The aims of the present study were to estimate the prevalence of heart failure (HF) with preserved ejection fraction (HF-PEF) in patients with HF and to compare their clinical characteristics with those with reduced ejection fraction in non-Western countries. The left ventricular ejection fraction ≥45% if measured <1 year before the visit was used to qualify the patients as having HF-PEF. Of the 2,536 consecutive outpatients with HF, 1990 (79%) had the EF values recorded. Of these patients, 1291 had HF-PEF, leading to an overall prevalence of 65% (95% confidence interval 63% to 67%). Compared to the patients with HF and a reduced ejection fraction, those with HF-PEF were more likely to be older (65 vs 62 years, p <0.001), female (50% vs 28%, p <0.001), and obese (39% vs 27%, p <0.001). They more frequently had a history of hypertension (78% vs 53%, p <0.001) and atrial fibrillation (29% vs 24%, p = 0.03) and less frequently had a history of myocardial infarction (21% vs 44%, p <0.001). Only 29% of patients with HF-PEF and hypertension had optimal blood pressure control. Left ventricular hypertrophy was less frequent in those with HF-PEF (58% vs 69%, p <0.001). The prevalence of HF-PEF was lower in the Middle East (41%), where coronary artery disease was more often found than in Latin America (69%) and North Africa (75%), where the rate of hypertension was greater. In conclusion, in the present diverse non-Western study, HF-PEF represented almost 2/3 of all HF cases in outpatients. HF-PEF mostly affects older patients, women, and the obese. Hypertension was the most frequently associated risk factor, highlighting the need for optimal blood pressure control.
Chronic heart failure (HF) is a major health problem associated with high morbidity and mortality. It is well recognized that a substantial proportion of patients with symptomatic HF have a relatively normal or preserved left ventricular (LV) ejection fraction (EF). HF with preserved EF (HF-PEF) is currently defined by the symptoms and signs of HF and an EF of >40% to 50%. The method used to assess the LVEF and the threshold used to divide preserved and reduced systolic function is likely to influence the prevalence of HF-PEF. Epidemiologic studies mainly conducted in Western countries have reported a prevalence of HF-PEF varying from 13% to 74%. Data on the prognosis of patients with HF-PEF compared to that of patients with HF and reduced EF (HF-REF) are contradictory, reporting either a better or similar survival in patients with HF-PEF. Most of these studies have been conducted in Western countries, although limited information is available in other regions of the world. Developing countries are undergoing the first phase of an epidemic of coronary artery disease that will probably lead to an increased incidence of HF. The epidemiology of HF in these other parts of the world could be quite different from that in Western countries with respect to the ethnic background and etiology. Therefore, the purpose of the Identification of patients with heart failure and PREserved systolic Function: an Epidemiological Regional (I PREFER) study was to determine the prevalence and characteristics of patients with HF-PEF in Latin American, Middle Eastern, and North African regions.
Methods
The I PREFER study was a multiregional, cross-sectional, observational study conducted in Latin America, the Middle East, and North Africa and designed to evaluate the prevalence of HF-PEF.
Consecutive eligible outpatients aged ≥21 years, with either a new diagnosis of HF or a previously documented diagnosis of chronic HF, were enrolled during a 3-month recruitment period. The diagnosis of HF had to be established according to the clinical Framingham criteria. Patients with acute decompensation of HF and those <21 years old or already enrolled in a clinical trial were excluded.
The study was planned according to the Declaration of Helsinki and approved by the local ethics committee in each country, and all patients provided signed informed consent.
Clinical data, including the medical history, cardiovascular risk factors, and associate co-morbidities, were collected using a standardized international case report form that was completed at the study visit. Complementary data collection included chest radiography, electrocardiography, echocardiography, Doppler, and laboratory tests, if performed, during the 12 months preceding the visit. Prescription of the main cardiovascular therapeutic classes was recorded.
Patients with an EF of ≥45% were classified as having HF-PEF, and those with an EF of <45% were classified as having HF-REF. Obesity was defined by a body mass index of ≥30 kg/m 2 . Current smokers were patients who smoked or had smoked within the previous 12 months, ≥1 cigarette/day; former smokers were patients who had stopped smoking >12 months before entering the study. Electrocardiographic LV hypertrophy was defined as a Sokolow index of >35 mm or Cornell index of >28 mm in men and >20 mm in women. The LV mass was estimated using the Penn formula and normalized by the body surface area. Echocardiographic LV hypertrophy was defined as an LV mass index of >134 g/m 2 for men and 110 g/m 2 for women. A relative wall thickness of >0.44 was used as a cutoff value for concentric and eccentric LV hypertrophy.
Physicians were randomly selected at the country level among an initial representative list of office- or hospital-based cardiologists and internists. The random process provided a list of potentially participating physicians of 200% of the final number required to take into account physicians who declined to participate. When a physician declined to participate, he was replaced by the next listed physician on the randomized list. Each physician recruited ≥5 consecutive patients. In each country, after random selection, 10% of all sites underwent a data audit by completing a site visit, with 100% of case report forms monitored for source documentation and accuracy.
The sample size determination assumed that HF-PEF prevalence could range from 10% to ≤50%. A sample of 2,100 patients was required to estimate the 95% confidence interval (CI) of the percentage with 1.5% to 2.5% precision. It was decided to enroll about 2,335 subjects to account for 10% of nonevaluable patients. Continuous variables are expressed as the mean ± SD, ratios as the mean and 95% CI, and categorical data as the absolute numbers and percentages. Differences between groups were evaluated using the Student t test for continuous data (or the Wilcoxon test, if appropriate), the chi-square test for binomial or nominal variables (or Fisher’s exact test, if appropriate), and the Cochran-Mantel-Haenszel test using rank scores for ordinal data. Statistical significance was considered as a 2-tailed probability of <0.05. Statistical analysis was performed using SAS software, version 9.1 (SAS Institute, Cary, North Carolina).
Results
A total of 2,539 patients were selected. Of these patients, 3 were excluded (because of age <21 years or acute HF). Therefore, 227 investigators (28% office based, 72% hospital based; 98% cardiologists), 92 in Latin America (33% office based, 67% hospital based; 95% cardiologists), 63 in the Middle East (8% office based, 92% hospital based; 99% cardiologists), and 72 in North Africa (40% office based, 60% hospital based; 100% cardiologists), included 2,536 patients in 10 countries: Latin America, Chile (n = 199), Colombia (n = 211), and Mexico (n = 458); Middle East, Iran (n = 105), Lebanon (n = 181), Saudi Arabia (n = 198), and United Arab Emirates (n = 93); and North Africa, Algeria (n = 400), Egypt (n = 434), and Tunisia (n = 257). The population sample included 42% women, and the mean age was 64 ± 13 years (range 22 to 99). Of these patients, 546 did not have an EF measurement (229 in Latin America, 84 in the Middle East, and 233 in North Africa). Thus, the remaining 1,990 patients (79%) constituted the study population. Of these, 33% had New York Heart Association functional class III or IV.
Of the 1,990 patients with EF values, 1,291 had an EF of ≥45% and were classified as having HF-PEF and 699 had an EF <45% and were classified as having HF-REF. The prevalence of HF-PEF is listed in Table 1 . The average prevalence was 65% (95% CI 63 to 67), ranging from 41% (95% CI 37% to 46%) in the Middle East to 69% (95% CI 65% to 72%) in Latin America and 76% (95% CI 73% to 78%) in North Africa.
| Variable | Patients (n) | Patients With Available EF Data (n) | Patients With EF ≥45% ⁎ |
|---|---|---|---|
| Overall | 2,536 (100%) | 1,990 (79%) | 1,291 (65%); 63–67% |
| Latin America | 868 (34%) | 639 (74%) | 440 (69%); 65–72% |
| Middle East | 577 (23%) | 493 (85%) | 203 (41%); 37–46% |
| North Africa | 1,091 (43%) | 858 (79%) | 648 (76%); 73–78% |
⁎ Ratio between number of patients with EF ≥45% and number of patients with available EF data.
The baseline characteristics of the patients with HF-PEF and HF-REF are listed in Table 2 . The HF-PEF population was significantly older and included a larger proportion of women. The systolic blood pressure and diastolic blood pressure at the visit was significantly greater in the patients with HF-PEF than in those with HF-REF (137/81 mm Hg vs 123/75 mm Hg; p <0.001). Moreover, 51% of the patients with HF-PEF had a systolic blood pressure/diastolic blood pressure value of ≥140/90 mm Hg compared to 25% in the patients with HF-REF (p <0.001). The patients with HF-PEF were also more likely to be obese, with a greater body mass index (29 ± 6 vs 27 ± 5 kg/m 2 ), to have a history of hypertension or valvular heart disease, and to present with atrial fibrillation. The blood pressure control (systolic blood pressure/diastolic blood pressure <140/90 mm Hg in nondiabetics and <130/80 mm Hg in diabetics) in hypertensive patients was 29% in the HF-PEF group versus 50% in the HF-REF group (p <0.001). In contrast, a history of coronary artery disease, a previous myocardial infarction or previous coronary revascularization was more common in patients with HF-REF.
| Variable | HF-PEF Group (EF ≥45%; n = 1,291) | HF-REF Group (EF <45%; n = 699) | p Value | HF Patients Without EF Data (n = 546) |
|---|---|---|---|---|
| Age (years) | 65 ± 12 | 62 ± 13 | <0.001 | 65 ± 13 |
| Women | 648 (50%) | 193 (28%) | <0.001 | 227 (42%) |
| Smoking (current and former) | 388 (30%) | 322 (46%) | <0.001 | 230 (42%) |
| Obesity | 499 (39%) | 184 (27%) | <0.001 | 237 (44%) |
| Diabetes mellitus | 498 (39%) | 267 (38%) | 0.87 | 204 (37%) |
| Hypercholesterolemia | 622 (48%) | 308 (44%) | 0.08 | 272 (50%) |
| Hypertension | 1,008 (78%) | 370 (53%) | <0.001 | 410 (75%) |
| Coronary artery disease | 587 (46%) | 428 (62%) | <0.001 | 238 (44%) |
| Myocardial infarction | 272 (21%) | 310 (44%) | <0.001 | 144 (26%) |
| Valvular heart disease | 412 (32%) | 181 (26%) | 0.005 | 112 (21%) |
| Atrial fibrillation | 372 (29%) | 170 (24%) | 0.03 | 132 (24%) |
| Stroke or transient ischemic attack | 81 (6%) | 53 (8%) | 0.27 | 55 (10%) |
| Peripheral artery disease | 101 (8%) | 52 (8%) | 0.77 | 39 (7%) |
| Chronic obstructive pulmonary disease | 183 (14%) | 88 (13%) | 0.33 | 93 (17%) |
| Heart failure duration ⁎ | <0.001 | |||
| <1 year | 358 (37%) | 177 (30%) | 100 (25%) | |
| 1–5 years | 445 (47%) | 305 (51%) | 218 (55%) | |
| >5 years | 152 (16%) | 116 (19%) | 81 (20%) | |
| New York Heart Association class III/IV | 383 (30%) | 275 (39%) | <0.001 | 167 (31%) |
| Echocardiographic data | ||||
| Left ventricular ejection fraction (%) | 57 ± 10 | 32 ± 8 | <0.001 | |
| Left ventricular end-diastolic diameter (mm) | 50 ± 10 | 62 ± 9 | <0.001 | |
| Left ventricular end-systolic diameter (mm) | 38 ± 11 | 52 ± 9 | <0.001 | |
| Posterior wall thickness (mm) | 12 ± 3 | 10 ± 2 | <0.001 | |
| Interventricular septum thickness (mm) | 12 ± 3 | 10 ± 3 | <0.001 | |
| Relative wall thickness >0.44 † | 557 (53%) | 62 (11%) | <0.001 | |
| Left ventricular hypertrophy ‡ | 588 (58%) | 357 (69%) | <0.001 | |
| Left atrium diameter (mm) | 43 ± 8 | 45 ± 8 | <0.001 | |
| Doppler data (E/A ratio) § | 1.0 ± 0.6 | 1.5 ± 0.9 | <0.001 |
⁎ Available for patients with a previously documented diagnosis of HF.
† Available for 1,601 patients (80%).
‡ Available for 1,526 patients (43%).
Coronary artery disease (44% vs 61%) and dilated cardiomyopathy (11% vs 30%) were reported as etiologies of HF in a greater percentage of patients with HF-REF, and hypertensive heart disease (74% vs 46%) and hypertrophic cardiomyopathy (4.7% vs 1.7%) were more frequent in those with HF-PEF. The patients with HF-PEF were less likely to have New York Heart Association class III or IV and had less frequently been hospitalized during the previous year (32% vs 55% patients and 0.8 ± 1.7 vs 1.3 ± 2.2 hospitalizations; p <0.001).
Chest radiographic data were available for 1,181 patients (59%). Cardiomegaly was more frequently reported in the HF-REF group (77% vs 88%; p <0.001), although the rate of pulmonary edema was the same (41%). Electrocardiography was performed during the previous year in 1,655 patients (83%). The heart rate was significantly greater in the patients with HF-REF (82 ± 19 vs 85 ± 18 beats/min; p = 0.002). Electrocardiographic LV hypertrophy, assessed using the Sokolow index, was present in 17% of those with HF-PEF versus 18% of those with HF-REF (p = 0.7) and in 22% versus 30% (p = 0.002) using the Cornell index, respectively.
As expected, the mean EF was in the normal range in those with HF-PEF (57%) and was 32% in those with HF-REF. Echocardiographic LV hypertrophy was less frequent in the patients with HF-PEF.
Natremia, creatininemia, and hemoglobin was reported in 59%, 69%, and 64% of the patients, respectively. No significant difference was found between the 2 groups. The median brain naturiuretic peptide (reported in 3.8% of patients) was 265 pg/ml in those with HF-PEF and 487 pg/mL in those with HF-REF (p = 0.04).
The characteristics of the patients with HF without EF values were close to that of those with HF-PEF ( Table 2 ).
The characteristics of the patients with HF-PEF varied according to geographic regions ( Table 3 ). The Middle East patients differed from those in Latin America and North Africa because they were more frequently men and smokers more often had diabetes, hyperlipidemia, a history of coronary artery disease or peripheral artery disease, and previous stroke or transient ischemic attack. The rate of HF-PEF was lower, and echocardiographic LV hypertrophy was more frequent. Blood pressure control in the patients with hypertension and HF-PEF varied from 23% in North Africa to 27% in Middle East and 39% in Latin America.
| Variable | Latin America (n = 440) | Middle East (n = 203) | North Africa (n = 648) |
|---|---|---|---|
| Age (years) | 68 ± 13 | 63 ± 12 | 63 ± 12 |
| Women | 60% | 40% | 47% |
| Smoking (current and former) | 24% | 42% | 31% |
| Obesity | 31% | 44% | 43% |
| Diabetes mellitus | 31% | 45% | 42% |
| Hypercholesterolemia | 46% | 66% | 44% |
| Hypertension | 80% | 79% | 76% |
| Coronary artery disease | 31% | 63% | 50% |
| Myocardial infarction | 16% | 34% | 20% |
| Valvular heart disease | 33% | 32% | 31% |
| Atrial fibrillation | 22% | 28% | 34% |
| Stroke or transient ischemic attack | 4% | 10% | 5% |
| Peripheral artery disease | 4% | 12% | 9% |
| Chronic obstructive pulmonary disease | 19% | 16% | 10% |
| Heart failure duration | |||
| <1 year | 37% | 43% | 36% |
| 1–5 years | 45% | 36% | 51% |
| >5 years | 17% | 21% | 13% |
| New York Heart Association class III/IV | 15% | 32% | 39% |
| Echocardiography data | |||
| Left ventricular ejection fraction (%) | 58 ± 10 | 52 ± 7 | 58 ± 10 |
| Relative wall thickness >0.44 | 66% | 61% | 40% |
| Left ventricular hypertrophy | 61% | 66% | 54% |
| Left atrium diameter (mm) | 42 ± 8 | 43 ± 9 | 44 ± 8 |
| Doppler data (E/A ratio) | 0.9 ± 0.5 | 1.1 ± 0.6 | 1.0 ± 0.6 |
The rates of medication used according to LVEF are listed in Table 4 . The HF-PEF group had a greater prescription rate of calcium channel blockers and angiotensin 2 receptor blockers but angiotensin-converting enzyme inhibitors, β blockers, aldosterone receptor antagonists, loop diuretics, and digoxin were prescribed significantly more often in patients with HF-REF.
| Drugs | HF-PEF Group (EF ≥45%; n = 1,291) ⁎ | HF-REF Group (EF <45%; n = 699) † | p Value |
|---|---|---|---|
| Angiotensin-converting enzyme inhibitor | 552 (43%) | 444 (64%) | <0.001 |
| Angiotensin II antagonist | 590 (46%) | 217 (31%) | <0.001 |
| Loop diuretic | 612 (48%) | 512 (74%) | <0.001 |
| Thiazide diuretic | 409 (32%) | 84 (12%) | <0.001 |
| Aldosterone receptor antagonist | 363 (28%) | 359 (52%) | <0.001 |
| β Blocker | 753 (59%) | 492 (71%) | <0.001 |
| Calcium antagonist | 339 (26%) | 57 (8%) | <0.001 |
| Aspirin | 849 (66%) | 454 (65%) | NS |
| Anticoagulant | 273 (21%) | 189 (27%) | 0.03 |
| Other antiplatelet agent | 225 (18%) | 103 (15%) | NS |
| Statin | 648 (50%) | 393 (57%) | 0.009 |
| Nitrate | 331 (26%) | 191 (27%) | NS |
| Digoxin | 262 (20%) | 268 (39%) | <0.001 |
| Amiodarone | 146 (11%) | 112 (16%) | 0.003 |
| Other antiarrhythmic agent | 14 (1.1%) | 5 (0.7%) | NS |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
