Am. J. Trop. Med. Hyg., 77(3), 2007, pp. 495–499
Copyright © 2007 by The American Society of Tropical Medicine and Hygiene
Electrocardiographic Characteristics in a Population with High Rates of Seropositivity
for Trypanosoma cruzi Infection
S. Williams-Blangero,* T. Magalhaes, E. Rainwater, J. Blangero, R. Corrêa-Oliveira, and J. L. VandeBerg
Department of Genetics, Southwest Foundation for Biomedical Research, San Antonio, Texas; Centro de Pesquisas Rene Rachou,
Fundação Oswaldo Cruz, Belo Horizonte, Minas Gerais, Brazil; Southwest National Primate Research Center, San Antonio, Texas
Abstract. This study was conducted in Posse, a rural community in Goiàs, Brazil. Persons were recruited into the
study through house-to-house sampling of all houses in the sampled area. Blood samples were collected for seropositivity
assessments for Trypanosoma cruzi and an electrocardiogram was assessed using a portable system. The results demonstrate significant differences between seropositive and seronegative persons for electrocardiographic (ECG)–derived
traits. Seropositive persons had substantially longer QRS and QT intervals than seronegative persons. The PR interval
was significantly different between seropositive and seronegative persons. Conduction abnormalities were observed
more frequently in seropositive than seronegative persons. Right bundle branch block, an ECG abnormality typical of
Chagas disease, was observed in 15% of seropositive persons compared with less than 1% of seronegative persons.
Results indicate that T. cruzi infection and subsequent Chagas disease will continue to be major health problems for the
foreseeable future in this typical rural area of Brazil.
disease.29,30 Statistics indicate that more than half of those
infected with T. cruzi can expect to progress to chronic Chagas disease and most will experience the cardiac form of the
disease.5
There are few large-scale studies that have assessed ECGrelated variables in populations in areas endemic for Chagas
disease (however, see the study by Maguire and others28).
The purpose of the current study was to assess the epidemiology of cardiac disease as shown by ECG characteristics in a
population with high rates of infection with T. cruzi.
INTRODUCTION
Chagas disease is the leading cause of heart disease in Latin
America, affecting approximately 17 million people.1–4 Trypanosoma cruzi, the parasitic cause of Chagas disease, is
transmitted to the human host by reduiivid bugs that live in
the walls and roofs of houses in many rural areas of Latin
America.5 There are no effective therapies for the chronic
long-term phase of the disease and there are no vaccines or
drugs to prevent infection with T. cruzi.
Substantial resources have been dedicated to eliminating
the vector from household structures in Latin America. However, active transmission occurs in many areas,6–11 and between 100 and 120 million persons are considered to be at risk
for infection with T. cruzi.2 Approximately 300,000 persons
are newly infected each year2 and approximately 60% of
those infected with T. cruzi will experience a long-term
chronic disease characterized by highly variable cardiac outcomes, and less frequently by mega-colon and/or megaesophagus.5
The presence of active transmission combined with the current lack of vaccines or effective prophylactic drugs5,12–14
means that Chagas disease will remain a significant health
burden and major economic drain in South America for the
foreseeable future. In addition, Chagas disease is an emerging
health threat in the United States and Canada, particularly in
regions with large South and Central American immigrant
populations.15–21
The cardiac form of Chagas disease is evidenced in electrocardiographic (ECG) data by conduction abnormalities
that include bradyarrythmias, premature ventricular contractions, and bundle branch blocks, particularly right bundle
branch blocks.5,22 Case-control studies have found that right
bundle branch block, left anterior hemiblock, and atrioventricular blocks occur at a much higher frequency in seropositive persons than in seronegative persons.23–28 An elevated
Q-T interval and T-wave axis deviation have both been implicated as predictive of mortality in patients with Chagas
MATERIALS AND METHODS
The study was conducted in Posse, a rural community in the
state of Goiás, Brazil located approximately 350 km northeast
of Brasilia, the capital of Brazil. The population consists primarily of subsistence farmers, most of whom were born in the
municipality. The population is highly admixed and derived
from European, African, and Amerindian ancestry.
Persons were recruited into the study through house-tohouse sampling using a protocol that was reviewed and approved by the Institutional Review Boards of the University
of Texas Health Science Center in San Antonio and the Rene
Rachou Research Center of the Oswaldo Cruz Foundation in
Belo Horizonte, Minas Gerais, Brazil. The sample is population based and includes all eligible adults in the sampling area
who were willing to participate. The study was explained at
each household and persons who elected to participate provided a consent form that was either signed or fingerprinted.
A total of 1,389 persons were successfully recruited.
A 10-mL blood sample was collected into a red-top tube for
assessment of seropositivity status. Samples were allowed to
clot, and then were kept on wet ice until they were separated
into serum and clot no more than eight hours after collection.
Serum samples were then aliquotted and frozen in liquid nitrogen until transfer to the laboratories in Belo Horizonte and
San Antonio on dry ice.
Serum samples from all persons were assessed by three
standardized tests (enzyme-linked immunosorbent assay
[ELISA], hemagglutination, and immunofluorescence) by the
Laboratory of Cellular and Molecular Immunology at the
Rene Rachou Research Center, FIOCRUZ. Persons were
considered seropositive if two or more of these standardized
* Address correspondence to S. Williams-Blangero, Department of
Genetics, Southwest Foundation for Biomedical Research, PO Box
760549, San Antonio, TX 78245-0549. E-mail: [email protected]
495
496
WILLIAMS-BLANGERO AND OTHERS
tests indicated seropositivity for T. cruzi infection. A total of
1,190 persons (86%) were typed with all three tests, and were
determined to be positive if two or more tests showed positive
results. The remaining persons were scored based on the results of two tests: 121 persons were scored as negative based
on results of the hemagglutination and immunofluoresence
tests, 3 persons were scored as positive on the basis of positive
ELISA and immunofluorescence test results, and 75 persons
were scored as positive on the basis of positive hemagglutination and immunofluorescence test results.
The ECGs were collected from all participants using the
portable Marquette MAC5000 System (GE Medical Systems
Information Technologies, Milwaukee, WI). In addition to
the automated quantitative data output by the machine,
qualitative ECG variables were verified by a clinical cardiologist experienced in the treatment of Chagas disease.
For the quantitative ECG traits, statistical analysis using a
linear model was performed to test for differences between
seropositive and seronegative persons. All analyses were performed simultaneously adjusting for the effects of several covariates including sex, age, age2, sex × age, and sex × age2.
Results are provided both as the covariate-adjusted means for
seropositive and seronegative persons and as standardized
differences measured in standard deviation units. The latter
index allows for easier comparison across variables and provides a measure of the biologic importance of a given statistical difference. Because the data set consists of large numbers of related persons, we explicitly allowed for kinshipbased non-independence by performing all analyses
conditional upon the known pedigree structure. Nonindependence was modeled through a familiality parameter
that was treated as a nuisance parameter given that the focus
of the current report is on the changes in mean effects induced by T. cruzi infection. Dichotomous discrete ECG variables such as bundle branch abnormalities were similarly analyzed using a probit regression model allowing for nonindependence due to kinship. The program SOLAR,31 which
enables arbitrary modeling of mean effects in the presence of
structured non-independence, was used for all statistical
analyses.
RESULTS
Seropositivity and ECG measures were available for 1,389
persons. There were 690 males and 699 females. These persons were members of 110 families although almost all (1,193)
can be placed into a single, large, extended family. The average age of the persons was 41.9 years. The rate of T. cruzi
seropositivity observed was high in this population. There
were 722 seropositive persons and 667 seronegative persons;
52% of the sampled persons were seropositive for T. cruzi
infection.
A number of traits were determined from the ECG readings. Quantitative intervals were determined directly from the
ECG. The quantitative intervals evaluated included the QRS
interval, the QT interval, and the PR interval. The mean values for each interval in seropositive and seronegative persons
are shown in Table 1. Also shown is the difference in standard
deviation units between seropositive and seronegative persons.
These results show that there were significant differences
TABLE 1
Mean electrocardiographic (ECG) parameters in Trypanosoma
cruzi–seropositive and -seronegative persons adjusted for age
and sex*
ECG trait
QRS
QT
PR
Ventricular
rate
Mean value
in seronegative
persons
Mean value
in seropositive
persons
91.08 ms
382.49 ms
152.15 ms
98.55 ms
396.97 ms
155.73 ms
70.21 bpm
67.04 bpm
Difference
in SDUs
P
0.4423
0.3940
0.1376
9.3 × 10−12
5.1 × 10−10
0.0275
−0.2460
0.00004
* SDUs ⳱ standard deviation units.
between seropositive and seronegative persons for most of
the quantitative ECG variables that were assessed in the
population. All quantitative ECG-derived variables showed
significant differences between seropostive and seronegative
persons. Seropositive persons had substantially longer QRS
and QT intervals than seronegative persons. The observed
highly significant differences are of a magnitude approximately 0.4 of a standard deviation for these traits. Such a large
standardized difference is quite striking. The PR interval was
also significantly different in length between seropositive and
seronegative persons. However, this increase is rather modest
and represents a displacement of approximately 0.14 of a
standard deviation unit. Seropositive persons also had significantly lower ventricular rates than persons seronegative for T.
cruzi infection. This highly significant difference in ventricular rate represents a displacement of the seropositive mean
approaching nearly a 0.75 of a standard deviation unit. Taken
together, these results on a large population show dramatic
differences in cardiovascular-related parameters between seropositive and seronegative persons.
We also assessed the frequency of a variety of qualitative
Chagas disease–related traits in seropositive and seronegative
persons. These traits are shown in Table 2. As is the case with
the quantitative ECG measures, there were significant differences between persons who were seropositive for T. cruzi
infection and uninfected persons.
Conduction abnormalities were more frequently observed
in infected than uninfected persons. For example, right
bundle branch block is an ECG characteristic that is typical of
Chagas disease.5 As expected, right bundle branch block is
TABLE 2
Frequency of qualitative electrocardiographic (ECG) traits in seropositive and seronegative persons adjusted for age and sex
ECG trait
Right bundle
branch block
Bifascicular block
Left anterior
fascicular block
Marked sinus
bradycardia
Left axis deviation
Left ventricular
hypertrophy
Atrial fibrillation
Abnormal ECG
Frequency in
seronegative persons
Frequency in
seropositive persons
P
0.0090
0.0000
0.1524
0.0526
6.0 × 10−20
2.8 × 10−9
0.0045
0.0789
8.6 × 10−7
0.0150
0.0150
0.0249
0.0540
0.3874
0.0060
0.0525
0.0015
0.1829
0.0845
0.0069
0.4349
0.7128
0.2248
1.2 × 10−9
ELECTROCARDIOGRAPHIC TRAITS IN A POPULATION INFECTED WITH T. CRUZI
much more frequent in seropositive persons than in seronegative persons, with less than 1% of seronegative persons
having a right bundle branch block compared with 15% of
seropositive persons having this branch block. This difference
was highly significant (P ⳱ 6 × 10−20). Thus, right bundle
branch block appears to be closely related to infection in this
population.
Figure 1 shows the cumulative proportion of persons with
right bundle branch block by age in males and females. There
was no significant difference in the frequency of right bundle
branch block between males and females in seronegative persons. However, in seropositive persons, right bundle branch
block increases significantly with age, and also shows a substantial difference between males and females. Among seropositive persons, males are more likely to develop a right
bundle branch block than females.
Other conduction abnormalities are altered by T. cruzi infection. As shown in Table 1, bifascicular block, left anterior
fascicular block, and left axis deviation are present in seropositive persons at significantly higher frequencies than in
seronegative persons. Significant differences in the frequencies of marked sinus bradycardia, left ventricular hypertrophy, and atrial fibrillation were not observed. However, for
each of these abnormalities, infected persons showed nominally higher frequencies than uninfected persons. There was a
substantial difference in the frequency of ECG abnormalities,
with seropositive persons having a prevalence of approximately 43% versus only 18% in seronegative persons (P ⳱
1 × 10−9). These results clearly demonstrate the effect of T.
cruzi infection on sub-clinical heart abnormalities.
Figure 2 shows the plot of presence of any type of conduction block by age in seropositive and seronegative persons
from the Posse population. Seropositive persons had much
higher rates of conduction blocks at all ages. Although some
conduction blocks are expected in uninfected persons, sero-
FIGURE 1. Cumulative proportion of individuals with right bundle
branch block by age and sex in persons seropositive and seronegative
for Trypanosoma cruzi. The difference in prevalence between seropositive and seronegative persons was highly significant (P ⳱ 6 ×
10−20). Age at electrocardiographic (ECG) testing is in years.
497
FIGURE 2. Cumulative proportion of persons with any electrocardiographic (ECG) abnormality by age (years) and sex in persons
seropositive and seronegative for Typanosoma cruzi. The difference
in prevalence of an ECG abnormality between seropositive and seronegative persons was highly significant (P ⳱ 1.2 × 10−9).
positive persons have higher prevalences of conduction
blocks at all ages.
DISCUSSION
The study region in Posse is typical of many rural areas in
central Brazil. This large-scale study of the local population
provides insights into the implications of Chagas disease for
local health services that surveys of cases and controls conducted in hospital settings cannot. The high rates of seropositivity in this rural area indicate that Chagas disease will be a
continuing public health problem in the region for the foreseeable future. Despite intensive vector control efforts, Chagas disease will persist as a major consumer of health care
resources because of the large population of persons who are
already infected. Congenital transmission can result in new
infections in regions with no vector-transmitted disease.32 Active transmission still occurs in many areas and congenital
transmission also can cause infections. With the decreased
emphasis on vector control in Brazil, it is possible that active
transmission will recur in areas that were previously considered to be controlled.
The population variation in QT and QRS intervals may be
informative in predicting the future health care needs in the
region. Several investigators have noted the potential value of
QT interval parameters for predicting outcome in Chagas disease.29,33 The QT and QRS intervals were significantly increased in seropositive persons in the Posse population. As
previously reported,34–36 persons with ECG abnormalities
predictive of a negative outcome in Chagas disease may be
apparently healthy. Population level surveys conducted
through general health clinics may be an excellent mechanism
for identifying asymptomatic seropositive persons whose cardiac health should be monitored on a regular basis.
Epidemiologic studies of European populations that were
498
WILLIAMS-BLANGERO AND OTHERS
not exposed to T. cruzi infection showed that ECG abnormalities are more common in males than in females.36 That
pattern was seen in the Posse population in seronegative persons. Interestingly, the divergence in prevalence of abnormalities between males and females was greater in seropositive persons. This finding may be reflective of increased
physical labor by males in this agricultural population, which
can put additional strain on the cardiovascular system. In
particular, right bundle branch blocks are more frequent in
seropositive persons than seronegative persons. The difference in ECG abnormalities between seropositive males and
females reflects the general trend for increased rates of abnormalities in men.
In conclusion, population level surveys of seropositivity for
T. cruzi infection and ECG variables are useful in predicting
the future health care needs of local populations in Brazil.
The epidemiologic data on seropositivity and ECG variables
in the Posse population clearly indicate that Chagas disease
will persist for a long time as a major public health burden.
Despite the successes of the vector control programs in diminishing active transmission of T. cruzi when assessed at the
national level, local populations in areas that are endemic for
Chagas disease are still struggling with this devastating disease and will continue to do so for the foreseeable future.
Received February 11, 2007. Accepted for publication May 27, 2007.
Acknowledgments: We thank Antonio R.L. Teixeira for his contributions to initiating this project, FIOCRUZ and the Brazilian Ministry of Health for logistical assistance, and the people of Posse for
their generous cooperation with this long-term study.
Financial support: This research was supported by grants HL66480 to
S. Williams-Blangero and MH59490 to J. Blangero from the National
Institutes of Health. This work was conducted in part in facilities
constructed with support from the Research Facilities Improvement
Program under grant C06 RR017515 from the National Center for
Research Resources of the National Institutes of Health.
Authors’ addresses: S. Williams-Blangero, E. Rainwater, and J.
Blangero, Department of Genetics, Southwest Foundation for Biomedical Research, PO Box 760549, San Antonio, TX 78245-0549. T.
Magalhaes and R. Corrêa-Oliveira, Centro de Pesquisas Rene Rachou, Fundação Oswaldo Cruz, Belo Horizonte, Minas Gerais, Brazil. J. L. VandeBerg, Department of Genetics, Southwest Foundation
for Biomedical Research, PO Box 760549, San Antonio, TX 782450549 and Southwest National Primate Research Center, PO Box
760549, San Antonio, TX 78245-0549.
REFERENCES
1. Dias JC, Silveira AC, Schofield CJ, 2002. Impact of Chagas disease control in Latin America: a review. Mem Inst Oswaldo
Cruz 97: 603–612.
2. World Health Organization, 2002. Control of Chagas disease.
World Health Organ Tech Rep Ser 905: 1–109.
3. Moncayo A, 2003. Chagas disease: current epidemiological trends
after the interruption of vectoral and transfusional transmission in the Southern Cone countries. Mem Inst Oswaldo Cruz
98: 577–591.
4. Dutra WO, Rocha MO, Teixeira MM, 2005. The clinical immunology of human Chagas disease. Trends Parasitol 21: 581–587.
5. Kirchhoff LV, 1999. American trypanosomiasis (Chagas disease).
Guerrant RL, Walker DH, Weller PF, eds. Tropical Infectious
Diseases: Principles, Pathogens, and Practice. New York:
Churchill Livingstone, 785–796.
6. Coll-Cardenas R, Espinoza-Gomez F, Maldonado-Rodriguez A,
Reyes-Lopez PA, Huerta-Viera M, Rojas-Larios F, 2004. Active transmission of human Chagas disease in Colima, Mexico.
Mem Inst Oswaldo Cruz 99: 363–368.
7. Rizzo NR, Arana BA, Diaz A, Cordon-Rosales C, Klein RE,
Powell MR, 2003. Seroprevalence of Trypanosoma cruzi infection among school-age children in the endemic area of Guatemala. Am J Trop Med Hyg 68: 678–682.
8. Grijalva MJ, Escalante L, Paarades RA, Costales JA, Padilla A,
Rowland EC, Aguilar JM, Racines J, 2003. Seroprevalence and
risk factors for Trypanosoma cruzi infection in the Amazon
region of Ecuador. Am J Trop Med Hyg 69: 380–385.
9. Pinto AY, Valente SA, Valente VC, 2004. Emerging acute Chagas disease in Amazonian Brazil: case reports with serious
cardiac involvement. Braz J Infect Dis 8: 454–460.
10. Gurtler RE, Cecere MC, Lauricella MA, Petersen RM, Chuit R,
Segura EL, Cohen JE, 2005. Incidence of Trypanosoma cruzi
infection among children following domestic reinfestation after insecticide spraying in rural northwestern Argentina. Am J
Trop Med Hyg 73: 95–103.
11. Borges JS, Machada de Assis GF, Gomes LV, Dias JC, Pinto ID,
Martins-Filho OA, Torres RM, Vinas PA, Bahia MT,
Machado-Coelho GL, Lana M, 2006. Seroprevalence of Chagas disease in school children from two municipalities of
Jequitinhonha Valley, Minas Gerais, Brazil: six years following
the onset of epidemiological surveillance. Rev Inst Med Trop
Sao Paulo 48: 81–86.
12. Garg N, Bhatia V, 2005. Current status and future prospects for
a vaccine against American trypanosomiasis. Expert Rev Vaccines 4: 867–880.
13. Urbino JA, Docompo R, 2004. Specific chemotherapy of Chagas
disease: controversies and advances. Trends Parasitol 19: 495–
501.
14. Kirchhoff LV, 2003. Changing epidemiology and approaches to
therapy for Chagas disease. Curr Infect Dis Rep 5: 59–65.
15. McCarthy M, 2003. American Red Cross to screen blood for
Chagas disease. Lancet 362: 1988.
16. Busch MP, Kleinman SH, Nemo GJ, 2003. Current and emerging
infectious risks of blood transfusions. JAMA 289: 959–961.
17. Beard CB, Pye G, Steurer FJ, Rodrigues R, Campman R, Peterson AT, Ramsey J, Wirtz RA, Robinson LE, 2003. Chagas
disease in a domestic transmission cycle, southern Texas, USA.
Emerg Infect Dis 9: 103–105.
18. DiPentima MC, Hwang LY, Skeeter CM, Edwards MS, 1999.
Prevalence of antibody to Trypanosoma cruzi in pregnant Hispanic women in Houston. Clin Infect Dis 28: 1281–1285.
19. Leiby DA, Herron RM, Read EJ, Lenes BA, Stumpf RJ, 2002.
Trypanosoma cruzi in Los Angeles and Miami blood donors:
impact of evolving donor demographics on seroprevalence and
implications for transfusion transmission. Transfusion 42: 549–
555.
20. Centers for Disease Control and Prevention, 2002. Chagas disease after organ transplantation—United States, 2001. MMWR
Morb Mortal Wkly Rep 51: 210–212.
21. Kirchhoff LV, Weiss LM, Wittner M, Tanowitz HB, 2004. Parasitic diseases of the heart. Front Biosci 9: 706–723.
22. Prata A, 2001. Clinical and epidemiological aspects of Chagas
disease. Lancet Infect Dis 1: 92–100.
23. Goldbaum M, Ajimura FY, Litvoc J, Alves de Carvalho S, ElufNeto J, 2004. American trypansomiasis and electrocardiographic alterations among industrial workers in Sao Paulo,
Brazil. Rev Inst Med Trop Sao Paulo 46: 299–302.
24. Yacoub S, Birks EJ, Slavik Z, Henein M, 2003. Early detection of
myocardial dysfunction in Chagas disease using novel echocardiographic indices. Trans R Soc Trop Med Hyg 97: 528–534.
25. Sosa-Hurado F, Mazariego-Aranda M, Hernandez-Becerril N,
Garza-Murillo V, Cardenas M, Reyes PA, Hirayama K, Monteon VM, 2003. Electrocardiographic findings in Mexican Chagasic subjects living in high and low endemic regions of Trypansoma cruzi infection. Mem Inst Oswaldo Cruz 98: 605–610.
26. Rangel-Flores H, Sánchez B, Mendoza-Duarte J, Barbnabe C,
Breniere FS, Ramos C, Espinoza B, 2001. Serologic and parasitologic demonstration of Trypansoma cruzi infections in an
urban area of central Mexico: correlation with electrocardiographic alterations. Am J Trop Med Hyg 65: 887–895.
27. Jorge MT, Macedo TA, Janones RS, Carizzi DP, Heredia RA,
Acha RE, 2003. Types of arrhythmia among cases of American
trypanosomiasis compared with those in other cardiology patients. Ann Trop Med Parasitol 97: 139–148.
ELECTROCARDIOGRAPHIC TRAITS IN A POPULATION INFECTED WITH T. CRUZI
28. Maguire JH, Hoff R, Sherlock I, Guimaraes AC, Sleigh AC,
Ramos NB, Mott KE, Weller TH, 1987. Cardiac morbidity and
mortality due to Chagas disease: prospective electrocardiographic study of a Brazilian community. Circulation 75: 1140–
1145.
29. Salles G, Xavier S, Sousa A, Hasslocher-Moreno A, Cardosa C,
2003. Prognostic value of QT interval parameters for mortality
risk stratification in Chagas disease: results of a long-term follow-up study. Circulation 108: 305–312.
30. Salles G, Xavier S, Sousa A, Hasslocher-Moreno A, Cardosa C,
2004. T-wave axis deviation as an independent predictor of
mortality in chronic Chagas disease. Am J Cardiol 93: 1136–
1140.
31. Almasy L, Blangero J, 1998. Multipoint quantitative trait linkage
analysis in general pedigrees. Am J Hum Genet 62: 1198–1211.
32. Brutus L, Schneider D, Postigo J, Delgado W, Mollinedo S, Chippaux JP, 2007. Evidence of congenital transmission of Trypa-
33.
34.
35.
36.
499
nosoma cruzi in a vector-free area of Bolivia. Trans R Soc
Trop Med Hyg: (Epub ahead of print).
Corbucci HA, Haber DM, Bestetti RB, Cordeiro JA, Fioroni
ML, 2006. QT interval dispersion in patients with chronic heart
failure secondary to Chagas cardiomyopathy: correlation with
clinical variables of prognostic significance. Cardiovasc Pathol
15: 18–23.
Soto-Rojas G, Cortes JM, Medrano GA, 1984. Electrocardiographic changes in 29 apparently healthy subjects with positive
serological tests for Chagas disease. Arch Inst Cardiol Mex 54:
579–583.
Pimenta J, Miranda M, Pereira CB, 1983. Electrophysiologic
findings in long-term asymptomatic chagasic individuals. Am
Heart J 106: 374–380.
DeBacquer D, DeBacker G, Kornitzer M, 2000. Prevalences of
ECG findings in large population based samples of men and
women. Heart 84: 625–633.