Continued assessment of temporal trends in mortality and epidemiology of specific cardiovascular diseases in South America is needed to provide a scientific basis for rational allocation of the limited healthcare resources and introduction of strategies to reduce risk and predict the future burden of cardiovascular disease. The epidemiology of cardiomyopathies, adult valve disease and heart failure (HF) in South America is reviewed here. Diseases of the circulatory system are the main cause of death based on data from about 50% of the South American population. Among the cardiovascular causes of death, cerebrovascular disease is predominant followed by ischaemic heart disease, other heart diseases and hypertensive disease. Of note, cerebrovascular disease is the main cause of death in women, and race also influenced cardiovascular mortality rates. HF is the most important cardiovascular reason for admission to hospital due to cardiovascular disease of ischaemic, idiopathic dilated cardiomyopathic, valvular, hypertensive and chagasic aetiologies. Also, mortality due to HF is high, especially owing to Chagas’ disease. HF and aetiologies associated with HF are responsible for 6.3% of deaths. Rheumatic fever is the leading cause of valvular heart disease. The findings have important public health implications because the allocation of healthcare resources, and strategies to reduce the risk of HF should also consider controlling Chagas’ disease and rheumatic fever in South American countries.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
In developed countries reduction in mortality from cardiovascular disease (CVD) has been reported over the past decades.1 2 However, major differences in cardiovascular event rates were reported in different populations.3 Although general trends in mortality from coronary heart disease (CHD) and CVD were favourable in several developed areas of the world, major geographical differences were reported, and in some eastern European countries, mortality from CHD and CVD remains exceedingly high.4 Despite this information, chronic diseases have been considered a public health problem only in developed countries and among the elderly.5
This review was undertaken to examine trends in cardiovascular mortality and death from all causes in Brazil, the largest country in South America, focusing on the burden of cardiomyopathies, adult valve disease and heart failure (HF) in South America.
A comprehensive literature search was performed using electronic bibliographic databases (eg, ClinicalTrials.gov, PubMed, Cochrane, Elsevier/ScienceDirect, SciELO, LILACS) and the following keywords: cardiomyopathy, heart failure, alcoholic cardiomyopathy, arrhythmogenic right ventricular cardiomyopathy, hypertrophic cardiomyopathy, restrictive cardiomyopathy, infiltrative cardiomyopathy, amyloidosis, endomyocardial fibrosis, valve disease and rheumatic fever. We limited this search to articles and trials from South America. The review was based on selected articles between 1998 and 2008. Reference lists from selected articles and reviews were also examined for further relevant articles, and we have cited those with information about epidemiology. Mortality ratios were derived from the Brazilian Ministry of Health data available electronically from 1983 to 2003.6 Also, data concerning HF and cardiomyopathy as causes of death in 2006 in the São Paulo State (Brazil) were obtained from the SEADE Foundation (Fundação Sistema Estadual de Análise de Dados).7
TOTAL MORTALITY AND CVD MORTALITY
The leading cause of death in Brazil was diseases of the circulatory system. The most important cause of cardiovascular death was cerebrovascular disease followed by ischaemic heart disease, other heart disease, acute myocardial infarction, hypertensive disease, atherosclerosis and rheumatic disease (fig 1).6
Analysis using Poisson regression models demonstrated an increase in ischaemic heart disease (p<0.001), acute myocardial infarction (p<0.001) and hypertensive disease (p<0.001). Reduction in mortality was observed for cerebrovascular disease (p<0.001), other heart diseases (p<0.001), atherosclerosis (p<0.001) and rheumatic disease (p<0.001) (fig 1).6
The lifetime risk of HF is a public health issue.8 Most knowledge of HF is derived from North American and European studies. In contrast, in the developing world no population-based study has been reported, and scarce information comes from data gathered in clinical trials or hospital-based studies.9 Data obtained from the SEADE Foundation on the mortality in São Paulo State (Brazil) with an estimated 41 654 020 inhabitants and limited to year 2006, demonstrated that HF or aetiologies associated with HF, except primary valvular disease, are responsible for 6.3% of the total deaths (table 1, fig 2).7 Thus, the real magnitude of the HF burden in South America is partially known (table 2). However, the increase in life expectancy in developing countries implies a rise in the incidence and prevalence of HF globally.10
The reduction in mortality related to improvement in prevention and treatment of acute coronary syndromes, coupled with an increase of overall life expectancy has led to a relative reduction in the prevalence of HF, especially from ischaemic cardiomyopathy.11 In developing countries an ischaemic aetiology is also important, but to a lesser extent. Other causes such as rheumatic (Africa, Asia, Oceania and Latin America), chagasic (Latin America) and hypertensive (African and African-Americans) are also prominent (table 1, fig 2). In Brazil and all of South America epidemiological data for HF are scarce and mostly related to populations from reference centres comprising inpatients and outpatients receiving ambulatory care.9 Also, the relative prevalence of systolic HF and HF with preserved systolic left ventricular function is not well documented. In a small hospitalised elderly Argentinian population with HF, HF with preserved systolic function was more common, especially in women.12
STUDIES OF HF INPATIENTS
The most robust Brazilian population information comes from the Ministry of Health (DATASUS) and is related to the number of admissions to hospital due to HF.6 Overall, cardiovascular diseases are the third most common cause of hospitalisation (fig 3). Of 743 763 admissions to hospital due to cardiovascular diseases in 2007, 39.4% were associated with HF (70% of the cases in the age group >60 years) (fig 4).6 Mortality ranged from 6.58% to 6.95% and the mean hospitalised period was 5.8 days.
In a Brazilian report including 903 patients from a tertiary centre admitted to hospital with a diagnosis of HF, the mean age was 53 years, with the ischaemic aetiology being the most common (34%), followed by the idiopathic (26%), valvular (22%), hypertensive (7%) and chagasic aetiology (6%).13 However, in a general hospital of Antigua,14 reporting on a population of 293 patients with HF as the diagnosis of discharge, hypertensive aetiology was the most prevalent aetiology (41%), followed by the ischaemic aetiology (33%).
An analysis of five Argentinian registries with decompensated HF over the past decade showed that only in the first study (from 1993) was the prevalence of hypertensive aetiology higher than that of ischaemic aetiology (32.3% vs 28.7%).15 In the other four studies the ischaemic aetiology (range 27.4–38.4%) was more common than the hypertensive aetiology (from 18.2–23.7%). In those studies the medium age was >60 years, with a high valvular aetiology (range 16.4–21.7%), and a low chagasic aetiology (range 1.3–8.4%). The presence of preserved systolic function in this population was reduced (varying from 20% to 36%).
The Brazilian EPICA–Niterói study compared decompensated HF populations from public versus private tertiary hospitals, showing a mean age of 61 years and 66% of ischaemic aetiology for public hospitals, and mean age 72 years with 62% of ischaemic aetiology for the private services.16 Another Brazilian study compared HF populations from tertiary hospitals in the south of Brazil and in northeastern USA reporting a high prevalence of ischaemic aetiology in both samples (39% Brazilian and 46% North American); the hypertensive aetiology amounted to 25% of the Brazilian and 14% of the North American population.17
The mean age in the Brazilian sample was higher than in the North American HF sample (73 vs 67 years). In another Brazilian study of 212 patients with decompensated HF from an emergency department of a tertiary hospital, the mean age was 60 years, with a high prevalence of the ischaemic aetiology (30%), followed by hypertensive (21%), valvular (15%) and chagasic (15%) aetiologies, and 45% of patients with preserved systolic function.18
The BELIEF multicentre study (Brazilian Evaluation of Levosimendan Infusion Efficacy) of 182 patients with decompensated HF recently published by Bocchi et al, reported a younger age (55 years) and more critically ill patients, since the required inclusion criteria included the need for vasoactive drugs.19 The HF aetiology was ischaemic in 34.1%, chagasic 21.4% and hypertensive in 13.2% of patients. The distribution of races was Caucasian in 58.2% and Afro-Brazilian in 36.8%.20
STUDIES OF HF OUTPATIENTS
In the 1990s, a Brazilian cohort study of 1220 outpatients followed up in a specialised HF clinic showed a mean age of 46 years; the idiopathic aetiology was present in 37%, followed by chagasic (20%), ischaemic (17%) and hypertensive in 14% of patients.21 Of note, the chagasic aetiology was associated with the worst prognosis.22 In contrast, in the Argentinean GESICA study 40% of the patients had previous myocardial infarction, the idiopathic aetiology was present in 20% and the chagasic in 10.5–8.1%.23 In Mexico, from a population sample of 45 patients with systolic HF, with a mean age of 61 years, at an HF clinic of a tertiary hospital, 47% of patients had ischaemic aetiology and 44% idiopathic aetiology.24
In the Brazilian REMADHE clinical trial of patients with systolic HF followed up at a tertiary hospital, Bocchi et al described a higher prevalence of ischaemic aetiology (22–28%). The chagasic aetiology was also quite common (21–16%), as well as the hypertensive (22–18%).25 The mean age was 51 years. Also, in the InCor-HCFMUSP HF and transplant outpatient clinic the aetiology was ischaemic in 28.2%, idiopathic in 28.2%, hypertensive in 20.6%, chagasic in 8.6%, tachycardiomyopathic in 2.1%, valvular in 6.5%, alcoholic in 2.1% and peripartum in 3.2%.26 At this clinic, comorbidities were found in significant proportions of patients: diabetes mellitus was diagnosed in 20.8%, chronic renal failure in 15.6%, dyslipidaemia in 28.9%, hypothyroidism in 9.3% and hyperuricaemia in 5.2%.
Community Latin-American HF studies are scarce. In a family health programme in the city of Niterói (Brazil) with 170 patients with a diagnosis of HF, mean age 61 years, there was a female predominance (58%), and the aetiology was hypertensive in 84%, and ischaemic in 21%.27 HF with preserved systolic function was seen in 64.2%. Smoking, coronary artery disease, diabetes mellitus and chronic renal failure were more prevalent in systolic HF. The similar prevalence, 86.1% in HF with preserved systolic function versus 86.4%, suggests that the systemic hypertension may be a risk factor in both types of HF.
PROGNOSIS OF HF
Prospective data about the prognosis of patients with HF in Latin America can be obtained only from clinical trials. Before β blockers were prescribed for HF, the GESICA trial reported a 24–42% mortality during 13 months’ follow-up of patients with advanced HF.23 In contrast, during the β-blocker era the Argentinian DIAL trial reported 15–16% mortality during a mean 16 months’ follow-up. More recently in the β-blocker era the Brazilian REMADHE trial reported 36–43% mortality during a mean (SD) follow-up of 2.47 (1.75) years.25 Thus, in the β-blocker era, mortality rates seen in Latin American trials are similar to those reported in the recently published ACCLAIM trial.28 Moreover, assessment of the clinical status of Brazilian HF outpatients in the InCor-HCFMUSP HF Clinic according the New York Heart Association functional class showed 32.3% in class I, 42.3% in II and 25.4% in III.26
MANAGEMENT OF HF
In South America no community-based studies of the management of HF have been published. The DIAL study on outpatients from the Argentine showed the use of diuretic agents in 82.5%, digoxin in 47%, amiodarone in 29.1%, spironolactone in 32.3%, angiotensin converting enzyme inhibitors in 79.6%, angiotensin receptor blockers in 13.4% and β blockers in 61.8%.29 The Brazilian data from the REMADHE trial showed prescription of amiodarone in 9.65%, amlodipine in 5.4%, angiotensin receptor blockers in 14.5%, oral anticoagulants in 13.7%, angiotensin converting enzyme inhibitors in 82.5%, β blockers in 66%, spironolactone in 55%, hydralazine in 6.4%, nitrates in 11.7%, digoxin in 63.1%, diuretic agents in 81.7%, statins in 12.9%, resynchronisation in 3.4% and a cardioverter–defibrillator in 0.64%.25 Also, data from the HF and transplant outpatient clinic of the Heart Institute of the University of São Paulo School of Medicine reported that most patients were receiving doses of drugs close to ideal, according to established guidelines.26 Renin–angiotensin–aldosterone system (RAAS) inhibitors were prescribed to 94.7% of patients, whereas the remainder received the hydralazine–nitrate combination owing to renal failure. A total of 84.7% received a β blocker; of those who were not using this class of drugs, 5.2% had stopped owing to clinical signs of intolerance to the drug (dizziness and hypotension), whereas the remainder presented clear contraindications to its use (bronchospasm and severe peripheral artery disease).
Overall, treatment of patients with HF in South America is not uniform. In comparison with data from the recent published ACCLAIM trial (Canada, Denmark, Germany, Israel, Norway, Poland and USA) the use of β blockers, lipid-lowering drugs, antiplatelet agents, anticoagulants, implanted cardioverter–defibrillator and cardiac resynchronisation is lower in South America.26 In addition, in patients admitted to hospital for treatment of decompensated HF the BELIEF study in Brazil reported a lower use of β blockers, nitrates, angiotensin receptor blockers, angiotensin converting enzyme inhibitors, and a higher use of spironolactone.19 Several factors are likely to explain these discrepancies between South America and the USA or Europe, including more limited resources, less adherence to guidelines and differences in aetiologies and racial factors.
VALVULAR HEART DISEASE
Rheumatic fever (RF) is the leading cause of valvular heart disease in Brazil, and is still widely prevalent.6 30 It is currently responsible for significant morbidity–mortality, being the cause of 90% of cardiac operations in children and over 30% of cardiac operations in adults, most of them of young age. An autoimmune disease caused by exposure to streptococcus, rheumatic fever has different clinical expressions geographically. This may be explained by the different strains of streptococci infecting a population of different genetic backgrounds and the influence of environmental factors.31 The prevalence of Streptococcus pyogenes may differ among Brazilian regions.32 Epidemiological features of group A Streptococcus from a paediatric population were very different in Belgian in comparison with a Brazilian city, not only in its clinical presentation, but also in its genetic diversity and distribution of emm patterns.33 Besides the traditional adaptative immune markers of susceptibility (such as HLA typing), it has recently has been shown that innate immunity markers such as mannose-binding lectin are correlated with an increased risk of developing chronic rheumatic heart disease and aortic regurgitation.34
Data from the Brazilian government health system show that RF and rheumatic heart disease (RHD) are still a major health problem, even in the 21st century (fig 1).6 There are annually 2200 hospital admissions for acute RF and it is possible to estimate that annually 73 000 patients have acute RF in Brazil. If we accept that 30% of the patients with acute RF will develop valvular sequelae, we end up with 21 900 new cases of RHD yearly. The increased risk of progressing to severe chronic valvular heart disease is associated with moderate or severe carditis, recurrences of acute RF and the mother’s low educational level.35 In developed countries it has been possible to halt progression of RHD to more severe valvular heart lesions, if subclinical carditis is diagnosed early in the course of the disease and treated with secondary prophylaxis.36 Unfortunately, this is not the case in Brazil, because patients remain asymptomatic for 10–20 years, and only seek medical attention after presenting symptoms of HF secondary to RHD. Genetic markers might contribute to particularities of the clinical manifestations of RF in Brazil, where acute rheumatic carditis has become very rare, and over 90% of the patients have asymptomatic acute carditis. Even children and adolescents may have significant haemodynamic rheumatic sequelae without symptomatic carditis, leading to cardiac surgery, an early exposure to valvular prosthesis-related complications and reoperations.37 In this context, a vaccine to prevent RF is highly desirable and may be a reality in the future.38
Reliable data on the incidence of RF are scarce, especially in South America. There have been reports ranging from 7.9 per 1000 habitants in La Paz (Bolivia) to 2.9 per 1000 habitants in Cuba (table 3).36 39 In 2003 101 822 children with RHD, aged between 5 and 14 years, were reported from Latin America.40
Currently, in Brazil and in most of Latin America, there is a trend toward decremental incidences rates of RF and RHD. However, while, most of the healthcare resources are spent in paying for high-complexity cardiac operations, little or no money has been directed towards prevention of the disease. There are presently in Brazil no specific programmes for the prevention or early diagnosis of RF. Moreover, most doctors outside major centres are not sufficiently trained and have no access to diagnostic tests (such as echocardiography). These factors explain why many patients will seek medical attention at major tertiary centres when they already have severe RHD. As a consequence, most such patients would require cardiac surgery.
In a retrospective study involving seven paediatric rheumatology centres in the Sao Paulo State, Brazil, from January 1989 to December 1994, of 786 children and adolescents diagnosed with RF 404 were boys and 382 were girls, with a mean age of 9.4 years.30 Arthritis was present in 453 (57.6%) patients, carditis in 396 (50.4%) and chorea in 274 (34.9%). Erythema marginatum and subcutaneous nodules were seen in only 13 (1.7%) and 12 (1.5%) patients, respectively. Typical migratory polyarthritis occurred in 290/453 (64%) patients. Considering the 50.4% children with a diagnosis of rheumatic carditis, the valvular lesion most often seen was mitral insufficiency (75%), followed by aortic insufficiency (25%), tricuspid insufficiency (9%) and aortic stenosis (0.1%). Silent carditis and recurrences were common findings. In other reports, rheumatic carditis was the most common manifestation of RF, and poor compliance with antibiotic treatment contributed to recurrence and cardiac sequelae.41
According to Brazilian Health Ministry data, over a period of 18 months comprising the whole year of 1995 and the second semester of 1996, 18 500 cases of RF were treated, resulting in 1.8 million visits to hospitals and 4500 cardiac operations. Currently 11 000 cardiac operations are performed annually in Brazil for the treatment of valvular heart disease, and this number is increasing. Moreover, these figures are significantly lower than the demand for that type of surgery, with a consequently long waiting list for valvular surgery; rough estimates are that for every patient who undergoes an operation, five others are on the waiting list.
In our institution, Heart Institute (InCor) of the University of São Paulo Medical School, which is the single largest centre performing valvular surgery in Brazil, 550 valvular operations are performed annually (including valvular repair and replacement) (Heart Institute Annual Surgery Report). Despite these figures there is a waiting list of approximately 2500 patients for valvular heart disease surgical treatment. The mean (SD) age of patients undergoing valvular surgery at the Heart Institute is 48.9 (17) years, 55.3% are female and 41% are reoperations. The aetiology of the valvular heart disease is rheumatic in 65% of the patients, 11% being secondary to mitral valve prolapse, 10% atherosclerotic (mostly aortic stenosis), 9% are due to infective endocarditis and only 5% congenital valvular heart disease. The large number of reoperations can be explained by the young age of the patients at the first operation (due to the predominantly rheumatic aetiology).
Table 4 shows the most common comorbidities in our patients. If valve replacement is needed, 84% have a bioprosthesis implanted. The institutional preference for bioprosthesis is linked to the poor social condition of the patients, most of them living far from medical facilities that are able to perform ambulatory monitoring of oral anticoagulation. The high prevalence of bioprosthesis also contributes to the high incidence of reoperation in the Heart Institute. Overall 30-day surgical mortality is 6.8%, reflecting a high-risk population with significant comorbidities, despite their young age.
Reliable data on the prevalence and incidence of cardiomyopathies in South America are lacking. Data limited to 2006 about causes of death in the São Paulo State (Brazil) showed that cardiomyopathy is an important cause of death (table 1, fig 2). In general, aetiology data are obtained from patients admitted to hospital with HF, patients receiving tertiary care or those included in clinical trials. Chagas’ heart disease and idiopathic cardiomyopathy are common diagnoses especially in some countries. Chronic Chagas’ disease was responsible for 8% of deaths (table 1, fig 2). Alcoholic cardiomyopathy, hypertrophic cardiomyopathy and endomyocardial fibrosis are also reported, but their prevalences are low in comparison with other causes of HF.42 Other restrictive and infiltrative cardiomyopathies are even more rarely found in reported series.43 According to the SEADE data from the São Paulo State (Brazil), restrictive and pericardial aetiologies (endomyocardial fibrosis, Löffler’s endocarditis, endocardial fibroelastosis, amyloidosis, pericardial diseases) were responsible for 0.41% of total HF deaths during 2006 (table 1).
In South America endomyocardial fibrosis is the most common cause of restrictive cardiomyopathies.44 A series of patients with endomyocardial fibrosis has been published in Brazil.45 Angiographic data from patients with endomyocardial fibrosis showed that 59% of them have biventricular involvement.46 Tricuspid regurgitation was seen in 58% of these patients and mitral regurgitation in 60%. Analysis of factors influencing the course of endomyocardial fibrosis showed that biventricular involvement (moderate or severe), right ventricular fibrosis, ascites, atrial fibrillation and the presence of tricuspid and mitral regurgitation were associated with greater mortality.42 47 The presence of atrial fibrillation was associated with a greater prevalence of tricuspid regurgitation and fibrosis of the right ventricle. In Venezuela the most common clinical feature of endomyocardial disease was HF associated with mitral regurgitation.48 The mortality of patients with endomyocardial fibrosis in New York Heart Association functional class I/II and III/IV was 10% and 48% at 5 years’ follow-up.49 Surgical excision of endocardial fibrous tissue has been performed early in the clinical course of endomyocardial fibrosis, thus allowing the preservation of the atrioventricular valves.50 Successful heart transplantation was reported for treatment of end-stage HF due to endomyocardial fibrosis.51
It is estimated that nearly 300 000 people have hypertrophic cardiomyopathy in Brazil.52 However, the cardiovascular outcome was favourable in a Brazilian population with hypertrophic cardiomyopathy, which differs from reports in other countries.53 Early diagnosis and medical therapeutic strategies including amiodarone were considered to favourably influence the prognosis in patients with hypertrophic cardiomyopathy.
Epidemiological data point to an increment in HF incidence in the future in South America as a consequence of factors of epidemiological transition, advances in healthcare, ageing of the population and prevalence of coronary artery disease, cigarette smoking, hypertension, diabetes and obesity.54 Favourable trends for reduction in risk factors leading to reduction in deaths due to cardiovascular disease in developed countries should be a model for South American countries.55 Unfortunately, our data show a increment in hypertensive causes of death (fig 1). Also, epidemiological HF studies are urgently needed to guide the implementation of preventive interventions and appropriate treatments, because of limited available resources. If timely preventive and treatment interventions are not introduced, HF could become one of the main contributors to the burden of morbidity, mortality, and health costs in South America.55 In recent years, public health priorities in developing countries,56 which are set by government politics makers in those countries and by international agencies, have been the reduction of infant and maternal mortality and the control of infectious disease. The results of this study call for serious attention to be given to the consequences of chronic diseases such as HF in developing countries.57
There is a lack of good-quality prevalence surveys of RHD in South America.31 58 Data suggest that RF remains an unsolved public heath problem.59 The persistently high morbidity–mortality due to RHD in Brazil, with high-cost procedures to treat valvular sequelae, warrant procedures to improve primordial prevention (housing, hygiene), primary prevention prophylaxis (sore-throat treatment), secondary prevention prophylaxis, early diagnosis of asymptomatic rheumatic carditis and tertiary prevention (drugs for HF, valve surgery, anticoagulation).36 Also, research should examine the factors associated with RHD such as the organism (group A streptococci), genetics and host and environmental factors (socioeconomic status, living conditions, overcrowding, urbanisation, day care,32 nutrition and access to medical services). The World Health Organization has recommended secondary prophylaxis, most effectively delivered within a coordinated programme using a registry of patients60; however, most developing countries still do not have a secondary prophylaxis programme. The beneficial effects of a programme for prevention and control of RF and RHD were published recently.61
The morbidity and mortality secondary to Chagas’ heart disease make it necessary to develop and maintain programmes to control the vector, eliminate transmission, and carry out clinical trials to improve HF treatment and prevention of sudden death (fig 2, table 1). Additionally, clinical trials to evaluate specific treatment to prevent Chagas’ heart disease or its worsening are highly desirable.62
This review indicates that cardiovascular disease is the predominant cause of death in South America. Additionally, transition from infection pandemics to predominantly degenerative chronic disease is taking place in most South American developing countries. HF is likely to become a major public burden, but the exact magnitude of the problem in South America is unknown and its causes may vary widely among countries. RHD and Chagas’ heart disease are still common causes of HF in many countries of South America. RHD and HF are important public health concerns and tackling these problems deserves high priority.
Competing interests: None.