Predominance of Heart Failure in the Heart of Soweto Study Cohort
Emerging Challenges for Urban African Communities
Background— There is a paucity of data to describe the clinical characteristics of heart failure (HF) in urban African communities in epidemiological transition.
Methods and Results— Chris Hani Baragwanath Hospital services the 1.1 million black African community of Soweto, South Africa. Of 1960 cases of HF and related cardiomyopathies in 2006, we prospectively collected detailed demographic and clinical data from all 844 de novo presentations (43%). Mean age was 55±16 years, and women (479 [57%]) and black Africans (739 [88%]) predominated. Most (761 [90%]) had ≥1 cardiovascular risk. Mean left ventricular ejection fraction was 45±18%. Overall, 180 patients (23%) had isolated diastolic dysfunction, 234 (28%) tricuspid regurgitation, 121 (14%) isolated right HF, and 100 (12%) mitral regurgitation. The most common diagnoses were hypertensive HF (281 [33%]), idiopathic dilated cardiomyopathy (237 [28%]), and, surprisingly, right HF (225 [27%]). Black Africans had less ischemic cardiomyopathy (adjusted odds ratio, 0.12; 95% CI, 0.07 to 0.20) but more idiopathic and other causes of cardiomyopathy (adjusted odds ratio, 4.80; 95% CI, 2.57 to 8.93). Concurrent renal dysfunction, anemia, and atrial fibrillation were found in 172 (25%), 72 (10%), and 53 (6.3%) cases, respectively.
Conclusions— These contemporary data highlight the multiple challenges of preventing and managing an increasing and complex burden of HF in urban Africa. In addition to tackling antecedent hypertension, a predominance of young women and a large component of right HF predicate the development of tailored therapeutic strategies.
Received April 14, 2008; accepted September 2, 2008.
The clinical syndrome of heart failure (HF) is a final common pathway of most forms of cardiovascular disease.1 Affected patients typically experience poor quality of life, recurrent hospitalizations, and premature mortality.2 A sustained epidemic of HF fueled by high levels of atherosclerosis and hypertension within the aging populations of high-income countries3,4 has necessitated the development of cost-effective preventative1 and disease management programs.5 There is, however, a paucity of data describing the etiology and underlying cardiac structure and function associated with HF in low- to middle-income countries. In Africa, few clinical data are derived from echocardiography,6,7 and the single largest study of 1000 subjects emanates from the early 1960s.8 In sub-Saharan Africa, the conventional view is that HF is caused by untreated rheumatic heart valvular disease, peripartum and idiopathic cardiomyopathy, and hypertension.9 However, in the Heart of Soweto Study of one of Africa’s largest urban concentrations of black Africans, we demonstrated the potential emergence of more affluent forms of cardiovascular disease due to epidemiological transition.10 It is within this context that we describe, to our knowledge, the largest and most comprehensive study of HF in Africa to date.
Editorial p 2323
Clinical Perspective p 2367
As described previously,10,11 the Heart of Soweto Study is investigating emergent heart disease in the townships of Soweto (1.1 million people), South Africa. Like other parts of sub-Saharan Africa, greater wealth has created the potential for an epidemiological transition toward more affluent disease in Soweto, and modifiable cardiovascular risks are now highly prevalent.12 Alternatively, Soweto is being steadily populated by rural migrants in whom traditional lifestyles remain the norm. South Africa has a population of different ethnic backgrounds encompassing a predominance of black Africans, Europeans, Asians, and those with mixed ancestry (coloreds), and this is reflected in our study cohort.
Significantly, among the 4162 cases captured by the prospective Heart of Soweto Study clinical registry of established and de novo presentations to the Cardiology Unit of the 3500-bed Chris Hani Baragwanath Hospital in 2006 (all were subjected to echocardiography), 1960 cases (47%) were diagnosed with HF.10 This article provides a detailed description of the clinical and demographic characteristics of the 844 de novo cases of HF (18% of 2006 cases).10 Our objectives were to better understand any fundamental changes in the pathways to HF in the region and to develop effective public health programs that optimize early detection and treatment.13
Because the hospital is the only provider of specialist cardiac services for Soweto, its case load represents a barometer of the underlying spectrum of cardiovascular disease (mild to severe).10 In 2006, the hospital managed 129 633 inpatients (35% via the Department of Medicine).
Currently, the Cardiology Unit manages 21 000 patients per annum and is staffed by internal medicine specialists in cardiology training and experienced cardiologists. Collectively, they apply advanced diagnostic investigations and gold-standard treatment to all patients seen via the cardiology outpatient department and Coronary Care Unit. The Cardiology Unit also manages patients referred from all other inpatient (eg, general medical ward) and outpatient (eg, diabetic clinic) units as well as those referred from 12 local primary care clinics.
In addition to detailed demographic and clinical profiling (eg, Minnesota coding of 12-lead ECGs14), all patients with HF were subjected to a detailed echocardiographic assessment of ventricular function, valvular integrity and function, and regional wall abnormalities (specific measurements available in all but 67 of de novo cases [8%]). Two-dimensional targeted M-mode echocardiography with Doppler color flow mapping was performed with a Hewlett-Packard Sonos 5500 echocardiograph attached to a 2.5- or 3.5-MHz transducer by trained operators according to the American Society of Echocardiography guidelines.15 Diastolic mitral flow was assessed by pulsed-wave Doppler echocardiography from the apical 4-chamber view. The E wave deceleration time was measured as the interval between the peak early diastolic velocity and the point at which the steepest deceleration slope is extrapolated to the zero line.
To exclude ischemic etiology, every patient (regardless of age and including those ultimately diagnosed with idiopathic dilated cardiomyopathy) with clinical suspicion of coronary artery disease based on ECG (eg, pathological Q waves) and echocardiography (eg, regional wall motion abnormalities) is routinely subjected to a combination of stress testing, cardiac nuclear imaging, and cardiac catheterization; this involved ≈350 nuclear perfusion scans and 550 coronary angiograms in 2006. Data availability and specific criteria applied are provided in table legends.
Table 1 summarizes the criteria that were applied, via independent review and consensus agreement by K.S. and S.S., to stratify patients according to the probable etiology of HF. These are based on European Society of Cardiology guidelines1 and the recent European Heart Survey.16
This study was approved by the local ethical committee and relevant administrative bodies and conforms to the principles outlined in the Declaration of Helsinki. Each patient in the study was assigned a unique identifying code (9 digits), all documents were labeled accordingly to maintain anonymity, and each patient verbally agreed to participate at the time of data collection.
All study data were documented and entered into the study database by a dedicated team of experienced cardiac nurses. Data were then verified and analyzed with SAS version 9.1 (SAS Institute Inc, Cary, NC). Normally distributed continuous data are presented as mean±SD and nongaussian distributed variables as the median plus interquartile range. Percentages are presented with 95% CIs where appropriate. Comparisons according to demographic and clinical profile involved χ2 analysis with calculation of odds ratios and 95% CIs for discrete variables and the Student t test and ANOVA for normally distributed continuous variables. Multiple logistic regression analyses (entry model) were performed on age, sex, ethnic origin, and risk factors to derive adjusted odds ratios. Significance was accepted at the 2-sided level of 0.05.
The authors had full access to and take responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Clinical and Demographic Profile
In 2006, 1960 patients (163 cases per month) presented with a primary or secondary diagnosis of HF. Of these, 844 patients (43%) were de novo presentations. Most patients were referred either from another outpatient clinic (384 [46%]) or as a hospital inpatient (350 [42%]); the remainder (96 cases [11%]) primarily were referred from a primary care clinic. Women (479 cases [57%] and black Africans (739 cases [88%]) predominated. Black African women (n=437) were the largest patient subgroup of patients, representing 59% versus 40% of their ethnic grouping (P<0.0001; odds ratio, 1.49; 95% CI, 1.25 to 1.78) and 52% of the entire cohort. Almost two thirds were from Soweto and lived there for a median of 40 (interquartile range, 30 to 52) years. Table 2 compares the broad characteristics of the cohort according to sex and race. Women were on average 3 years younger than men and had significantly increased odds of having a positive family history and to be obese. Men were at increased odds of having a history of smoking. More women reported advanced symptoms of dyspnea and peripheral edema and had higher resting heart rates. Compared with other ethnic groups, black Africans had received less formal education, and fewer reported a positive family history or a personal history of smoking, stroke, or diabetes. On presentation, black Africans had substantially higher resting heart rates but more preserved renal function.
Risk Factor Profile
Overall, 761 patients (90%) had ≥1 cardiovascular risk, and 519 (62%) had multiple risks, the most common being diastolic and/or systolic hypertension (503 cases [60%]), a history of smoking (396 cases [47%], obesity (129 cases [34%] of those with both height and weight data recorded), and dyslipidemia (66 cases [19%] of those subject to fasting cholesterol tests).
Five major forms of HF were diagnosed overall, with mixed left and right forms of HF found in 104 (12%) of cases: (1) 281 cases of hypertensive HF (33%; 95% CI, 30% to 36%); (2) 237 cases of idiopathic cardiomyopathy (28%; 95% CI, 25% to 31%); (3) 225 cases of right HF overall (27%; 95% CI, 24% to 30%) and 121 primary diagnoses of the same; (4) 77 cases of ischemic cardiomyopathy (9%; 95% CI, 7% to 11%); and (5) 67 cases of valvular HF (8%; 95% CI, 6% to 10%). A miscellaneous group of 61 cases of other discrete forms of HF (combined with idiopathic dilated cardiomyopathy in the tables and Figure) included 29 cases of peripartum cardiomyopathy, 29 cases of HIV-related cardiomyopathy (13 with concurrent tuberculosis), and 4 cases of hypertrophic obstructive cardiomyopathy.
The Figure shows clinically important differences in the etiology of HF according to gender and race. With adjustment for age, race, and risk factor profile (family history, smoking status, lipid profile, blood pressure, and diabetic status), the odds of a woman (predominantly black Africans) presenting with an ischemic cardiomyopathy (6% versus 13%; adjusted odds ratio, 0.49; 95% CI, 0.28 to 0.85; P=0.011) were significantly lower compared with men. The odds of a black African presenting with an ischemic cardiomyopathy (6% versus 34%; adjusted odds ratio, 0.12; 95% CI, 0.07 to 0.20) were markedly lower than other ethnic groups. Conversely, black Africans had 5 times higher odds of presenting with an idiopathic dilated cardiomyopathy or other form of HF (38% versus 12%; adjusted odds ratio, 4.80; 95% CI, 2.57 to 8.93) compared with the rest (P<0.0001 for both comparisons).
Table 3 shows the clinical and demographic features of the 5 major groups of HF with some important differences evident. Those with idiopathic cardiomyopathy and other causes of HF were significantly younger than the rest (P<0.001). Only 10% of patients (72 cases) had concurrent anemia, and >10% had a history of stroke and/or atrial fibrillation (58 and 53 cases, respectively). Overall, 141 (20%) had an estimated glomerular filtration rate indicative of moderate to severe renal dysfunction. Black Africans were more likely to have preserved renal function (35% versus 23%; P=0.009) compared with the rest, which is partly explained by the fact that those with an ischemic cardiomyopathy had significantly lower estimated glomerular filtration rate than those with a dilated cardiomyopathy or right HF (P<0.01 for both comparisons).
Table 4 shows the echocardiographic features according to primary underlying etiology. Overall, 417 (54%) of those with a quantified left ventricular ejection fraction had moderate to severe systolic dysfunction, and 225 (29%) had impaired diastolic function (180 [80%] with isolated diastolic dysfunction). In addition to those with primary valvular disease, patients with an idiopathic cardiomyopathy and right HF were most likely to have moderate to severe mitral or tricuspid regurgitation.
There were many complex cases. For example, 237 patients were diagnosed with idiopathic dilated cardiomyopathy after exclusion of coronary artery disease and ongoing investigations for familial cardiomyopathy. Although prior hypertension was reported in 101 of these cases (43%), most were normotensive and had no evidence of left ventricular hypertrophy. In addition, 29 black African women presented with peripartum cardiomyopathy, and 24 cases were found to be HIV-related.
Perhaps most significantly, there were many cases of right HF, either in its isolated form (ie, 121 patients without left-sided pathology, particularly valve disease, which could have explained their clinical state) or in combination with other forms of HF (104 cases). Beyond those with cor pulmonale (84 of 121 isolated right HF cases [69%]), there were a range of pathologies contributing to these cases. This included 20 cases (9%) of rheumatic heart disease (primary etiology of valvular HF) and 13 cases (12 were women) of pulmonary hypertension secondary to connective tissue disorders (6%) in addition to 13 cases (6%) of pulmonary complications secondary to tuberculosis.
Initial Pharmacological Treatment
At clinical presentation, 517 patients (61%) were prescribed at least 1 form of HF-specific treatment (43% combination therapy) with no differences on the basis of race or gender. In those with systolic dysfunction (n=417), the 5 most commonly prescribed agents were loop or thiazide diuretic (85%), angiotensin-converting enzyme inhibitor (70%), β-blocker (64%), aldosterone inhibitor (60%), and digitalis (19%). In those with preserved systolic function (n=373), these were loop or thiazide diuretics (49%), angiotensin-converting enzyme inhibitor (44%), β-blocker (25%), aldosterone inhibitor (22%), and a calcium antagonist (18%).
This study represents one of the largest and most comprehensive studies of HF in Africa focusing on cases derived from urban-dwelling black Africans in epidemiological transition. The volume and pattern of cases challenge a number of assumptions about the natural history of HF on the continent. Although resembling some aspects of the HF epidemic in high-income countries,3 the spectrum of documented cases represented a broad combination of traditional and affluent pathways to the syndrome in greater numbers than first anticipated. These data have important clinical and public health implications for Soweto and other populations in epidemiological transition.
Unlike the typical pattern of HF in high-income countries,2–4 this urban African cohort was dominated by young black African women, and we found a broad diversity of heart disease attributable to a combination of infectious and noncommunicable diseases.10 The surprisingly high number of cases (while no doubt inflated by increased detection rates due to routine echocardiography) is indicative of a much higher prevalence of HF than previously suspected in urban Africans. Consistent with previous African reports, many cases were attributable to idiopathic dilated cardiomyopathy,17 HIV-related cardiomyopathy,18 peripartum cardiomyopathy,19 and endomyocardial fibrosis.20 Equivalent studies from Nigeria (315 patients)21 and Cameroon (167 patients)22 have also highlighted the importance of communicable diseases and valve disease in the development of HF in Africa. Importantly, these data reinforce the central (and most probably increasing) role of hypertension (found in 60% of cases) in the development of HF,6,7 which represents a high-priority target for HF prevention in urban Africa. Fortunately, hypertension is readily identifiable,13 and 15% of patients had ECG evidence of left ventricular hypertrophy; although the progression from concentric left ventricular hypertrophy to HF has not been well defined in Africa, data from hypertensive African Americans suggest that this is a common outcome.23 There is a clear need, therefore, to develop African-specific prevention programs that optimize the detection and management of hypertension in addition to targeting those most at risk of developing peripartum cardiomyopathy19 and HIV-related cardiomyopathy.18
These data also reaffirm that valvular disorders play a pivotal role in the development of HF in Africa, with 8% of de novo HF cases attributable to primary valve disease (mostly chronic rheumatic heart disease) and ≈25% of most forms of HF having concurrent valve dysfunction. Significantly, not a single case of acute rheumatic fever, a previously common diagnosis in the region,24 was documented, most probably reflecting less overcrowding, better sanitation, and access to antibiotic therapy in children and young adults. However, children are managed elsewhere, and such cases remain high in rural areas of sub-Saharan Africa.6 Regardless, there is clear potential for early screening programs to detect rheumatic disease–related cardiac murmurs in urban communities. Furthermore, the latent potential for increasing numbers of cases with ischemic cardiomyopathy25 is underscored by a near 5-fold difference (2% versus 9%) in reported cases compared with previous reports.6 With many so complex cases, it was essential to distinguish between primary versus functional valve disease (eg, idiopathic dilated cardiomyopathy resulting in annular dilatation). Surprisingly, few HF studies (eg, the EuroHeart Failure Survey II reported that 72% of de novo acute HF was complicated by mitral regurgitation) make such a distinction.16 In Africa, such distinctions are important given that the therapeutic approach to HF due to severe rheumatic mitral regurgitation (surgery) versus functional mitral regurgitation (eg, in selected patients with idiopathic dilated cardiomyopathy who may benefit from a multisite pacemaker26) will markedly differ. This highlights the challenge of adapting and prioritizing therapeutic strategies according to the changing etiology of HF.
The surprisingly high prevalence of isolated and concomitant right HF (21% of cases combined) relative to reports from high-income countries (only 3.2% in the EuroHeart Failure Survey II cohort16) is also worthy of comment. This disparity might reflect poor access to healthcare facilities; more than two thirds of right HF cases presented as “late” pulmonary complications. However, consistent with the BOLD Study showing that South Africa had a particularly high prevalence of chronic obstructive pulmonary disease,27 ≈20% of right HF cases were of indeterminate origin. The high prevalence of right HF in individuals who never smoked requires further investigation, particularly in respect to the role of air pollution and potentially greater genetic susceptibility in certain individuals. Certainly, understanding and addressing the potentially preventable antecedents of right HF represent a major target for reducing the burden of HF in this population. Once again, this would require the development of specific treatments for a form of HF rarely reported and managed in high-income countries.
Our study has a number of limitations. First, this study cohort only reflects those who seek specialist care and therefore reflects those with more advanced forms of HF. These data undoubtedly underrepresent adults living with milder forms of HF, and we have no data on sudden fatal events secondary to undiagnosed HF. Unfortunately, there are no gold-standard methods for definitively categorizing the etiology and clinical characteristics of HF. We applied a clinically orientated approach based on published criteria.16 We acknowledge potential biases in our classification of cases despite our careful delineation of primary versus secondary valve disease/dysfunction. Finally, as a clinical registry we did not systematically validate diagnostic data but (wherever possible) have adhered to the recently published Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines in our reporting of study data.28
In conclusion, this study provides important insights into the broadening range of causes (ie, beyond the traditional causes of HF in Africa) and characteristics of HF arising from a large urban population of black Africans in epidemiological transition. Although the current burden of HF in Soweto does not (yet) resemble that found in high-income countries, we have highlighted the emerging challenge of dealing with an expected rise of HF on the continent (particularly in relation to early prevention and treatment strategies). The complexity of cases has forced us to more critically examine case classifications beyond that typically reported. Ultimately, these data challenge us to recognize and respond to HF in Africa by recognizing unique aspects concerning its natural history (particularly antecedent hypertension and a major component of right HF) and the need for culturally sensitive interventions.
The authors wish to thank all the doctors, nurses, and patients who participated in the registry. We acknowledge in particular Elisabeth Tshele, Bridget Phooko, Maureen Kubheka, and Phutuma Mathusi.
Sources of Funding
The registry was supported by unconditional research grants from Adcock-Ingram, the Medtronic Foundation, and Servier. Dr Stewart is supported by the National Health and Medical Research Council of Australia.
Swedberg K, Cleland J, Dargie H, Drexler H, Follath F, Komajda M, Tavazzi L, Smiseth OA, Gavazzi A, Haverich A, Hoes A, Jaarsma T, Korewicki J, Levy S, Linde C, Lopez-Sendon JL, Nieminen MS, Pierard L, Remme WJESC. Guidelines for the diagnosis and treatment of chronic heart failure: executive summary (update 2005): the Task Force for the diagnosis and treatment of CHF of the European Society of Cardiology. Eur Heart J. 2005; 26: 1115–1140.
McMurray JJ, Stewart S. Epidemiology, etiology and prognosis of heart failure. Heart. 2000; 83: 596–602.
Stewart S, MacIntyre K, Capewell S, McMurray JJV. Heart failure and the ageing population: an increasing burden in the 21st century? Heart. 2003; 89: 49–53.
Wang TJ, Evans JC, Benjamin EJ, Levy D, LeRoy EC, Vasan RS. Natural history of asymptomatic left ventricular systolic dysfunction in the community. Circulation. 2003; 108: 977–982.
Damasceno A, Cotter G, Sliwa K, Mayosi BM. Heart failure in sub-Saharan Africa: time for action. J Am Coll Cardiol. 2007; 17: 16988–16993.
Mayosi BM. Contemporary trends in the epidemiology and management of cardiomyopathy and pericarditis in sub-Saharan Africa. Heart. 2007; 93: 1176–1183.
Cosnett JE. Heart disease in the Zulu: especially cardiomyopathy and cardiac infarction. Br Heart J. 1962; 24: 76–82.
Sliwa K, Damasceno A, Mayosi BM. Epidemiology and etiology of cardiomyopathy in Africa. Circulation. 2005; 112: 3577–3583.
Stewart S, Wilkinson D, Becker A, Askew D, Ntyintyane, McMurray J, Sliwa K. Mapping the emergence of heart disease in a black, urban population in Africa: the Heart of Soweto Study. Int J Cardiol. 2006; 101: 101–108.
Tibazarwa K, Ntyintyane L, Sliwa K, Gerntholtz T, Carrington M, Wilkinson D, Stewart S. A time bomb of cardiovascular risk factors in South Africa: results from the Heart of Soweto Study “Heart Awareness Days.” Int J Cardiol. February 1, 2008. doi:10.1016/j.ijcard.2007.11.067. Accessed September 1, 2008.
Prineas RJ, Crow RS, Blackburn H. The Minnesota Code Manual of Electrocardiographic Findings: Standards and Procedures for Measurement and Classification. Boston Mass: John Wright; 1982.
Sahn DJ, DeMaria A, Kisslo J, Weyman A, for the Committee on M-mode Standardization of the American Society of Echocardiography. Recommendations regarding quantification in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation. 1978; 58: 1072–1083.
Komajda M, Lassus J, Lopez-Sendon JL, Ponikowski P, Tavazzi L, Nieminen MS, Brutsaert D, Dickstein K, Drexler H, Follath F, Harjola VP, Hochadel M, on behalf of the EuroHeart Survey Investigators. EuroHeart Failure Survey II (EHFS II): a survey on hospitalized acute heart failure patients: description of population. Eur Heart J. 2006; 27: 2725–2736.
Mayosi BM, Wiysonge CS, Ntsekhe M, Volmink JA, Gumedze F, Maartens G, Aje A, Thomas BM, Thomas KM, Awotedu AA, Thembela B, Mntla P, Maritz F, Blackett KN, Nkouonlack DC, Burch VC, Rebe K, Parish A, Sliwa K, Vezi BZ, Alam N, Brown BG, Gould T, Visser T, Shey MS, Magula NP, Commerford PJ. Clinical characteristics and initial management of patients with tuberculous pericarditis in the HIV era: the Investigation of the Management of Pericarditis in Africa (IMPI Africa) registry. BMC Infect Dis. 2006; 6: 2.
Amira C, Ntyintyane L, Wilkinson D, Stewart S, Sliwa K. Emerging epidemic of cardiovascular disease among urban Africa: acute coronary syndrome at Baragwanath Hospital, Soweto. S Afr Heart J. 2006; 3: 7–12.
Buist AS, McBurnie MA, Vollmer WM, Gillespie S, Burney P, Mannino DM, Menezes AM, Sullivan SD, Lee TA, Weiss KB, Jensen RL, Marks GB, Gulsvik A, Nizankowska-Mogilnicka E, on behalf of the BOLD Collaborative Research Group. International variation in the prevalence of COPD (the BOLD Study): a population-based prevalence study. Lancet. 2007; 370: 741–750.
In one of the largest studies of heart failure (HF) in Africa, we studied 844 de novo cases in Soweto, South Africa. The results challenge a number of previous assumptions regarding the natural history of HF in Africa. Compared with more affluent countries, there was a higher prevalence of women (57%), and patients were younger (the mean age was 55±16 years.) The unexpectedly high burden of HF was attributed to a combination of etiologies, with the expected cases of hypertensive HF, idiopathic dilated cardiomyopathy, and HF due to structural valve disease supplemented by right HF and ischemic heart failure. These data have important clinical and public health implications for urban African and other communities in epidemiological transition where old and new forms of HF will emerge. Treatment strategies need to keep pace with the changing etiology of HF, and preventative strategies need focus on both old (eg, late recognition of rheumatic heart disease) and new antecedents (eg, atherosclerosis) of HF. Hypertension, in particular, represents a traditional antecedent for HF in Africa that will gain even more clinical importance as a preventative target. Significantly, we found a surprisingly high burden of right HF (either in isolation or in combination with left HF). The increasing burden of HF in vulnerable populations in Africa and beyond will necessitate the development of relatively inexpensive but effective programs of care that provide treatments adapted to the local healthcare system and the specific population.