530 Am. J. Trop. Med. Hyg., 59(4), 1998, pp. 530–534 Copyright  1998 by The American Society of Tropical Medicine and Hygiene EARLY ELECTROCARDIOGRAPHIC ABNORMALITIES IN TRYPANOSOMA CRUZI–SEROPOSITIVE CHILDREN ANA LUCIA SAMPAIO SGAMBATTI DE ANDRADE, FABIO ZICKER, ANIS RASSI, ALEXANDRE GABRIEL RASSI, RENATO MAURICIO OLIVEIRA, SIMONNE ALMEIDA SILVA, SORAYA SGAMBATTI DE ANDRADE, AND CELINA MARIA TURCHI MARTELLI Communicable Diseases Program, Pan American Health Organization, Washington, District of Columbia; Instituto de Patologia Tropical e Sau ´de Pu ´blica, Departamento de Medicina Tropical Sau ´de Coletiva e Dermatologia, e Faculdade de Medicina, Universidade Federal de Goia ´s, Goia ˆnia, Goia ´s Brazil; Special Program for Research and Training in Tropical Diseases, World Health Organization, Geneva, Switzerland; Hospital Sa ˜o Salvador, Setor Oeste, Goia ˆnia, Brazil Abstract. As part of a major epidemiologic study on Chagas’ disease, we compared the prevalence of electro- cardiographic (ECG) abnormalities among 141 school children 7–12 years of age and seropositive for Trypanosoma cruzi, and 282 age-, sex-, and school-matched seronegative children in an endemic area in Brazil. The prevalence of ECG abnormalities was 11.3% among seropositive children and 3.5% among seronegative children (odds ratio = 3.5, 95% confidence interval [CI] = 1.5–8.4). The prevalence rate of ECG alterations was 10.7% for seropositive males versus 8.9% for seropositive females. Complete right bundle branch block (CRBBB), which is highly suggestive of Chagas’ disease cardiopathy, was diagnosed in nine (6.4%) seropositive children and in only one (0.3%) seronegative child (odds ratio = 18.5, 95% CI = 2.3–146.5, attributable fraction = 58.3%). Five incident new cases of CRBBB were diagnosed after a 36-month follow-up of seropositive children who were enrolled in an independent clinical field trial. No case of frequent and/or multifocal ventricular premature beats was found in the cohort of children. The surprisingly high frequency of early ECG abnormalities, which indicates a rapid evolution from infection to disease, suggests the existence of endemic areas with a particular accelerated disease progression that was not described before. Under such conditions, a public health chemotherapy program focusing on the treatment of young seropositive children would be recommended. The knowledge of the natural history of Chagas’ disease cardiopathy has been built on hospital- and clinic-based ob- servations and from a small number of population-based long-term investigations conducted in endemic areas. 1–3 The comparison between studies and the understanding of evolv- ing patterns of the disease, however, have been limited to some extent due to the geographic difference of vector spe- cies, levels of endemicity, and the lack of standard proce- dures for selecting and examining study participants. 4 In endemic areas, infection by Trypanosoma cruzi occurs early in life; the prevalence rate of positive serology increas- es with age up to the fifth decade of life and decreases af- terwards. 5 Electrocardiographic (ECG) alterations are usu- ally the first clinical evidence of disease progression, but the onset of these abnormalities is rarely detected, since they tend to be asymptomatic at the beginning and heart involve- ment is usually clinically detected in adult life. 6,7 In cross-sectional observations the prevalence of ECG ab- normalities increases with age with a peak at the fourth to fifth decade of life, while the differential risk of ECG ab- normalities comparing seropositive with seronegative sub- jects decreases with age. 8 Cardiopathy seems to occur in family clusters, but no single risk factor has been clearly associated with disease progression at individual or popu- lation levels. 4 It is estimated that approximately 20–30% of seropositive subjects will develop heart disease during the course of their lives. 9 The prevalence of heart disease in rural populations tends to be underestimated due to selective ex- clusion from the endemic communities of diseased individ- uals who migrate for medical treatment or have died. We report here the results of an ECG cross-sectional sur- vey carried out among seropositive and seronegative school children, and a subsequent ECG evaluation of a cohort of seropositive children, who enrolled into a three-year clinical follow-up in an endemic area in Brazil. 10 METHODS Study area and population. The study protocol was eth- ically and technically reviewed and approved by the Re- gional Medical Council in accordance with World Health Organization guidelines for biomedical research. Signed in- formed consent was obtained from the parents or guardians of each child. Participants were recruited from children at- tending 60 rural schools in three small municipalities (Posse, Guarani de Goias, and Simolandia) in the northern part of Goias State in central Brazil, where T. cruzi infection rates for children 7–12 years old ranged from 12% to 23%, as reported in a national survey, 1975–1980. 11 From March to September 1991, a total of 1,900 school children were examined. Eluates of blood samples collected onto filter paper were processed for anti-T. cruzi anti- bodies simultaneously by indirect immunofluorescence, in- direct hemagglutination, and ELISA. Details of this screen- ing and methods used have been described elsewhere. 12 The study included 141 of 153 children who were positive by all three tests on eluates and later confirmed by testing serum samples collected by venipuncture. An age- (7–9 and 10–12 years), sex-, and school-matched group of 282 seronegative children (two seronegative to one seropositive) were ran- domly selected from the screened population as a compari- son group. This procedure was achieved by using a com- puter-generated stratified assignment into subgroups of age, sex, and school, and individually matching a seropositive child to the next two seronegative children in numerical or- der of screening. Data collection. Parents of all children were interviewed to collect information on living conditions and child’s med- ical history. Anthropometric measures were recorded: height measured to the nearest 0.5 cm and weight to the nearest 100 grams. All participants were physically examined and