyanide (Chapter 4). Luku and saka-saka are the known and the main sources of dietary cyanide, but smoke inhalation from wood fires inside the primitive housing and probably also inhalation of HCN escaped during soaking and sun drying of cassava roots might be considered as additional sources of cyanide exposure that need further

Konzo and cassava toxicity 110 study. The conversion of cyanide to thiocyanate in the human body requires sulphur originating from dietary sulphur amino acids. The rate of detoxification is therefore limited by the supply of a sulphur donor. When the body is regularly exposed to cyanogens, the increased synthesis of rhodanese, enzyme responsible for cyanide detoxification in the human body by forming thiocyanate, makes extra demands on the body's reserves of sulphur amino acids. If this demand is prolonged as in the regular consumption of cassava insufficiently processed, and the diet is inadequate, the synthesis of taurine may be impaired (Chapter 4).

Taurine and thiocyanate are excreted in the urine as end products of the catabolism of sulphur containing amino acids methionine and cysteine. Production of thiocyanate may affect the quantity of taurine excreted. In the case when insufficiently processed cassava is consumed as staple food, the low methionine content may aggravate the risk for cyanide toxicity and konzo disease. Moreover, the consumption of cassava leaves as the main source of dietary protein can not compensate for the methionine deficiency in konzo-affected areas where the dietary requirement for methionine needs to be adjusted for the loss caused by cyanide detoxification

(Chapter 4).

This dietary imbalance can be corrected if other components of the diet contribute to a better balanced amino acid composition, especially the level of sulphur amino acids. Consumption of cereals and sesame is found in Chapter 2 to be protective factors against konzo. Similar protection by methionine rich cereals was also found for neurolathyrism (Getahun et al,

2003). In the regions neighbouring the konzo-affected areas in Bandundu, where traditionally corn or millet flour is mixed with cassava as staple food, no cases of konzo have been reported. This may corroborate our views as to the importance of methionine for a healthier balanced diet.

Konzo and cassava toxicity 111

V.3 Conclusions and recommendations

The high urinary thiocyanate levels found in our studies show that the population of

Popokabaka is still highly exposed to dietary and perhaps environmental cyanogens and to the risk of konzo. A better balanced diet, especially richer in methionine is required to allow efficient detoxification of cyanide in the body. Therefore, we recommend:

a) Supplementation of methionine to contribute to the detoxification of cyanide should

be done in parallel with the promotion of consumption of foods rich in methionine

locally available (cereals, sesame, soybean, pumpkin seeds, eggs, meat, etc). A study

of the effect of methionine administered on urinary levels of taurine and on

thiocyanate can help to better understand the role of taurine, as an antioxidant, a

neuroprotectant, and an inhibitory neuromodulator in the central nervous system, and

the relationship between taurine and thiocyanate in konzo.

b) Promotion of safe cassava processing to reduce significantly dietary cyanogen

exposure. Sufficient soaking (retting) combined with sun drying has been proven to be

effective in reducing cyanogen. Organised public information should be promoted.

Attention should be drawn on the duration of cassava processing especially during the

dry season which is the period of low ambient daily temperature and low water

supply. Days are cloudy and the temperatures are below 20°C; HCN is evaporated

more slowly at lower temperature as the boiling point of HCN is 25.7°C, thus removal

of HCN is not optimal during the cold dry season. The dry season is reported to be the

period of high incidence of konzo. We suggest the increase of the period of soaking

and drying during this season (at least 5 days each).

c) Promotion of consumption of luku obtained from the mixing of cassava with maize

flour or other cereals. Maize is locally available but pounding by hand to transform

grain to flour seems to be hard for the population. Therefore, an adapted milling

Konzo and cassava toxicity 112

technology should be implemented. Better information on the benefit of mixing

cassava and maize or other cereals should be promoted.

d) Food diversification should be promoted especially during the dry season. Nutritional

resources are scarce during the dry season. Almost all crops are rain-fed and cannot

survive during the dry season. Watering or an irrigation system is not practicable in

this area because of absence of inputs and of major river systems for irrigation. Roads

should be better maintained to open the region to the market.

Konzo is a very much neglected disease in D. R. C., where political instability and military activity are factors attracting the international attention. The patients suffering from this incurable disease become a socio-economic liability for their family. Prevention of this disease and its dramatic socio-economic consequences can be attained by simple education of the basics of nutrition and the supply of the simple means to produce a more varied diet. A national campaign should be organised to identify isolated pockets of konzo affected communities and to distribute preventive information.

Konzo and cassava toxicity 113

REFERENCES

Konzo and cassava toxicity 114

References

Agrahar-Murugkar D, Pal PP (2004). Intake of nutrients and food sources of nutrients among

the khasi tribal women of India. Nutrition, 20, 268- 273.

Almazan AM, Theberge RL (1989). Influence of cassava mosaic virus on cassava leaf-

vegetable quality. Tropical Agriculture (Trinidad), 66, 305 – 308.

Altman PL, Dittmer DS (1974). Amino acid requirements: vertebrates. In Biology data book

Second edition, vol. 3; Federation of American Societies for experimental Biology:

Bethesda, Maryland, USA, pp 1473.

AOAC (1990). Food composition; Additives; Natural Contaminants; Official Methods of

Analysis, 15th Ed, Vol. 2, Association of Official Analytical Chemists: Washington DC,

pp 1096-1097.

ATSDR (1989). Toxicological profile for cyanide. Atlanta, GA, US Department of Health and

Human services, Public Health, Agency for Toxic Substances and Disease Registry

ATSDR (1997). Toxicological profile for cyanide. Atlanta, GA, US Department of Health and

Human services, Public Health, Agency for Toxic Substances and Disease Registry

Attia RS, El-Tabey Shehata MA, Aman ME, Hamza MA (1994). Effect of cooking and

decortication on the physical properties, the chemical composition and the nutritive

value of chickpea (C. arietinum, L). Food Chemistry, 50, 125-131

Bainbridge Z, Harding S, French L, Kapinga R, Westby A (1998). A study of the role of

tissue disruption in the removal of cyanogens during cassava root processing. Food

Chemistry, 62, 291 – 297.

Banea M, Bikangi N, Nahimana G, Nunga M, Tylleskär T, Rosling H (1992a). Haute

prevalence de konzo associée à une crise agro-alimentaire dans la region de Bandundu

au Zaïre. Annales de la Société Belge de Médecine Tropicale, 72, 295 – 309.

Konzo and cassava toxicity 115

Banea M, Poulter NH, Rosling H (1992b). Shortcuts in cassava processing and risk of dietary

cyanide exposure in Zaire. Food and Nutrition Bulletin, 14, 137 – 143.

Banea-Mayambu JP, Tylleskär T, Gitebo N, Matadi N, Gebre-Mehdin, Rosling H (1997a).

Geographical and seasonal association between linamarin and cyanide exposure from

cassava and the upper motor neuron disease konzo in former Zaire. Tropical Medicine

and International Health, 2, 1143 – 1151.

Banea M, Tylleskär T, Rosling H (1997b). Konzo and ebola in Bandundu region of Zaïre.

Lancet, 349, 621.

Banea-Mayambu JP (1997c). Dietary exposure to cyanogens from cassava. A challenge for

prevention in Zaïre. Acta Universitatis Upsaliensis. Comprehensive Summaries of

Uppsala dissertations from the Faculty of Medecine 689, 64 p.

Banea-Mayambu JP, Nkiabungu B, Tylleskar T, Rosling H (1998). High cassava

consumption without cyanide exposure in Kinshasa, Zaire. Ecology of Food Nutrition,

37, 363-377.

Banea-Mayambu JP, Tylleskär T, Tylleskär K, Gebre-Medhin M, Rosling H (2000). Dietary

cyanide from insufficiently processed cassava and growth retardation in children in the

Bandundu Region. Annals of Tropical Paediatrics, 20, 34 – 40.

Bezkorovainy A, Rafelson ME (1996). Protein and amino acid metabolism. In: Concise

Biochemistry, A. Bezkorovainy and M.E. Rafelson, Jr, editors. Marcel Deker, Inc, New

York, pp.535 – 579.

Bokanga M (1994a). Processing of cassava leaves for human consumption. Acta

Horticulturae, 375, 203-207.

Bokanga M, Ekanayake IJ, Dixon AGO, Porto MCM (1994b). Genotype-environment

interactions for cyanogenic potential in cassava. Acta Horticulturae, 375, 131 – 139.

Konzo and cassava toxicity 116

Bonmarin I, Nunga M, Perea WA (2002). Konzo outbreak, in the South-west of the

Democratic Republic of Congo, 1996. Journal of Tropical Pediatrics, 48, 234 – 238.

Bradbury JH, Holloway WD (1988). Chemistry of tropical root crops: significance for

nutrition and agriculture in the Pacific. ACIAR Monograph No 6, 201 p.

Bradbury MG, Egan SV & Bradbury JH (1999). Determination of all forms of cyanogens in

cassava roots and cassava products using picrate paper kits. Journal of the Science of

Food and Agriculture, 79, 593-601.

Bradbury JH (2004). Konzo counts. CCDN news, 3, 1.

Brand HS, Jörning GGA, Chamuleau RAFM, Abraham-Inpigjn L (1997). Effect of a protein-

rich meal on urinary and salivary free amino acid concentrations in human subjects.

Clinica Chimica Acta, 264, 37 – 47.

Brody T (1994). Nutritional aspects of protein catabolism. In: Nutritional Biochemistry;

Academic press, Inc.: San Diego, California, pp 336-338.

Brough SH (1991). The nutritional evaluation of cassava (Manihot esculenta Crantz). Ph D

thesis, University of Nottingham

Cardoso AP, Ernesto M, Cliff J, Bradbury H (1999). High levels of total cyanogens in cassava

flour related to drought in Mozambique. Roots, 6, 4 – 6.

Cardoso AP, Ernesto M, Nicala D et al (2004) Combination of cassava flour cyanide and

urinary thiocyanate measurements of school children in Mozambique. International

Journal of Food Science and Nutrition, 55, 183 – 190.

Carlsson L, Mlingi N, Ronquist G, Rosling H (1995). A specific and sensitive method for the

determination of linamarin in urine. Natural Toxins, 3, 378 – 382.

Carlsson L, Mlingi N, Juma A, Ronquist G, Rosling H (1999). Metabolic fates in humans of

linamarin in cassava flour ingested as stiff porridge. Food and Chemical Toxicology, 37,

307 – 312.

Konzo and cassava toxicity 117

Carod-Artal FJ, Vargas AP, del Negro C (1999). Spastic paraparesis due to long term

consumption of wild cassava (Manihot esculenta): a neurotoxic model of motor-neurone

disease. Revista De Neurologia, 29, 610 – 613.

Carter SE, Fresco LO, Jones PG (1997). Introduction and diffusion of cassava in Africa.

(http://www.cgiar.org/iita/inform/trn_mat/irg49/irg49.html) 6/4/2000

Carton H, Kayembe K, Kabeya, Odio, Billiaus A, Maertens K (1986). Epidemic spastic

paraparesis in Bandundu (Zaïre). Journal of Neurology, Neurosurgery and Psychiatry,

49, 620 – 627.

Cassab J, Penalva-de-Oliveira AC (2000). The pathogenesis of tropical spastic parparasesis/

human T-cell leukemia type I-associated myelopathy. Brazilian Journal of Medical and

Biological Research, 33, 1395 – 1401.

CEPLANUT (1982). Rapport de mission sur le “buka-buka” (paralysie spastique) dans la

vallée de la Lukula, Luie et Inzia dans la region de Bandundu. CEPLANUT, Rapport

XXX.

CEPLANUT (1988). Collection de recettes zaïroises. 1ère Ed, Centre National de

Planification de Nutrition Humaine, Kinshasa, pp 61-62.

Clemente A, Sanchez-Vioque R, Vioque J, Bautista J and Millan F (1998). Effect of cooking

on protein quality of chickpea (Cicer arietinum) seeds. Food Chemistry, 62, 1-6.

Cliff J, Lundquist P, Mårtensson J, Rosling H, Sörbo B (1985). Association of high cyanide

and low sulphur intake in cassava-induced spastic paraparesis. Lancet, ii, 1211 – 1213.

Cliff J, Nicala D (1997a). Long term follow-up of konzo patients. Transactions of the Royal

Society of Tropical Medicine and hygiene, 91, 447 – 449.

Cliff J, Nicala F, Saute R, Givragy R, Azambuja G, Taela A, Chavane L, Howarth J (1997b).

Konzo associated with war in Mozambique. Tropical Medicine and International Health,

2, 1068 – 1074.

Konzo and cassava toxicity 118

Conn EE (1994). Cyanogenesis – a personal perspective. Acta Horticulturae, 375, 31 – 43.

De la Cruz C, Lopez-Hernadez J, Gonzalez-Castro MJ, Rodriguez-Bernaldo de Quiros AI and

Simal-Lozano J (1999). Effect of various culinary treatment on the amino acid content

of green beans. Deut Lebensm-Rundsch, 12, 507-510.

Diasolua Ngudi D, Kuo Y-H & Lambein F (2002) Food safety and amino acid balance in

processed cassava “cossettes”. Journal of Agricultural and Food Chemistry, 50, 3042-

3049.

Diasolua Ngudi D, Kuo Y-H & Lambein F (2003a) Amino acid profiles and protein quality of

cooked cassava leaves or ‘saka-saka’. Journal of the Science of Food and Agriculture,

83, 529 – 534.

Diasolua Ngudi D, Kuo Y-H & Lambein F (2003b) Cassava cyanogens and free amino acids

in raw and cooked leaves. Food Chemistry and Toxicology, 41, 1193 – 1197.

Djazuli M, Bradbury HJ (1999) Cyanogen content of cassava roots and flour in Indonesia.

Food Chemistry, 65, 523 – 525.

Donaghy M (1999). Classification and clinical features of motor neurone diseases and motor

neuropathies in adults. Journal of neurology, 246, 331- 333.

Egan SV, Yeoh HH, Bradbury JH (1997). Simple picrate kit for determination of the

cyanogenic potential of cassava flour. Journal of the Science of Food and Agriculture,

75, 258-262.

Ermans AM, Mbulamoko NM, Delange F, Ahluwalia R (1980). Role of cassava in the

etiology o of endemic goitre and cretinism, IDRC-136e, Ottawa, Canada.

Ernesto M., Cardoso AP, Nicala D., Mirione E., Massaza F., Cliff J., Haque M.R., Bradbury

J.H. (2002a). Persistent konzo and cyanogens toxicity from cassava in northen

Mozambique. Acta Tropica, 82, 357 – 362.

Konzo and cassava toxicity 119

Ernesto M., Cardoso AP, Nicala D., Mirione E., Massaza F., Cliff J., Haque M.R., Bradbury

J.H. (2002b). Strategy for the elimination of konzo in Mozambique. Roots, 8, 8 – 11.

Essers AJA (1989). Reducing toxicity by simple changes in the traditional preparation of

cassava leaves- a small trial in Mozambique. Tropical Science, 29, 269 -272.

Essers AJA, Alsen P, Rosling H (1992). Insufficient processing of cassava induced

intoxications and the paralytic disease konzo in a rural area of Mozambique. Ecology of

Food and Nutrition, 27, 17 – 27.

Essers AJA (1995). Removal of cyanogens from cassava roots: studies on domestic sun-

drying and solid-substrate fermentation in rural Africa. Ph D Thesis, Wageningen

Agricultural University, The Netherlands, 131p.

Essers AJA, Van der Grift RM, Voragen AGJ (1996). Cyanogen removal from cassava roots

during sun-drying. Food Chemistry, 55, 319 – 325.

Essers AJA, Alink GM, Speijers GJA et al. (1998) Food plant toxicants and safety, risk

assessment and regulation of inherent toxicants in plants. Environmental Toxicology

and Pharmacology, 5, 155-172.

FAO (1990). Roots, Tubers, Plantains and Bananas in Human Nutrition. FAO, Food and

Nutrition series, No 24; FAO, Rome.

FAO (1991). Protein quality evaluation. Report of the Joint FAO/WHO Expert Consultation,

FAO Food and Nutrition paper no 51. Food Agriculture Organization, Rome.

FAO (2004). World Production and trade of cassava products.

(http://www.fao.org/docrep/x5032E/x5032E08.htm)

FAO/WHO (1991). Joint FAO/ WHO food standards programme. Rome: Codex alimentarius,

Commission XII, Supplement 4, FAO/ WHO.

Konzo and cassava toxicity 120

FAO/WHO/UNU (1985). Energy and Protein requirements. Report of Joint FAO/WHO/UNU

Expert consultation: WHO technical report series no 724. World Health Organization:

Geneva, Switzerland.

Finetin, M. (2001). Protéines et acides aminés essentiels in Végétarisme pratique : Le guide

de la complémentation protéique. Muriel Finetin, diététicienne- La diététique en

question (http://perso.club-internet.fr/fderad/tables/aae.htm).

Firmin, A, Kamenan, A (1996). Changes in the protein and amino acid composition of

cassava roots (Manihot esculenta, Crantz) during processing. Tropical Science, 36, 138-

142.

Formunyam RT; Adegbola AA, Oke OL (1985).. The stability of cyanohydrins. Food

Chemistry, 17, 221-225.

Friedman M (1996). Nutritional value of proteins from different food sources. A review.

Journal of Agricultural and Food Chemistry, 44, 6-29.

Gaitonde MK, Dovey K (1970). A rapid and direct method for the quantitative determination

of tryptophan in the intact protein. Biochemistry Journal, 117, 907- 911.

Geervani P and Theophilus F (1980). Effect of home processing on the protein quality of

selected legumes. Journal of Food Science, 45, 707-710

Getahun H, Mekonen A, Haimanot R, Lambein F (1999). Epidemic of neurolathyrism in

Ethiopia. Lancet, 354, 306 – 307.

Getahun H, Lambein F, Vanhoorne M (2002a). Neurolathyrism in Ethiopia: Assessment and

comparison of knowledge and attitude of health workers and rural inhabitants. Social

Sciences and Medicine, 54, 51 – 62

Getahun H, Lambein F, Vanhoorne M, Van Der Stuyft (2002b).Pattern and associated factors

of the neurolathyrism epidemic in Ethiopia. Tropical Medicine and International Health,

7, 118 – 124.

Konzo and cassava toxicity 121

Getahun H, Lambein F, Vanhoorne M, Van der Stuyft P (2003). Food-aid cereals to reduce

neurolathyrism related to grass-pea preparations during famine. The Lancet, 362, 1808 –

1810.

Glew RH, Vanderjagt DJ, Lockett C, Corinetti LE, Smith GC, Pastuszyn A, Millson M

(1997). Amino Acid, fatty acid and mineral composition of 24 indigenous plants of

Burkina Faso. Journal of Food Composition and Analysis, 10, 205-217.

Goldsmith RF, Earl JW, Ouvrier RA, Procopis P, Cummingham CSF (1990). Amino acids in

childhood epilepsy. In: Lubec G, Rosenthal GA, (Editors), Amino acids: chemistry,

biology and medicine. Escom science publishers B.V., Leiden, The Netherlands, pp

415-421.

Goossens F (1996). The cassava production system. In: Cassava Production and Marketing in

Zaïre: The market of Kinshasa. Leuven University Press: Leuven, Belgium, pp 53-71.

Hahn SK (1989).An overview of traditional cassava processing and utilization. Outlook on

Agriculture, 18, 110-118.

Haque MR, Bradbury H (1999). Simple method for determination of thiocyanate in urine.

Clinical Chemistry, 45, 1459 –1464.

Hoffer LJ (2002). Methods for measuring sulfur amino acid metabolism. Current opinion in

clinical nutrition and metabolic care, 5, 511 – 517.

Hou C, Wykes LJ, Hoffer LJ (2003). Urinary sulphur excretion and the nitrogen/ sulphur

balance ratio reveal nonprotein sulphur amino acid retention in piglets. Journal of

Nutrition, 133, 766 – 772.

Howlett WP, Brubaker GR, Mlingi N, Rosling H (1990). Konzo an upper motor neurone

disease studied in Tanzania. Brain, 113, 223 – 235.

Howlett WP (1994). Konzo: A new human disease entity. Acta Horticulturae, 375, 323-329.

Konzo and cassava toxicity 122

IITA/ UNICEF (1990). Cassava in tropical Africa; A reference manual. IITA, Ibadan,

Nigeria, 176 p.

Kakes P (1994). The function of cyanogenesis in cassava. Acta Horticulturae, 375, 79 – 85.

Kama F (1970). La République Démocratique du Congo, Géographie 3e secondaire.

Programme Officiel, Collection André Journaux. Hatier , Paris 6 e, 95p

Khan JK, Kuo Y-H, Kebede N, Lambein F (1994) Determination of nonprotein amino acids

and toxins in Lathyrus by high performance liquid chromatography with precolumn

phenylisothiocyanate derivatization. Journal of Chromatography A, 687, 113- 119.

Kimaryo VM, Massawe GA, Olasupo NA, Holzapfel WH (2000). The use of a starter culture

in the fermentation of cassava for production of "Kivunde", a traditional Tanzanian food

product. International Journal of Food Microbiology, 56, 179 - 190.

Koch BM, Sibbesen O, Swain E, Kahn RA, Liangcheng D, Bak S, Halkier A, Moller LB

(1994). Possible use of a biotechnological approach to optimize and regulate the content

and distribution of cyanogenic glucosides in cassava to increase food safety. Acta

Horticulturae, 375, 45 – 60.

Kormanisky LA, Christopherson RJ, Basu TK (2003). Sulfur: its clinical and toxicologic

aspects. Nutrition, 19, 54 – 61.

Kussendrager KD, Van Hooijdonk ACM (2000) Lactoperoxidase: physico-chemical

properties, occurrence, mechanism of action and applications. British Journal of

Nutrition, 84, suppl. 1, S19 – S25.

Lambein F, Diasolua Ngudi D, Kuo Y-H (2001) Vapniarca revisited: lessons from an

inhuman human experience. Lathyrus Lathyrism Newsletter, 2, 5 - 7.

http://go.to/lathyrus

Lambein F, Defoort B, Kuo Y-H (2004). Konzo or neurolathyrism, is there a difference?

CCDN news, 3, 2-4

Konzo and cassava toxicity 123

Lancaster PA, Brooks E (1983). Cassava leaves as human food. Economic Botany, 37, 331 –

348.

Laube B, Maksay G, Schem M, Betz H (2002) Modulation of glycine receptor function: a

novel approach for therapeutic intervention at inhibitory synapses? Trends in

Pharmacological Science, 23, 519 - 527

Lawrence J (1999) The scourge of konzo. ACIAR newsletter, 35, p 8.

Lucasse (1952). Le “Kitondji”, une paralysie spastique. Annales de la Société Belge de

Médecine Tropicale, 33, 391 – 400.

Ludolph AC, Spencer PS (1996). Toxic models of upper motor neuron disease. Journal of the

Neurological Sciences, 139, 53 – 59.

Maloney EM, Morgan OSC, Widner B, Werner ER, Fuchs D (2000). Central nervous system

activation of the indoleamine-2,3-Dioxygenase pathway in human T cell lymphotropic

virus Type I- Associated Myelopathy/ Tropical spastic paraparesis. The Journal of

Infectious Diseases, 181, 2037 – 2040.

Mbemba F, Remacle J (1992). Inventaire et composition chimique des aliments et denrées

alimentaires traditionnels du Kwango-Kwilu au Zaïre. Presses Universtaires de Namur

(Belgique), p 80.

McMahon J.M., White W.L.B., Sayre R.T. (1995). Cyanogenesis in cassava (Manihot

esculenta Crantz). Journal of Experimental Botany, 46, 731 – 741.

McMurry, J; Castellion ME (1996). The food pyramid and proteins in the diet. In:

Fundamentals of general, organic and biological chemistry Second edition; Prentice-

Hall, Inc.: New Jersey, USA, pp 543 and 745.

Minten B, Kyle S (1999). The effect of distance and road quality on food collection,

marketing margins and traders' wages: evidence from the former Zaire. Journal of

Development and. Economics, 60, 467-495.

Konzo and cassava toxicity 124

Mlingi NLV, Bainbridge ZA, Poulter NH, Rosling H (1995). Critical stages in cyanogen

removal during cassava processing in southern Tanzania. Food Chemistry, 53, 29 – 33.

Muzanila YC, Brennan JG, King RD (2000). Residual cyanogens, chemical composition and

aflatoxins in cassava flour from Tanzanian villages. Food Chemistry, 70, 45 – 49.

Mwanza JC, Tshala-Katumbay D, Tylleskär T (2005). Neuro-ophthalmolgic manifestations of

konzo. Environmental Toxicology and Pharmacology, 19, 491 - 496.

Mwela P.M (2002). Présentation de la zone de santé rurale de Popokabaka, rapport annuel.

Nakamura H, Kajikawa R & Ubuka T (2002) A study on the estimation of sulphur- containing

amino acid metabolism by the determination of urinary sulphate and taurine. Amino

Acids, 23, 427 – 431.

Nartey F (1978) Manihot esculenta (cassava): cyanogenesis ultrastructure and seed

germination. Munksgaard, Copenhagen, Denmark, 262 p.

Nweke FI, Bokanga M (1994). Importance of cassava processing for production in Sub-

Saharan Africa. Acta Horticulturae, 375, 401 – 412.

O'Brien GM, Mbome L, Taylor AJ, Poulter NH (1992). Variations in cyanogens content of

cassava during village processing in Cameroon. Food Chemistry, 44, 131-136.

Oke OL (1968). Cassava as food in Nigeria. World. Review of Nutrition and Dietetic, 9, 227-

250.

Olive MF (2002) Interactions between taurine and ethanol in the central nervous system.

Amino Acids, 23, 345 – 357.

Onwuka CFI, Akinsoyinu AO, Tewe OO (1992). Role of sulphur in cyanide detoxification in

ruminants. Small Ruminant Resource, 8, 177-284.

Padmaja G (1989). Evaluation of techniques to reduce assayable tannin and cyanide in

cassava leaves. Journal of Agricultural and Food Chemistry, 37, 712-716.

Konzo and cassava toxicity 125

Padmaja G (1995). Cyanide detoxification in cassava for food and feed uses. Critical Review

in Food Science and Nutrition, 35, 299-339.

Padmaja G (1996). The culprit in cassava toxicity: cyanogens or low protein? CGIAR news,

3, 3. (http://www.worldbank.org/html/cgiar/newsletter/oct96/6cassava.html).

Parker M. (2004). Webster’s online dictionary. (http://www.websters-online-

dictionary.org/definition/english/Pa/Paraparesis.html)

Pavy PR, Bardet JI, Rabler DM, Kamoun PP (1988). Age related reference values for free

amino acids in the first morning urine specimens. Clinical Chemistry, 34, 2092 – 2095.

Poulton J. (1990). Cyanogenesis in plants. Plant physiology, 94, 401 – 405.

Przedborski S, Vila M, Jackson-Lewis V (2003). Neurodegeneration: what is it and where are

we? The Journal of Clinical Investigation, 111, 3- 10.

Ravi V., Aked J., Balagopalan C. (1996) Review on tropical root and tuber crops. I. Storage

methods and quality changes. Critical reviews in Food Science and Nutrition, 36, 661 –

709.

Ravi V., Aked J. (1996) Review on tropical root and tuber crops. II. Physiological disorders in

freshly stored roots and tubers. Critical reviews in Food science and Nutrition, 36, 711 –

731.

Ravindran G, Ravindran V (1988).Changes in the nutritional composition of cassava

(Manihot esculenta Crantz) leaves during maturity. Food Chemistry, 27, 299 - 309

RDC (2000) Politique Nationale de Nutrition. 15p

RDC (2001). Enquête nationale sur la situation des enfants et des femmes MICS2, Rapport

synthèse. RDC/UNICEF/USAID, Kinshasa, 20 p.

Ressler C & Tatake JG (2001) Vicianin, prunasin, and ß-cyanoalanine in common vetch seed

as sources of urinary thiocyanate in the rat. Journal of Agricultural and Food Chemistry,

49, 5075 – 5080.

Konzo and cassava toxicity 126

Román GC, Spencer PS, Schoenberg BS (1985). Tropical myeloneuropathies: the hidden

endemias. Neurology, 35, 1158 – 1170.

Rose, W. C.; Wixon, R. L (1955). The amino acid requirement of man. Journal of Biological

Chemistry, 216, 763-773.

Rosling H (1988). Cassava toxicity and food security. A review of health effects of cyanide

exposure from cassava and of ways to prevent these effects. A report for

UNICEF/African Household Food Security Programme, second revised edition, 40 p.

Rosling H (1994) Measuring effects in humans of dietary cyanide exposure from cassava.

Acta Horticulturae, 375, 271-283.

Rosling H (1997). Konzo, an emerging toxico-nutritional paralytic disease in some rural parts

of Africa. Report from ENRECA-seminar on Emerging and Re-emerging disease.

ENRECA health Network, Copenhagen, pp 9 – 12.

Rozan P, Kuo Y-H, Lambein F (2000). Free amino acids present in commercially available

seedlings sold for human consumption: a potential hazard for consumers. Journal of

Agricultural and Food Chemistry, 48, 716-723.

Rubenstein E (2000). Biologic effects of and clinical disorders caused by non- protein amino

acids. Medicine, 79, 80-89.

Scott GJ, Rosegrant MW, Ringler C (2000) Roots and tubers for 21st century: Trends,

projections, and policy options. IFPRI/CIP. Food, agriculture and the environment

discussion paper 31.

Shelp BJ, Bown AW, McLean MD (1999). Metabolism and functions of gamma-

aminobutyric acid. Trends in Plant Science, 11, 446-452.

Shemer M, Perkins EG (1975). Degradation of methionine in heated soybean protein and the

formation of ß-methyl mercaptopropionaldehyde. Journal of Agricultural and Food

Chemistry, 23, 201-205.

Konzo and cassava toxicity 127

Sherr B, Lee CM, Jelesciewicz C (1989). Absorption and metabolism of lysine Maillard

products in relation to utilization of L-lysine. Journal of Agriculture and Food

Chemistry, 37, 119-122.

Sigma-RBI® (2001). Alzheimer’s disease and other neurodegenerative disorders. Sigma-

RBI®, St Louis, USA, p 1.

Sodek LS, Vecchia PTD, Lima MLG (1975). Rapid determination of tryptophan in beans

(Phaseolus vulgaris) by acid ninhydrin method. Journal of Agriculture and Food

Chemistry, 23, 1147-1150.

Spencer PS (1999). Food toxins, AMPA receptors, and motor neuron diseases. Drug

Metabolism Reviews, 31, 561 – 587.

Sreeja V, Nagahara N, Li Q, Minami M (2003). New aspects in pathogenesis of konzo: neural

cell damage directly caused by linamarin contained in cassava (Manihot esculenta

Crantz). British Journal of Nutrition, 90, 467 -472.

Stipanuk MH (2004). Sulfur amino acid metabolism: pathways for production and removal of

homocysteine and cysteine. Annual reviews of nutrition, 24, 539 – 577.

Swash M, Desai J, Misra VP (1999). What is primary lateral sclerosis? Journal of the

Neurological Sciences, 170, 5 – 10.

Talbot K (2002). Motor neurone disease. Postgraduate Medical Journal, 78, 513 - 519

Tewe O (1994). Indices of cassava safety for livestock feeding. Acta Horticulturae, 375, 241 –

249.

Thiele S, Mensink GBM, Beitz R (2004). Determinants of diet quality. Public Health

Nutrition, 7, 29- 37.

Tomozawa M, Yukihiro K, Yao WB (1998). Excretion of taurine and sulphate in rats fed with

a low protein diet. Acta Medica Okoyama, 52, 77 – 81.

Konzo and cassava toxicity 128

Tor-Agbidye J, Palmer VS, Sabri MI, Craig AM, Blythe L, Spencer PS (1998). Dietary

deficiency of cystine and methionine in rats alters thiol homeostasis required for cyanide

detoxification. Journal of Toxicological Environment and Health A, 55, 583-595.

Tor-Agbidye J, Palmer VS, Lasarev MR, Craig AM, Blythe LL Sabri MI, , Spencer PS

(1999). Bioactivation of cyanide to cyanate in sulphur amino acid deficiency: relevance

to neurological disease in humans subsisting on cassava. Toxicological Sciences, 50,

228-235.

Trolli G (1938). Paraplégie spastique épidémique, “konzo” des indigenes du Kwango. In:

Trolli G, ed. Résumé des observations reunies, au Kwango, au sujet de deux affections

d’origine indéterminée. Brussels: Fonds reine Elisabeth, 1 – 36.

Tshala-Katumbay D (2001a). On the site of the lesion in konzo: Clinical and

neurophysiological studies on a non-progressive upper motor neuron disorder. Acta

Universitatis Upsaliensis. Comprehensive Summaries of Uppsala Dissertations from the

faculty of Medecine, 1092, 72 pp.

Tshala-Katumbay D, Banea-Mayambu JP, Kazadi-Kayembe T, Nunga-Matadi R, Bikangi-

Nkiabungu F, Eeg-Olofsson KE, Tylleskär T (2001b). Neuroepidemiology of konzo – a

spastic para/tetraparesis of acute onset in a new area of the Democratic Republic of

Congo. African Journal of Neurological Sciences, 20, 8 – 12.

Tylleskär T, Banea M, Bikangi N, Fresco L, Persson LA, Rosling H (1991). Epidemiological

evidence from Zaire for a dietary etiology of konzo, an upper motor neuron disease.

Bulletin of the World Health Organization, 69, 581 – 590.

Tylleskär T, Banea M, Bikangi N, Cooke R, Poulter N & Rosling H (1992). Cassava

cyanogens and konzo, an upper motor neuron disease found in Africa. Lancet, 339, 208–

211.

Konzo and cassava toxicity 129

Tylleskär T (1994a). The association between cassava and the paralytic disease konzo. Acta

Horticulturae, 375, 331 – 339.

Tylleskär T (1994b). The causation of konzo. Studies on a paralytic disease in Africa. Acta

Universitatis Upsaliensis. Comprehensive Summaries of Uppsala Dissertations from the

faculty of Medecine, 413, 67 pp.

Tylleskär T, Ruiza HT, Banea M, Howlett WP, Aquilonius S-M, Persson L-Å, Rosling H

(1994c). Similarities between konzo and lathyrism suggest a common pathogenetic

mechanism. In The grass pea and lathyrism.; Abegaz B. M., Haimanot R. T., Palmer, V.

S., and Spencer, P. S., Eds.; Proceedings of the Second International Lathyrus/ lathyrism

Conference in Ethiopia; Third World Medical Research Foundation: New York, USA,

pp 26-32.

Tylleskär T (1994d). The association between cassava and the paralytic disease konzo. Acta

Horticulturae, 375, 331- 339.

Tylleskär T, Banea M, Rosling H (1995). Konzo- an upper motoneuron disease associated to

cassava agro-ecological collapse in Africa. In Recent Advances in Tropical Neurology,

Ed by Rose CF. Elsevier Science, BV, Amsterdam, pp 355-366

Tylleskär T, Banea M, Bottinger B, Thorstensson R, Biberfeld G, Rosling H (1996). Konzo,

an epidemic spastic paraparesis in Africa, is not associated with antibodies to HTLV-I,

HIV or HIV gag-encoded proteins. Journal of Acquired Immune Deficiency Syndromes

and Human Retrovirology, 12, 317 – 318.

Van der Beken A (1993). Proverbes yaka du Zaïre. Karthala ed. Paris: Harmattan.

Venta R (2001) Year-long validation study and reference values for urinary amino acids using

a reversed-phase HPLC method. Clinical Chemistry, 47, 573-583.

Verma A, Bradley WG (2001). Atypical motor neuron disease and related motor syndromes.

Seminars in neurology, 21, 177 – 187.

Konzo and cassava toxicity 130

Vesey CJ, McAllister H & Langford RM (1999) A safer method for the measurement of

plasma thiocyanate. Journal of Analytical Toxicology, 23, 134-136.

Vetter J. (2000). Plant cyanogenic glycosides. Toxicon, 38, 11-36.

Waterfield CJ, Jairath M, Asker DS & Timbrell JA (1995). The biochemical effects of

clenbuterol: with particular reference to taurine and muscle damage. European Journal

of Pharmacology, 293, 141–149.

White WLB, Arias-Garzon DI, McMahon JM, Sayre RT (1998). Cyanogenesis in cassava, the

role of hydroxynitrile lyase in root cyanide production. Plant Physiology 116, 1219 –

1225.

White WLB, McMahon JM, Sayre RT (1994). Regulation of cyanogenesis in cassava. Acta

Horticulturae, 375, 69 – 85.

WHO (1996). Konzo - a distinct type of upper motoneuron disease. Weekly Epidemiological

Records, 71, 225 – 232.

WHO (2004). Hydrogen cyanide and cyanides: Human health aspects. Concise international

Chemical Assessment Document 61. WHO/UNEP/ILO, Geneva, 64p.

Yeoh HH, Chew MY (1976). Protein content and amino acid composition of cassava leaf.

Phytochemistry 15, 1597-1599.

Yeoh HH, Truong VD (1996). Protein contents, amino acid composition and nitrogen-to-

protein conversion factors for cassava roots. Journal of the Science of Food and

Agriculture, 70, 51-54.

Zahnley JC, Davis JG (1973). Effect of high tyrosine content on the determination of

tryptophan in proteins by acid ninhydrin method. Biochemistry, 135, 59-61.

Zaninovic’ V (2001). Similarities between Tropical Spastic Paraparesis (TSP) and

neurolathyrism. Lathyrus Lathyrism Newsletter, 2, 11 – 14.

Konzo and cassava toxicity 131

Zvauya R, Ernesto M, Bvochora T, Tivana L, Francisco JC (2002). Effect of village

processing methods on cyanogenic potential of cassava flours collected from selected

districts of Nampula Province in Mozambique. International Journal of Food Science

and Technology, 37, 463 – 469.

CURRICULUM VITAE

Curriculum Vitae

Identity

Full name: Delphin DIASOLUA NGUDI

Place and date of birth: Kinshasa, May 2, 1961 Sex: Male

Nationality: Congolese (Democratic Republic of Congo)

Marital status: Married and 3 sons

Permanent address: Inga 57 Quartier 10, N’djilj/ Kinshasa, D R Congo

Present address: Filips Van Cleeflaan 386, 9000 Gent, Belgium

Email address: [email protected]

Education

ƒ Master of Science in Food Science and Technology, Universiteit Gent/ Katholieke

Universiteit Leuven, Belgium, 1999

Masters Thesis: Effect of sprouting and lactic acid fermentation on protein in finger millet

(Eleusine Coracan) and kidney beans (Phaseolus vulgaris

Promoter: Prof Dr Ir André Huyghebaert and Dr John Van Camp

ƒ Degree in Complementary Studies in Nutrition and Food Science, Universiteit Gent,

Belgium, 1997

Thematic study: Mise en place des activités de surveillance nutritionnelle à Kinshasa

ƒ Graduate in Medical Techniques: Nutrition and dietetics, Institut Supérieur de Techniques

Médicales, Kinshasa, R. D. Congo, 1983

Thematic study: Evaluation du critère de prise de poids ou du périmètre brachial pour la

récupération des enfants malnourris.

iii

Additional Training

ƒ Summer course on “biosafety assessment and regulation of agricultural biotechnology”,

Plant Biotechnology Institute for Developing Countries (IPBO), Ghent university,

Belgium, 2004.

ƒ International Training Course in Dairy Technology: “Dairy technology from rural to

industrial level”, B.A.D.C./ Universiteit Gent, Belgium, 1998

ƒ Socrates Intensive Course: Food Packaging, European Union/ Universiteit Gent, Belgium,

1998

ƒ International Training in Nutrition and Food Science, Target Program on food security:”

Micronutrients deficiencies, Université de Benin, Cotonou, Bénin,1994

ƒ Training in writing of health education school manuals, UNICEF/ Ministry of Primary

and Secondary Education, Kisantu, R. D. Congo, 1991

ƒ Training of Health Community Workers Trainers, (U.S.A.I.D./ SANRU Project),

Kimpese, R .D. Congo, 1985

Employement record

• 2000 to date: Research on cassava and konzo at Universiteit Gent, Belgium

• 1984 to date: Nutritionist and Food quality control Officer at the National Program of

Nutrition (PRONANUT/ D R. Congo) (former CEPLANUT)

• 1992 to 1996: Collaborating Assistant to the Director of CEPLANUT and Member of

the Technical Committee

• 1993 to 1995: Technical Assistant to the project TCP/ ZAI/ 2355(A) – FAO/

CEPLANUT) “Food vended street”

• 1989 to 1992: Assistant to the Activity coordinator of CEPLANUT Regional Office

of Bandundu Region in Kikwit

iv

• 1984 to 1989: Head of the Nutrition Education section and Contact for the regional

council for food and nutrition of Bandundu Province (CRANB), project

660-079 ( United States Agency for International .Development

(USAID) / CEPLANUT)

International meeting Attendance

ƒ The first International Conference on Food Systems, college of Food systems, United

Emirates University, Al- Ain (United Arab Emirates), October 19 – 21, 2003

ƒ World Health Organization (WHO) AFRO Regional awareness raising workshop on food

safety evaluation, Bamako, (Mali), December 4 – 6, 2002.

ƒ International Food Policy Research Institute (IFPRI) 2020 vision: Sustainable Food

Security for All by 2020, Bonn (Germany), September 4-6, 2001 (Participation sponsored

by GTZ- Echborn)

Publications

1. Diasolua Ngudi D., Kuo Y.H., Lambein F.: Cassava cyanogens and free amino acids

in raw and cooked leaves. Food and Chemical Toxicology 2003, 41, 1193-1197.

2. Diasolua Ngudi D., Kuo Y.H., Lambein F.: Amino acid profiles and protein quality of

cooked cassava leaves 'saka saka'. Journal of the Science of Food and Agriculture.

2003, 83, 529-534.

3. Diasolua Ngudi D., Kuo Y.-H., Lambein F. Cassava leaves, a non-negligible source of

dietary exposure to cyanogens. Cassava Cyanide Diseases Network (CCDN) News

2003, 2, 1-2.

4. Diasolua Ngudi D., Kuo Y.-H., Lambein F. : Food Safety and Amino Acid Balance in

Processed Cassava "cossettes" Journal of Agricultural and Food Chemistry 2002, 50,

3042-3049.

v

5. Lambein F., Diasolua Ngudi D., Kuo Y.-H.: Vapniarca revisited: Lessons from an

inhuman human experience. Lathyrus lathyrism newsletter 2001, 2, 5-7.(website URL:

http://go.to/lathyrus).

6. Mbithi-Mwikya S., Ooghe W., Van Camp, J., Ngudi, D., Huyghebaert A.: Amino acid

profiles after sprouting, autoclaving, and lactic acid fermentation of finger millet

(Eleusine coracan) and kidney beans (Phaseolus vulgaris L). J. Agr.Food Chem. 2000,

48, 3081- 3085.

7. Diasolua Ngudi D., Kuo Y.-H., Lambein F., Patrick Kolsteren. High risk of dietary

cyanogen exposure in a population living in a konzo – affected area of Democratic

Republic of Congo. Food and Chemical Toxicology (submitted for publication)

8. Diasolua Ngudi D., Banea-Mayambu J.-.P., Lambein F., Kolsteren P.:.Crippling konzo

in DRC, three generations later. The Lancet (submitted for publication)

Memberships

ƒ Member and country contact of cassava cyanide diseases network (CCDN)

ƒ Member of the Congolese Nutritionists and Dieticians Association and former Secretary

of the Council (1995-1996)

ƒ Member of Science Press, vzw

Our web page: http://www.sciencepress.be

Visit web from Delphin

ISBN 90-5989-073-6