Idiopathic cardiomegaly in Africa*

Idiopathic cardiomegaly in Africa* A. C. IKEME 1 Idiopathic cardiomegaly is probably the commonest single diagnosis other than hyper- tension made in tropical and subtropical African cardiovascular practice. Understanding of the nature of this disease has been hampered by failure to recognize the possibility that the term " idiopathic cardiomegaly" may embrace several disease entities. Evidence suggests that many factors, sometimes acting singly, but often acting in combination, may be responsible for the genesis of so-called idiopathic myocardial failure. The future attitude to research should not be one of excluding well-defined forms from the concept of idiopathic cardiomegaly, but one of clinicopathological classification, which should be a prelude to the search, within each moiety of this group of disorders, for a specific or dominant etiological factor. Idiopathic cardiomegaly is an amorphous area in tropical cardiology, largely because no specific diag- nostic clinical or pathological features have been defined in this disorder. Recognition of the disease, both in life and post mortem, still depends on the exclusion of known causes of heart failure. This approach would obviously lead to overdiagnosis in areas with poorly developed diagnostic facilities, and it is unfortunately in such parts of the world that a high prevalence of idiopathic cardiomegaly has been reported. Nevertheless, the difficulties posed by this approach to the recognition and understanding of the disease are not confined to tropical developing countries. Both Brigden (1) and Mattingly (2) have contributed interesting accounts of the development of thinking on similar disorders in developed coun- tries. One of the earliest reports of cardiomyopathies from Africa was communicated to the British Car- diac Society by Bedford & Konstam in 1946 (3). These authors described heart disease in West Afri- can troops aged 20 to 40 years, which in retrospect must have included the two conditions now known as endomyocardial fibrosis and idiopathic cardio- megaly. There then followed a series of descriptions of cardiomyopathies by Davies from Uganda in 1948 (4), and, from South Africa, by Gillanders in * Based on an address delivered at the CIBA-Geigy Symposium on Cardiovascular Disease in Africa, Ibadan, Nigeria, 15-18 March 1976. l Leader, WHO Cardiovascular Research Team, Univer- sity of Ghana Medical School, Accra, Ghana. Present address: Faculty of Medicine, University of Jos, Jos, Nigeria. 1951 (5) and Becker et al. in 1953 (6). These descrip- tions created much confusion because of the large number of terms used, and it was not until a seminar, held in South Africa in 1957 (7), that it was agreed that the cardiomyopathies described by vari- ous authors from South Africa were the same and that they were different from the condition described by Davies. This view was confirmed by D. H. Connor (unpublished observations, 1964) when he visited several countries in Africa on behalf of WHO, and it has been supported by others (8, 9), although there are still a few dissenting opinions (10, 11). In 1965 WHO held the first of its consultative meetings of clinical cardiologists on the cardiomyo- pathies (12). At this meeting, the clinical features of endomyocardial fibrosis and of cardiomegaly of unknown origin were defined. It was accepted that cardiomegaly of unknown origin included the dis- orders described by Gillanders (5) and by Becker et al. (6); the cryptogenic heart disease described by Higginson et al. (13), and the heart muscle disease described by Edington & Jackson (10) in Ibadan; as well as the cardiomegaly of obscure origin described by Stuart & Hayes (14) from Jamaica. It is generally assumed that cardiomegaly of unknown origin also includes the subgroup of congestive cardiomyo- pathy described by Goodwin et al. (15). At aWHO consultative meeting in 1968, the term " idiopathic cardiomegaly " was adopted for these disorders (16). The term therefore embraces all those subacute or chronic disorders of the myocardium associated with cardiomegaly and congestive cardiac failure occur- ring in the absence of valvular disease, hypertension, 3540 - 455 BULL. WORLD HEALTH ORGAN., Vol. 54, 1976 A. C. IKEME coronary artery disease, or other observable cause and presenting at autopsy with variable cardiac dilatation and hypertrophy and with inconstant endocardial lesions. This definition is offered for idiopathic cardiomegaly, because it is evident that, with the limited mode of response to injury by the myocardium, the end stage of myocardial disease from many causes may be myocardial failure, oc- curring at a stage when the pathogenetic factor may no longer be recognizable from macroscopic or routine microscopic study of the myocardium; and because it is this approach of multiple pathogenicity that is likely to yield results in the search for the etiology of idiopathic cardiomegaly. IDIOPATHIC CARDIOMEGALY IN AFRICA Unexplained myocardial failure has been reported from all parts of the world, but there is clear evidence of its preponderance in tropical and sub- tropical countries (17-19) despite the possibility of overdiagnosis in those countries. Reports from the tropics also show that the condition is more preva- lent in negroes than in non-negroes, and that there is a high prevalence among underprivileged sociocul- tural groups. This is confirmed by reports of heart disease among the Bantu or migrant workers in South Africa (20). These socioeconomic differences suggest a fruitful area for study in the search for the etiology of idiopathic cardiomegaly, and also that there may be a case for regarding the disease as a problem specific to the tropics. Edington & Gilles (21) have observed that, in Africa, idiopathic cardiomegaly occurs at all ages. Although cases of idiopathic cardiac enlargement with heart failure in children have been reported by Antia et al. (22) and by a number of other workers in Africa (23, 24), it is my belief that the use of the term " idiopathic cardiomegaly " for such cases has helped to obscure the significance of the disorder seen in children. In tropical countries the condition occurs predominantly in middle life and old age (25). Supporting epidemiological data from Accra, Ghana, have been presented at the CIBA-Geigy Symposium (see footnote on page 455). Experience from medical units in various parts of tropical Africa also indicates that idiopathic cardiomegaly occurs most frequently in the 5th and 6th decades of life. Data reported by Carlisle & Ogunlesi (27) on the dis- tribution by age of 38 consecutive cases seen at one hospital in Ibadan over a 24-month period suggest that there exist, in the tropics, at least two distinct groups of individuals with heart failure of obscure origin, pointing to the presence of different patho- genetic groups. Antia (28) in his follow-up of children with idio- pathic cardiomegaly, reports a natural history differ- ent from that observed in adults. Only one-third of the children returned for follow-up, and all of them were found to have returned to normal. Kelly et al. (29), on the other hand, reported a 71 % mortality in 5 years among 51 adult Bantu patients with this disorder. Antia's cases of obscure heart failure in infancy and childhood were characterized by systemic upset, heart failure, and a good progno- sis. It has been suggested that myocarditis associated with recent viral infections may be the underlying disease in such cases (16, 30-32). Until more infor- mation is available. I suggest that obscure heart failure in children and young adults should be regarded as an entirely different condition from that seen in middle life and old age. The chapter therefore will deal with idiopathic cardiomegaly as a disorder or group of disorders seen chiefly in tropical and subtropical Africa and occurring predominantly in adults, usually among the underprivileged. PATHOGENESIS OF IDIOPATHIC CARDIOMEGALY IN AFRICA In adults, the disease tends to occur earlier in women than in men, the greatest number of cases occurring in the 35-54-year age group in the former and in the 55-74-year age group in the latter (40). In a clinical study of 35 cases in Uganda by Ikeme et al. (33), 3 (out of 15) women had heart failure within 8 weeks of parturition. Their average age was 30.3 years, with a mean number of pregnancies of 5.3. In contrast, the mean age of the other 12 women was 49.9 years, which did not differ significantly from the male value of 51.7 years. The mean number of preg- nancies in the group of older women was 3.7. This question is referred to briefly also on pages 458-459. It is evident that the lower mean age of women with idiopathic cardiomegaly could be due to the influence of pregnancy and parturition on the natural history of this disorder (34). Many reports have suggested that the blood pres- sure may rise as a consequence of congestive heart failure in idiopathic cardiomegaly (12, 35). On the other hand, a number of observers have suggested that idiopathic cardiomegaly is a form of hyperten- sive heart disease (36, 37). The possible role of 456 IDIOPATHIC CARDIOMEGALY hypertension in congestive cardiomyopathy was exhaustively discussed by Oakely et al. (38). In a recent report, I have attempted to show that both the above-mentioned views on the relationship of hypertension to myocardial failure in idiopathic cardiomegaly are unsatisfactory (39). With regard to the elevation of the blood pressure as a consequence of myocardial failure, it is to be noted that a complex interaction exists between ar- terial pressure, stroke volume, and vascular imped- ance (Fig. 1). In the presence of a normally function- ing left ventricle, increased vascular impedance results in rapid autoregulatory adjustments in the left ven- tricle, so that stroke volume is maintained with a normal end-diastolic pressure. The end result of these changes is that elevation of pressure becomes an index of the state of the peripheral circulation. NORMAL LEFT VENTRICLE LEFT VENTRICLE DYSFUNCTION ) (I OUSPS ~UUDIOPATHIC CARDIOMEGALYL EN.ANCEO CONTRACT L'TY STROKE VOLUME STROKE VOLUME NORMAL STSTOLIC SYSTOLIC BLUFF IMPEDA CE FBLOOD PRESSURE / PRESSURE s ~~~~~~~~~~~~~~~SYSTOLICALIN OFF / v ~~~~~~~~~~~~ARTERIALD STESBLT i/ NCATEC.OLAMINES ~~~~~~~~~~~~~~~SALTAND WAER VACLR RESISTANCE Fig. 1. Scheme of cardiovascular response to changes in impedance In the presence of myocardial failure, however, the diseased ventricle is unable to respond normally to the increased load resulting from increased imped- ance, and the stroke volume falls. It is irrelevant whether the increased impedance is extrinsic or a result of the circulatory adjustments to cardiac fail- ure. In the presence of a severely diminished ventri- cular reserve, which may be assumed to be present when the primary disease is myocardial failure, it is difficult to see how increased impedance can result in elevation of the blood pressure. Clearly a vicious circle ensues, in which a fall in stroke volume results in a further increase in impedance, which further compromises myocardial function. Thus the blood pressure, if previously raised, would fall and an apparently normal blood pressure could result. In connexion with the role of hypertension in the pathogenesis of heart failure in idiopathic cardio- megaly, a study of 35 patients was carried out in Uganda. These cases were classified as idiopathic cardiomegaly because of heart failure of obscure etiology in the presence of normal blood pressure on admission to hospital (33). Elevation of the blood pressure requiring treatment was noted in 7 of the 25 persons followed up for over 3 months. It was seen also that the mean aortic arch width in these patients was higher than that in normotensive control sub- jects and similar to that in hypertensive patients. Furthermore, all but two of those who continued with normal blood pressure during follow-up also had evidence of continuing severe myocardial dys- function. It was concluded from these observations that persisting myocardial failure was a factor in maintaining normal blood pressure and that a large proportion of patients regarded as having idiopathic cardiomegaly on the basis of normal blood pressure on admission to hospital were in fact persons with hypertensive disease in whom intrinsic myocardial factors had led to a fall in blood pressure despite the increased impedance associated with congestive heart failure. All the foregoing observations may be unified into a working hypothesis on the pathogenesis of idiopathic cardiomegaly (39, 40). To explain why some individuals with hypertensive disease develop heart failure with a very raised blood pressure, whereas others do so with normal (or only moderately raised) blood pressure, one needs to postulate that, in the latter, myocardial damage already exists and that the effects of both hypertension and the other factor causing myocardial damage are additive. Indeed, Davies (41) observed " that the heart of many Africans for some reason cannot stand up to sustained hypertension ". The explanation probably lies in the fact that extrinsic factors play a role in precipitating early heart failure in hypertensive dis- ease in Africa, as ischaemic heart disease does in Europe (42). Basil et al. (43) have demonstrated that thiamine deficiency occurs in idiopathic cardiomegaly, and have suggested that this may be a cause of myocar- dial failure. Similar findings have also been reported as regards idiopathic heart failure among the Bantu of South Africa (44, 45). The extent to which thiamine deficiency and various other factors con- tribute to producing myocardial damage should form the basis for future research and etiological classification in idiopathic cardiomegaly. 457 A. C. IKEME RESEARCH In looking at our future attitude towards the cardiomyopathies in Africa, one can do no better than again quoting Davies (41). [While] " I agree ... on the limited range of reac- tions of the myocardium to disease processes, I still do not think that this is as limiting a factor as might be supposed. Even if tissue reactions are limited in the overall sense, there is still a sufficient inter-play of factors, of cell reactions, of tissue reactions, of cell infiltrates, of vascular, nervous, endocardial and epi- cardial reactions, of infiltrations and degenerations, that in conjunction with the factors of time, of inten- sity, of concurrent lesions elsewhere, and with the clinical findings, the chemical and functional findings, and with the armamentarium we now have for studying morphological changes, we can hope with time, with careful study, to determine much more specificity in lesions than we recognise at present." To all these I would add the search for pathogenetic factors, which may vary at different periods of life, operate more often in one geographical area than in another, and act singly or in combination to produce myocardial failure. As a first step in research, therefore, a concerted effort should be made to classify the group of disorders now referred to under the term " idio- pathic cardiomegaly ", excluding the bulk of similar disorders described in temperate countries, some of which have been observed to be familial (46) whereas others may follow a phase of hypertrophic obstruc- tion (47). A tentative subdivision of what is now called idiopathic cardiomegaly is presented below. (1) Disease affecting children and probably young adults. In this group the possibility of myocardial damage consequent on a viral infection should be examined. It has been pointed out that, in the late effects of viral damage, evidence of myocarditis need not be present (16). Persisting impairment of myo- cardial function has been observed five years after myocarditis (30). (2) Disease affecting middle-aged and elderly adults with unfolded aorta on chest radiographs or with mild hypertension. (3) Disease affecting middle-aged and elderly adults with normal blood pressure. In the two last-mentioned groups, vascular impedance is a more accurate index of the state of the cardiovascular system than blood pressure is. Both groups are often " hypertensive ", as has al- ready been shown, although in some individuals the hypertension may be masked by " pump " failure. In both groups, however, factors other than hyper- tension obviously contribute to myocardial failure; thus aneurin deficiency has been demonstrated among some patients in these two groups (43-45). Chronic protein energy malnutrition must be examined seriously as another possible contributory factor. Myocardial damage and heart failure have been claimed to occur in kwashiorkor (48-50); and a cardiomyopathy has been reported with a protein- losing enteropathy (51). Experimentally, acutely increasing vascular impedance leads in a few hours to increased myocardial protein synthesis (52), and inhibition of such synthesis leads to heart failure (53). Although it has been usual in the past to regard the heart as an organ protected from the effects of undernutrition and protein deficiency, this view is no longer generally held (54). The possible modification of myocardial metabolism by chronic protein energy malnutrition, especially in the presence of infective states, deserves greater attention. The role of immoderate use of alcohol in the pathogenesis of idiopathic cardiomegaly is being examined together with these other factors by the WHO Cardiovascular Research team in Ghana and members of the University of Ghana Medical School staff. Excessive alcohol intake has been reported in a proportion of the cases reported from South Africa (55). Among the cases of myocardial failure reported from East Africa, a history of regular-probably excessive-use of alcohol was obtained in about one- third. While these have not been regarded as cases of alcoholic cardiomyopathy, the contributory role of alcohol in the pathogenesis of heart failure in such cases has been emphasised (55). In groups (2) and (3), therefore, attention should be focused on the possible combined roles of hyper- tension, chronic aneurin deficiency-which may be associated with over-indulgence in alcohol-and chronic protein energy malnutrition. It is also pos- sible that some cases in this group may be associated with the late effects of myocarditis. (4) Women with post-partum myocardial failure. This condition has been included here because it is not thought to differ etiologically from the forms of idiopathic cardiomegaly described above. In a com- prehensive study of myocardial failure in Kampala (40), 24 women had diastolic pressure below 114 mmHg on admission to hospital; 9 of them were in the reproductive period (up to 45 years of age) and 5 of these 9 had heart failure within 8 weeks of parturition. Only one of these 5 patients had normal 458 IDIOPATHIC CARDIOMEGALY 459 blood pressure; the others were known to be hyper- tensive during pregnancy. This predisposition to myo- cardial failure in the puerperium has also been re- ported elsewhere (34, 56). Ikeme et al. (40) concluded that an additional burden occurred during the puer- perium and suggested that this might be nutritional, and that there is no such specific pathological entity as puerperal heart disease. These women simply had idiopathic cardiomegaly together with an additional load during the puerperium. The reversible nature of some cases of puerperal heart failure calls to mind the acute reversible heart failure in Africans de- scribed by Grusin (57). An extensive search for deficiencies occurring in the puerperium should be undertaken. It is known that folate deficiency may lead to megaloblastic anaemia, which not infre- quently occurs in the puerperium and is often cor- related, as is puerperal heart failure, with high maternal age and with parity (58, 59). There are, in fact, reports suggesting tryptophan deficiency (60, 61) and aneurin deficiency (62) as underlying factors in puerperal myocardial failure. The natural history of idiopathic cardiomegaly needs to be examined in greater detail. Cardiomegaly without heart failure in the absence of hypertension or valvular disease occurs in the general population in the tropics (27). That such people progress to develop idiopathic cardiomegaly has so far been confirmed only in Jamaica (63). This needs to be confirmed in other parts of the world where the disorder occurs. Studying this phase of cardiomegaly without heart failure should facilitate a complete categorization of individuals susceptible to cardio- megaly, as an adjunct both to studying the patho- genesis of idiopathic cardiomegaly and to attempting its prevention. Greater attention should be paid to extracardiac lesions in this disease, particularly in the liver and kidneys. Such studies may contribute to the under- standing of the role of malnutrition, hypertension, and toxic or infective factors in its pathogenesis. Finally, in idiopathic cardiomegaly, any attempt to demarcate a distinct entity of primary myocardial disease by excluding categories with clear-cut causes is bound to fail. This approach leads to an ever- contracting clinical complex and to incessant changes in diagnostic criteria. We should concentrate on clas- sifying what seems to be a group of disorders with an identical or very closely similar clinical picture. Only after this can clinical and epidemiological stud- ies examine the role of the various factors in the pathogenesis of this intriguing disorder. RtSUMt CARDIOMEGALIE IDIOPATHIQUE EN AFRIQUE Dans les pays d'Afrique tropicale et subtropicale, la cardiomegalie idiopathique est probablement avec l'hyper- tension le diagnostic isole le plus courant en pratique cardio-vasculaire. La comprehension de la nature de cette affection a et6 retardee parce qu'on n'avait pas admis qu'il etait possible que le terme de ( cardiomegalie idiopathique ) puisse couvrir plusieurs entites morbides. Les observations donnent 'a penser que de nombreux facteurs, agissant parfois isolement mais le plus souvent en association, peuvent etre responsables de la genese de ce qu'on appelle l'insuffisance cardiaque idiopathique. Dans I'avenir, I'attitude en matiere de recherche doit consister non pas a exclure les formes bien definies du concept de cardiomegalie idiopathique, mais au contraire a tenter, sur la base clinique et anatomopathologique, une classification qui serve de prelude a la recherche d'un facteur etiologique, specifique ou dominant, dans chaque composante de ce groupe d'affections. REFERENCES 1. BRIGDEN, W. Uncommon myocardial diseases: non- coronary cardiomyopathies. Lancet, 2: 1179, 1243 (1957). 2. MATTINGLY, T. W. Diseases of the myocardium (cardiomyopathies): the viewpoint of a clinical car- diologist. American journal of cardiology, 25: 79 (1970). 3. BEDFORD, D. & KONSTAM, G. Obscure heart disease in West African troops. British heart journal, 8: 236 (1946). 4. DAVIES, J. N. P. Endomyocardial fibrosis in East Africa. East African medical journal, 25: 10 (1948). 5. GILLANDERS, A. D. Nutritional heart disease. British heart journal, 13: 177 (1951). 460 A. C. IKEME 6. BECKER, B. J. P. ET AL. Cardiovascular collagenosis with parietal endocardial thrombosis. Circulation, 7: 345 (1953). 7. Some African cardiomyopathies (Seminar). South African medical journal, 31: 854 (1957). 8. DAVIES, J. N. P. The heart of Africa: cardiac pathology in the population of Uganda. Laboratory investigation, 10: 205 (1961). 9. CONNOR, D. H. ET AL. Endomyocardial fibrosis in Uganda (Davies disease). American heart journal, 74: 687 (1967) and 75: 107 (1968). 10. EDINGTON, G. M. &JACKSON, J. G. The pathology of heart muscle disease and endomyocardial fibrosis in Nigeria. Journal of pathology and bacteriology, 86: 333 (1963). 11. MCKINNEY, B. A comparative histological study of endomyocardial fibrosis and cardiomegaly of un- known origin. Pathologia et microbiologia, 35: 70 (1970). 12. HUTT, M. S. R. ET AL. Cardiomyopathies. Bulletin of the World Health Organization, 33: 257 (1965). 13. HIGGINSON, J. ET AL. The pathology of cryptogenic heart disease. Archives ofpathology, 70: 496 (1960). 14. STUART, K. L. & HAYES, J. A. A cardiovascular disorder of unknown aetiology in Jamaica. Quarterly journal of medicine, 32: 99 (1963). 15. GOODWIN, J. F. ET AL. Clinical aspects of cardiomyo- pathy. British medical journal, 1: 69 (1961). 16. Idiopathic cardiomegaly (account of an informal meeting). Bulletin of the World Health Organization, 38: 979 (1968). 17. SHAPER, A. G. On the nature of some tropical cardiomyopathies. Transactions of the Royal Society of Tropical Medicine and Hygiene, 61: 458 (1967). 18. FEJFAR, Z. Cardiomyopathies: an international prob- lem. Cardiologia, 52: 9 (1968). 19. HUTT, M. S. R. Pathology of the cardiomyopathies: geographical aspects. Postgraduate medical journal, 48: 738 (1972). 20. SCHIRE, V. The racial incidence of heart disease in South Africa with particular reference to Groote Schuur Hospital, Cape Town. Journal of the Indian medical profession, 13: 57 (1966). 21. EDINGTON, G. M. & GILLES, H. M. Pathology in the tropics. London, Arnold, 1969, p. 302. 22. ANTIA, A. U. ET AL. Idiopathic cardiomegaly in Nigerian children. British heartjournal, 31:178 (1969). 23. STEIN, H. ET AL. Cardiomyopathies in African chil- dren. Archives ofdisease in childhood, 39: 610 (1964). 24. ALTMAN, A. & STEIN, H. Idiopathic hypertrophy of the heart in African children. British medical journal, 1: 207 (1956). 25. IKEME, A. C. Tropical idiopathic cardiomegaly. East African medical journal, 49: 890 (1972). 27. CARLISLE, R. & OGUNLESI, T. 0. Prospective study of adult cases presenting at the cardiac clinic, Univer- sity College Hospital, Ibadan, 1968 and 1969. African journal of medical sciences, 3: 13 (1972). 28. ANTIA, A. U. Idiopathic cardiomegaly in African children: further experiences with cases seen in Nigeria. Pathologia et microbiologia, 35: 44 (1970). 29. KELLEY, K. J. ET AL. Tropical cardiomyopathies. Lancet, 2: 523 (1963). 30. BENGTSON, E. & LAMBERGER, B. Five-year follow-up study of cases suggestive of acute myocarditis. American heart journal, 72: 751 (1966). 31. BENGTSON, E. Myocarditis and cardiomyopathy: clinical aspects. Cardiologia, 52: 97 (1968). 32. SOMERVILLE, W. Postcarditic cardiomyopathy. Post- graduate medical journal, 48: 746 (1972). 33. IKEME, A. C. ET AL. The clinical features of idiopathic cardiomegaly in the tropics. Tropical cardiology, 1: 101 (1975). 34. STUART, K. L. Cardiomyopathy of pregnancy and the puerperium. Quarterly journal of medicine, 37: 463 (1968). 35. PARRY, E. H. 0. Idiopathic cardiomegaly, clinical diagnosis. Cardiologia, 52: 36 (1968). 36. FoSTER, R. M. The possible role of hypertension in unexplained forms of heart failure. East African medicaljournal, 42: 661 (1965). 37. BROCKINGTON, I. F. & BOHRER, S. P. Enlargement of the aortic shadow in Nigerian heart muscle disease. Acta cardiologica, 25: 344 (1970). 38. OAKELY, C. ET AL. Debate: that congestive cardio- myopathy is really hypertensive heart disease in disguise. Postgraduate medical journal, 48: 777 (1972). 39. IKEME, A. C. The pathogenesis of idiopathic cardio- megaly. In: Bennett, I. et al., ed. Degenerative dis- orders in the African environment: epidemiology and consequences. Proceedings of the 22nd Annual Scientific Conference of the East African Medical Research Council. Nairobi, East African Literature Bureau, 1976. 40. IKEME, A. C. El AL. The role of hypertension in the pathogenesis of idiopathic cardiomegaly in the tropics. Tropical cardiology, 1: 135 (1975). 41. DAVIES, J. N. P. Primary cardiomyopathies and congestive heart failure in clinical medicine: the African cardiomyopathies. Annals of the New York Academy of Sciences, 156: 498 (1969). 42. FRIEDBERG, C. K. Diseases of the heart, 3rd ed. Philadelphia, Saunders, 1976, p. 1474. 43. BASIL, U. ET AL. Thiamine deficiency and idiopathic cardiomegaly in Nigerian adults. African journal of medical sciences, 4: 465 (1973). 44. BRANDS, V. ET AL. Red cell thiamine concentration in idiopathic cardiomegaly. South African journal of medical sciences, 30: 64 (1965). 45. BRINK, A. J. ET AL. Thiamine deficiency and beri-beri heart disease. South African medical journal, 40: 581 (1966). 46. EMMANUEL, R. Familial cardiomyopathies. Post- graduate medical journal, 48: 742 (1972). IDIOPATHIC CARDIOMEGALY 461 47. OAKELY, C. Clinical definitions and classification of cardiomyopathies. Postgraduate medical journal, 48: 703 (1972). 48. SMYTHE, P. M. ET AL. The heart in kwashiorkor. British medical journal, 1: 67 (1962). 49. WHARTON, B. A. ET AL. The myocardium in kwash- iorkor. Quarterly journal of medicine, 38: 107 (1969). 50. PIZA, J. ET AL. Myocardial lesions and heart failure in infantile malnutrition. American journal of tropical medicine and hygiene. 20: 343 (1971). 51. OLIVER, I. ET AL. Cardiomyopathy with protein- losing enteropathy. Chest, 64: 513 (1973). 52. MEERSON, F. Z. The myocardium in hyperfunction, hypertrophy, and heart failure. Circulation research, 25 (Suppl. 2): 1 (1969). 53. ZUHLKE, V. ET AL. Inhibition of protein synthesis in cardiac hypertrophy and its relation to myocardial failure. Circulation research, 18: 558 (1966). 54. FOLLIS, R. H. In: Deficiency disease. Springfield IL, Thomas, 1958, p. 577. 55. SEFrEL, H. C. Cardiomyopathies in the Johannes- burg Bantu: Part II. Aetiology of idiopathic cardio- megaly. South African medical journal, 46: 1823 (1972). 56. SEFTEL, H. C. & SUSSER, M. Maternity and myocar- dial failure in African women. British heart journal, 23: 43 (1961). 57. GRUSIN, H. Acute reversible heart failure in Afri- cans. Circulation, 26: 27 (1957). 58. BARNEs, C. G. Medical disorders in obstetric prac- tice. Oxford, Blackwell Scientific Publications, 1965. 59. CHANARIN, J. The megaloblastic anaemias. Oxford, Blackwell Scientific Publications, 1969. 60. REID, J. V. 0. & BERJAK, P. Tryptophan and serotonin levels in patients with, or susceptible to, African cardiomyopathy. American heart journal, 74: 337 (1967). 61. REID, J. V. 0. Trial of tryptophan prophylaxis in patients liable to African cardiomyopathy. South African medicaljournal, 44: 732 (1970). 62. BLEGEN, S. D. Postpartum congestive heart failure. Acta medica Scandinavica, 178: 515 (1965). 63. STUART, K. L. & BRAs, C. Prognosis of idiopathic cardiomegaly in Jamaica with reference to the coro- nary arteries and other factors. British heart journal, 33 (Suppl.): 187 (1971).