You are here

Article Text

Article menu

Download PDFPDF

Cardiomyopathy
Original article
Predictors of outcome in 176 South African patients with peripartum cardiomyopathy
  1. Lori A Blauwet1,
  2. Elena Libhaber2,3,
  3. Olaf Forster4,
  4. Kemi Tibazarwa2,5,
  5. Alex Mebazaa6,
  6. Denise Hilfiker-Kleiner7,
  7. Karen Sliwa2
  1. 1Department of Medicine, Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota, USA
  2. 2Hatter Institute for Cardiovascular Research in Africa, Cape Town, South Africa
  3. 3School of Clinical Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa
  4. 4MDH Health Centre, Ramada, Kenya
  5. 5Groote Schuur Hospital, University of Cape Town, Cape Town, South Africa
  6. 6Department of Anesthesia and Critical Care, Lariboisière Hospital, Paris, France
  7. 7Molecular Cardiology, Department of Cardiology and Angiology, Medical School Hannover, Hannover, Germany
  1. Correspondence to Professor Karen Sliwa, Department of Medicine, Hatter Cardiovascular Research Institute, Medical School, Groote Schuur Hospital and University of Cape Town, Anzio Road, Observatory, Cape Town 7925, South Africa; sliwa-hahnlek{at}mdh-africa.org

Abstract

Objective Identify novel prognostic factors for patients with peripartum cardiomyopathy (PPCM).

Design and setting Prospective cohort study conducted in a single tertiary care centre in South Africa.

Patients 176 African women with newly diagnosed PPCM were studied.

Interventions Clinical assessment, echocardiography and laboratory results were obtained at baseline and at 6 months.

Main outcome measures Poor outcome was defined as the combined end point of death, left ventricular (LV) ejection fraction (LVEF) < 35%, or remaining in New York Heart Association (NYHA) functional class III/IV at 6 months. Complete LV recovery was defined as LVEF ≥55% at 6 months.

Results Forty-five (26%) patients had a poor outcome. Multiple logistic regression analysis revealed that, after adjustment for age, NYHA functional class, LVEF and systolic blood pressure, increased left ventricular end systolic dimension (LVESD), lower body mass index (BMI) and lower total cholesterol at baseline were independent predictors of poor outcome (adjusted OR 1.09, 95% CI 1.04 to 1.15, p=0.001; OR 0.89, 95% CI 0.83 to 0.96, p=0.004, and OR 0.50, 95% CI 0.34 to 0.73, p=0.0004, respectively). Thirty (21%) of the 141 surviving patients with echocardiographic follow-up recovered LV function at 6 months. Multiple logistic regression analysis revealed that, after adjustment for NYHA functional class, LVEF and left ventricular end diastolic dimension, older age and smaller LVESD at baseline were predictors of LV recovery (OR 1.08, 95% CI 1.01 to 1.17, p=0.02 and OR 0.92, 95% CI 0.86 to 0.98, p=0.007, respectively).

Conclusions This study suggests that increased LVESD, lower BMI and lower serum cholesterol at baseline may be independent predictors of poor outcome in patients with PPCM, while older age and smaller LVESD at baseline appear to be independently associated with a higher chance of LV recovery.

  • LIPIDS

https://doi.org/10.1136/heartjnl-2012-302760

Statistics from Altmetric.com

    Request Permissions

    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.

    Introduction

    Peripartum cardiomyopathy (PPCM) is a potentially life-threatening disease that occurs in women of child-bearing age. This disease is characterised by new onset of heart failure between several months before and 6 months after delivery in previously healthy women. Although patients with PPCM have a higher rate of spontaneous recovery of left ventricular (LV) function than patients with other forms of non-ischaemic cardiomyopathy,1 normalisation of LV function at 6 months has been reported to occur in only 23–54% of them.1–5

    Factors predicting poor outcome in case series include degree of LV systolic dysfunction2 ,3 ,6 ,7 and LV dilatation2 ,8 ,9 on presentation, as well as LV thrombus.8 Identifying additional prognostic factors, particularly factors that may be easily and relatively inexpensively assessed, would be beneficial in providing optimal care for patients with PPCM.

    Body mass index (BMI) has been shown to be a predictor of outcome in patients with chronic heart failure,10–13 as well as acute decompensated heart failure,14 but this association has not previously been assessed in patients with PPCM. Renal and liver dysfunction have also both been shown to predict poor outcome in patients with heart failure, but, again, this association has not previously been investigated in patients with PPCM. Inflammation has been implicated in the pathogenesis and prognosis of PPCM,5 ,15 and, while it has been shown that lipoproteins play a role in regulating cytokine production and the associated inflammatory response,16 an association between cholesterol and outcome in patients with PPCM has not previously been reported. We sought to determine whether these variables, among others, may be predictors of outcome in patients with PPCM.

    Methods

    Study design and patient recruitment

    The study was conducted at Chris Hani Baragwanath (CHB) Hospital, Soweto, South Africa. Patients were referred from local clinics, secondary hospitals and the Department of Obstetrics at CHB Hospital. History of pre-existing cardiac signs or symptoms, pre-eclampsia or eclampsia and mode of delivery were obtained from the patient and confirmed by examining the obstetric card carried by each patient. Signs and symptoms were recorded during first presentation at the cardiac unit at CHB Hospital (baseline) and after a follow-up period of 6 months. Clinical assessment, echocardiography and blood analysis were performed at baseline and at 6 months.

    Inclusion criteria were: (1) age ≥16 and ≤40 years; (2) symptoms of congestive heart failure that developed in the last month of pregnancy or during the first 5 months post partum; (3) no other identifiable cause of heart failure; (4) LV ejection fraction (LVEF) ≤45% by transthoracic echocardiography; and (5) sinus rhythm. Exclusion criteria were: (1) significant organic valvular heart disease; (2) systolic blood pressure >160 mm Hg or diastolic blood pressure >100 mm Hg; (3) clinical conditions other than cardiomyopathy that could increase plasma levels of inflammatory markers; (4) severe anaemia (haemoglobin <9 g/dl); and (5) any clinical condition that, according to the investigators, precluded inclusion in the study, such as ischaemic heart disease or malignancy.

    Among the 176 patients included in the study, 164 (93%) received treatment with furosemide and 141 (80%) were treated with an ACE inhibitor. Digoxin was being taken by 113 (64%) of the 176 patients. Patients with an LVEF < 25% or LV thrombus were treated with warfarin. Carvedilol was initiated in 100 (57%) of the 176 patients after resolution of overt heart failure. ACE inhibitor and carvedilol doses were titrated upward as tolerated throughout the 6-month study period. Furosemide dose was titrated upward or downward as indicated according to clinical assessment. This study was approved by the Human Research Ethics Committee of the University of the Witwatersrand, Johannesburg, South Africa and complies with the Declaration of Helsinki. All study participants gave written informed consent before study entry. A total of 176 consenting consecutive patients diagnosed with PPCM and fulfilling the inclusion criteria were enrolled in the study. All women were of African descent.

    Echocardiography, assessment of New York Heart Association (NYHA) functional class, and non-invasive blood pressure measurements

    Two-dimensional and targeted M-mode echocardiography with Doppler colour flow mapping were performed using either a Hewlett Packard Sonos 5500 (Royal Philips Electronics, Amsterdam, Netherlands) or a VIVID i (General Electric Company, Fairfield, Connecticut, USA) echocardiography machine. Systolic and diastolic LV dimensions were measured according to the American Society of Echocardiography (ASE) guidelines. LV dimensions and function were determined using the mean of three or more cycles. Echocardiography was taped on video or CD and stored within the Soweto Cardiovascular Research Unit Division for further reference and audit purposes.

    NYHA functional class of each patient at baseline and follow-up visits was evaluated by a physician, who was provided with clinical data but blinded to the protocol and unaware of the results of the laboratory tests. Blood pressure and heart rate were measured non-invasively with a Critikon Dinamap vital signs monitor 1846 and calculated as mean values from five readings. Measurements were made after a 30 min resting period in the sitting position with 2 min intervals between successive measurements.

    Research-specific blood tests

    Between 10 am and 12 noon, 8 ml blood was withdrawn from an antecubital vein and collected in prechilled tubes containing ethylenediaminetetra-acetic acid or clot activator and mixed rapidly. Plasma or serum was separated by centrifugation at 2500 rpm for 7 min within 10 min of collection. Aliquots were stored at −80°C. Full blood count, liver function, renal function and total cholesterol were assessed.

    Outcomes

    Poor outcome was defined as the combined end point of death, LVEF <35%, or remaining in NYHA functional class III/IV at 6 months. Complete LV recovery was defined as LVEF ≥55% at 6 months.

    Statistical analysis

    Database management and statistical analyses were performed with SAS V.9.2 software. Continuous data are expressed as mean±SD or median (range). Comparison of means and proportions between groups at baseline was performed by independent t test and χ2 statistics or Fisher exact test, respectively. A Wilcoxon rank-sum test was used where data were not normally distributed. Differences in class variables and continuous data between baseline and 6 months were assessed by a McNemar test and a paired t test or sign test (data not normally distributed), respectively.

    Univariate and stepwise multiple logistic regression analyses were performed to establish independent predictors of poor outcome and LV recovery with cholesterol, blood pressure and echocardiographic variables in separate models after adjustment for age and BMI. Univariate logistic regression was used to examine associations with death. Significance was assumed at a two-sided p value of <0.05.

    Results

    Baseline characteristics

    Baseline clinical characteristics are listed in table 1. Notably, the mean age was 30.7±6.9 years, mean parity was 2 (range 1–7), mean BMI was 25.6±5.2 kg/m2, and most of the women (82%) presented with NYHA functional class III or IV symptoms.

    View this table:
    Table 1

    Baseline characteristics of study population (n=176)

    Baseline versus 6-month characteristics among survivors

    Table 2 lists the characteristics of the survivors among the study population at baseline and 6 months. Mean LV end systolic dimension (LVESD) decreased significantly from 51.3±7.6 mm to 42.3±9.5 mm (p<0.0001), while mean LV ejection fraction (LVEF) increased significantly from 27.3±8.1% to 43.3±12.5% (p<0.0001). Mitral inflow E/A decreased significantly from 2.01±0.86 to 1.51±0.65, and mitral inflow deceleration time increased significantly from 140.0±66.1 ms to 185±67.9 ms (p<0.0001 for both). Haemoglobin, renal function and liver function test results had all improved at 6 months. Total cholesterol results were not obtained in most of the study population at 6 months; hence these results were not tabulated.

    View this table:
    Table 2

    Clinical, echocardiography and laboratory variables at baseline and 6 months among survivors (n=141)

    Combined measure of poor outcome

    During the 6-month study period, nine patients were lost to follow-up, three patients moved to remote areas where follow-up could not occur, and two patients did not undergo LVEF assessment at 6 months. Of the remaining 162 patients, 45 (28%) met the prespecified combined end point of death (21 patients, 13%), remaining in NYHA functional class III or IV (13 patients, 9%) or LVEF <35% at 6 months (40 patients, 25%).

    Predictors of poor outcome

    Univariate analysis revealed that predictors of the prespecified poor outcome include decreased systolic blood pressure, increased LVESD, decreased LVEF, increased mitral inflow E/A, decreased mitral inflow deceleration time, increased aspartate transaminase and alanine transaminase, and decreased total cholesterol (table 3).

    View this table:
    Table 3

    Univariate logistic regression analysis of predictors of poor outcome (n=162)

    Predictors of LV recovery

    Thirty (21%) of the 141 surviving patients had fully recovered LV function (LVEF ≥55%) at 6 months. Baseline characteristics of patients who fully recovered versus patients who did not are listed in table 4. Univariate analysis revealed that predictors of LV recovery included older age, decreased LV end diastolic dimension (LVEDD), decreased LVESD, higher haemoglobin and lower creatinine (table 5). Stepwise multiple logistic regression analysis showed that, after adjustment for NYHA functional class, medication, LVEF and LVEDD, older age and smaller LVESD were predictors of LV recovery (adjusted OR 1.08, 95% CI 1.01 to 1.17, p=0.02 and OR 0.92, 95% CI 0.86 to 0.98, p=0.007, respectively).

    View this table:
    Table 4

    Baseline characteristics for patients with complete recovery of LVEF versus patients with non-recovery of LVEF (n=141)

    View this table:
    Table 5

    Univariate logistic regression analysis of predictors of LV recovery (n=141)

    Predictors of death

    Univariate analysis showed that predictors of death included younger age (OR 0.93, 95% CI 0.87 to 1.00, p=0.04), lower BMI (OR 0.83, 95% CI 0.72 to 0.95, p=0.007), increased LVEDD (OR 1.08, 95% CI 1.01 to 1.15, p=0.02), increased LVESD (OR 1.07, 95% CI 1.01 to 1.14, p=0.02), and higher NYHA functional class (OR 2.35, 95% CI 1.12 to 4.94, p=0.02). The mean age of the patients who died (n=21) was 27.8±8.6 years, while the mean age of the patients who survived but were not lost to follow-up (n=141) was 31.0±6.6 years. The mean BMI of the patients who died was 22.6±3.3 kg/m2, while the mean BMI of the patients who survived but were not lost to follow-up was 25.9±5.5 kg/m2.

    Discussion

    This prospective single-centre study of 176 newly diagnosed patients with PPCM summarises data on clinical, echocardiographic and laboratory characteristics at the time of diagnosis and after 6 months of standard clinical care. We confirmed previous findings in our collective5 that a relatively high percentage of patients (79%) had failed to normalise LV function at 6 months compared with collectives from the USA.3 ,8 Increased LVESD, lower BMI and lower total cholesterol were identified as potential novel predictors of poor outcome on stepwise multivariate logistic regression analysis. Older age appears to be a novel predictor of LV recovery, while younger age, lower BMI, increased LVEDD, increased LVESD and higher NYHA functional class all seem to be predictors of mortality.

    Numerous factors predicting morbidity and mortality in patients with PPCM have previously been proposed but not validated. These factors include degree of decreased LVEF and LV dilatation at diagnosis,2 ,7 ,8 ,17 ,18 presence of LV thrombus,8 and being of African descent.8 ,19 Previous studies performed at our own institution have shown that NYHA functional class,5 N-terminal prohormone of brain natriuretic peptide,20 and increased plasma markers of inflammation and apoptosis5 ,20 at diagnosis are predictors of poor outcome as well.

    The present analysis, using a much larger sample size than in most previous analyses, reveals different findings with regard to some predictors of morbidity and mortality, but confirms other findings. As previously reported,2 increased LVESD at diagnosis was a significant predictor even when adjusted for other variables in multivariate analysis. In contrast with previous reports, LVEF and LVEDD at diagnosis,2 ,3 ,6 ,8 ,9 NYHA functional class5 and presence of LV thrombus8 were not predictors of poor outcome. A novel finding in the present study was that lower BMI and lower total cholesterol at baseline were both associated with poor outcome.

    Previous studies have shown that increased BMI is associated with decreased all-cause mortality in patients with chronic heart failure.10 ,12 ,13 The potentially beneficial effect of being overweight or obese has been termed the ‘obesity paradox’. Several hypotheses have been proposed to account for this paradox, including the suggestions that overweight and obese patients may have higher metabolic reserve, reduced cytokine and neuroendocrine activation, higher blood pressure, which may allow more aggressive upward titration of medication, and higher serum lipid levels. An association between higher BMI and improved survival in patients with acute decompensated heart failure has also been shown. An analysis of more than 100 000 patients enrolled in the Acute Decompensated Heart Failure National Registry showed that higher BMI was associated with lower in-hospital mortality.14 The mechanisms by which a higher BMI may be protective for patients with either acute or chronic heart failure remain unclear.

    Several studies have reported an inverse relationship between cholesterol and mortality in patients with chronic heart failure.21–23 Rauchhaus and colleagues were among the first to demonstrate this paradox in a derivative/validation study in 417 patients with chronic heart failure whereby lower serum total cholesterol was independently associated with worse prognosis.24 A more recent study demonstrated that higher serum high-density lipoprotein (HDL) cholesterol and higher serum triglycerides were associated with lower mortality in a cohort of 833 outpatients with chronic heart failure due to various aetiologies.25 In contrast with the results of these studies, Christ and colleagues found that, in patients with idiopathic dilated cardiomyopathy, low cholesterol levels are dependent on the severity of cardiac disease and do not independently predict adverse outcomes in these patients.26

    Low cholesterol has also been shown to be a predictor of adverse outcomes in acute heart failure. Findings from the Acute Heart Failure Database main registry show that low total cholesterol was one of several predictors for in-hospital mortality in patients admitted for acute heart failure.27 In a cohort of 207 older patients, low cholesterol was associated with increased length of hospital stay and was among the best predictors of in-hospital mortality.28

    Although the pathophysiological basis for the association of low cholesterol and impaired prognosis has not been fully elucidated, an ‘endotoxin–lipoprotein’ hypothesis has been proposed.16 Patients with chronic heart failure have increased cytokine activation, which depends, at least in part, on exposure to bacterial endotoxins due to mesenteric venous congestion.24 ,29 Lipoproteins serve as natural buffers because they bind to endotoxins. This, in turn, leads to reduced lipopolysaccharide activity and diminished immune activation.16

    We have previously shown that patients with PPCM display increased cytokine levels and that failure to decrease interferon γ was associated with poor outcome.5 ,20 On the basis of the results of the present study, low total cholesterol at diagnosis may be a marker for increased immune activation in patients with PPCM and portend a poor prognosis.

    Predictors of LV recovery

    In this study, only 21% of surviving patients had attained complete LV recovery at 6 months. Predictors of LV recovery were older age and smaller LVESD at diagnosis, confirming a previous study that reported that smaller LVESD at diagnosis is associated with better LV recovery.2 Other studies have reported that smaller LVEDD,8 ,9 higher LVEF,2 ,6 ,19 higher fractional shortening9 and absence of LV thrombus8 at diagnosis may predict LV recovery, but none of these factors predicted LV recovery in our patient population. Of note, the LV recovery rate at 6 months in this study was quite low compared with a previous study of patients with PPCM in the USA, of whom only 19% were black, which reported an overall LV recovery rate of 54%.3 More recently, another study from the USA in which 14 of 39 patients with PPCM were black reported that normalisation of LV function at 6 months occurred in 56% of white patients but only 30% of black patients.1 The LV recovery rate in our study of black African patients with PPCM is even lower than that reported for black patients living in the USA, but similar to that reported in patients living in Haiti4 and Turkey.2 These findings suggest that patients of African ancestry with PPCM may be less likely to normalise their LV function at 6 months than patients of other racial or ethnic backgrounds.

    Differences in nutrition between patients from various ethnic and sociodemographic backgrounds have not been evaluated and may also play a role. Fett and colleagues30 have shown that, among 32 Haitian patients with severe heart failure, only 6% had recovered at 6 months, but 100% had recovered by 48 months. The long-term recovery rate of the patients with PPCM in the present study has not been assessed, but it may be that predictors of late LV recovery may differ from predictors of early LV recovery. This issue merits further investigation.

    Age has not previously been reported as a predictor of LV recovery. The reason(s) why older patients with PPCM in this study had better LV recovery than younger patients remains unclear. It is possible that younger patients mounted a more severe immune response than older patients, resulting in more extensive myocardial damage, thereby decreasing the probability of functional recovery.

    Predictors of mortality

    Of the 162 patients with follow-up, 21 (13%) died within 6 months of diagnosis. Similar to previous reports, degree of LV dilatation2 and higher NYHA functional class5 at diagnosis were predictors of mortality in univariate analysis. Lower BMI and younger age at diagnosis seemed to be independent predictors of mortality, which are novel findings. Because of the relatively low number of deaths in the current cohort, multiple logistic regression analysis of mortality was not performed. Studies including larger numbers of patients with PPCM are warranted to further elucidate and validate predictors of death among this patient population.

    Limitations

    Total cholesterol was measured at baseline for all patients, but was not assessed at 6 months in the majority of patients, thus not allowing comparison of change over time. In addition, only total cholesterol was measured, so data on the levels of various types of cholesterol including HDL, low-density lipoprotein, triglycerides and non-HDL are not available. Assessment of cholesterol types at baseline and at follow-up may provide additional information in future studies. Finally, follow-up concluded at 6 months. Longer follow-up would probably increase our knowledge about predictors of poor outcome as well as LV recovery in patients with PPCM.

    Conclusion

    In this study, larger LVESD and low total cholesterol were found to be associated with a composite poor outcome in a large cohort of African women with PPCM, while smaller LVESD and older age were found to be associated with a better chance of LV recovery. These results confirm and expand upon previous findings by our group and others and merit further investigations on potential pathophysiological mechanisms underlying these findings.

    Acknowledgments

    We would like to thank the clinicians at CHB Hospital for referring the patients to us. The excellent work of Mrs Bridget Phooko is appreciated.

    References

    1. ↵
      1. Cooper LT,
      2. Mather PJ,
      3. Alexis JD,
      4. et al
      . Myocardial recovery in peripartum cardiomyopathy: prospective comparison with recent onset cardiomyopathy in men and nonperipartum women. J Card Fail 2012;18:28–33.
      OpenUrlCrossRefPubMed
    2. ↵
      1. Duran N,
      2. Gunes H,
      3. Duran I,
      4. et al
      . Predictors of prognosis in patients with peripartum cardiomyopathy. Int J Gynaecol Obstet 2008;101:137–40.
      OpenUrlCrossRefPubMed
    3. ↵
      1. Elkayam U,
      2. Akhter MW,
      3. Singh H,
      4. et al
      . Pregnancy-associated cardiomyopathy: clinical characteristics and a comparison between early and late presentation. Circulation 2005;111:2050–5.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Fett JD,
      2. Christie LG,
      3. Carraway RD,
      4. et al
      . Five-year prospective study of the incidence and prognosis of peripartum cardiomyopathy at a single institution. Mayo Clin Proc 2005;80:1602–6.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Sliwa K,
      2. Forster O,
      3. Libhaber E,
      4. et al
      . Peripartum cardiomyopathy: inflammatory markers as predictors of outcome in 100 prospectively studied patients. Eur Heart J 2006;27:441–6.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Goland S,
      2. Bitar F,
      3. Modi K,
      4. et al
      . Evaluation of the clinical relevance of baseline left ventricular ejection fraction as a predictor of recovery or persistence of severe dysfunction in women in the United States with peripartum cardiomyopathy. J Card Fail 2011;17:426–30.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Habli M,
      2. O'Brien T,
      3. Nowack E,
      4. et al
      . Peripartum cardiomyopathy: prognostic factors for long-term maternal outcome. Am J Obstet Gynecol 2008;199:415e1–5.
      OpenUrl
    8. ↵
      1. Amos AM,
      2. Jaber WA,
      3. Russell SD
      . Improved outcomes in peripartum cardiomyopathy with contemporary. Am Heart J 2006;152:509–13.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Witlin AG,
      2. Mabie WC,
      3. Sibai BM
      . Peripartum cardiomyopathy: a longitudinal echocardiographic study. Am J Obstet Gynecol 1997;177:1129–32.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Curtis JP,
      2. Selter JG,
      3. Wang Y,
      4. et al
      . The obesity paradox: body mass index and outcomes in patients with heart failure. Arch Intern Med 2005;165: 55–61.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Kenchaiah S,
      2. Pocock SJ,
      3. Wang D,
      4. et al
      . Body mass index and prognosis in patients with chronic heart failure: insights from the Candesartan in Heart failure: Assessment of Reduction in Mortality and morbidity (CHARM) program. Circulation 2007;116:627–36.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Kistorp C,
      2. Faber J,
      3. Galatius S,
      4. et al
      . Plasma adiponectin, body mass index, and mortality in patients with chronic heart failure. Circulation 2005;112: 1756–62.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Oreopoulos A,
      2. Padwal R,
      3. Kalantar-Zadeh K,
      4. et al
      . Body mass index and mortality in heart failure: a meta-analysis. Am Heart J 2008;156:13–22.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Fonarow GC,
      2. Srikanthan P,
      3. Costanzo MR,
      4. et al
      . An obesity paradox in acute heart failure: analysis of body mass index and inhospital mortality for 108,927 patients in the Acute Decompensated Heart Failure National Registry. Am Heart J 2007;153:74–81.
      OpenUrlPubMedWeb of Science
    15. ↵
      1. Sliwa K,
      2. Skudicky D,
      3. Bergemann A,
      4. et al
      . Peripartum cardiomyopathy: analysis of clinical outcome, left ventricular function, plasma levels of cytokines and Fas/APO-1. J Am Coll Cardiol 2000;35:701–5.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Rauchhaus M,
      2. Koloczek V,
      3. Volk H,
      4. et al
      . Inflammatory cytokines and the possible immunological role for lipoproteins in chronic heart failure. Int J Cardiol 2000;76:125–33.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Chapa JB,
      2. Heiberger HB,
      3. Weinert L,
      4. et al
      . Prognostic value of echocardiography in peripartum cardiomyopathy. Obstet Gynecol 2005;105:1303–8.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Witlin AG,
      2. Mabie WC,
      3. Sibai BM
      . Peripartum cardiomyopathy: an ominous diagnosis. Am J Obstet Gynecol 1997;176:182–8.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Modi KA,
      2. Illum S,
      3. Jariatul K,
      4. et al
      . Poor outcome of indigent patients with peripartum cardiomyopathy in the United States. Am J Obstet Gynecol 2009;201:171e1–5.
      OpenUrl
    20. ↵
      1. Forster O,
      2. Hilfiker-Kleiner D,
      3. Ansari AA,
      4. et al
      . Reversal of IFN-gamma, oxLDL and prolactin serum levels correlate with clinical improvement in patients with peripartum cardiomyopathy. Eur J Heart Fail 2008;10:861–8.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Afsarmanesh N,
      2. Horwich TB,
      3. Fonarow GC
      . Total cholesterol levels and mortality risk in nonischemic systolic heart failure. Am Heart J 2006;152:1077–83.
      OpenUrlCrossRefPubMed
    22. ↵
      1. Horwich TB,
      2. Hamilton MA,
      3. Maclellan WR,
      4. et al
      . Low serum total cholesterol is associated with marked increase in mortality in advanced heart failure. J Card Fail 2002;8:216–24.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. May HT,
      2. Muhlestein JB,
      3. Carlquist JF,
      4. et al
      . Relation of serum total cholesterol, C-reactive protein levels, and statin therapy to survival in heart failure. Am J Cardiol 2006;98:653–8.
      OpenUrlCrossRefPubMed
    24. ↵
      1. Rauchhaus M,
      2. Coats AJ,
      3. Anker SD
      . The endotoxin-lipoprotein hypothesis. Lancet 2000;356:930–3.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Freitas HF,
      2. Barbosa EA,
      3. Rosa FH,
      4. et al
      . Association of HDL cholesterol and triglycerides with mortality in patients with heart failure. Braz J Med Biol Res 2009;42:420–5.
      OpenUrlPubMed
    26. ↵
      1. Christ M,
      2. Klima T,
      3. Grimm W,
      4. et al
      . Prognostic significance of serum cholesterol levels in patients with idiopathic dilated cardiomyopathy. Eur Heart J 2006;27:691–9.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Spinar J,
      2. Parenica J,
      3. Vitovec J,
      4. et al
      . Baseline characteristics and hospital mortality in the Acute Heart Failure Database (AHEAD) Main registry. Crit Care 2011;15:R291.
      OpenUrlCrossRefPubMed
    28. ↵
      1. Arques S,
      2. Roux E,
      3. Stolidi P,
      4. et al
      . Usefulness of serum albumin and serum total cholesterol in the prediction of hospital death in older patients with severe, acute heart failure. Arch Cardiovasc Dis 2011;104:502–8.
      OpenUrlCrossRefPubMed
    29. ↵
      1. Niebauer J,
      2. Volk HD,
      3. Kemp M,
      4. et al
      . Endotoxin and immune activation in chronic heart failure: a prospective cohort study. Lancet 1999;353:1838–42.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Fett JD,
      2. Sannon H,
      3. Thelisma E,
      4. et al
      . Recovery from severe heart failure following peripartum cardiomyopathy. Int J Gynaecol Obstet 2009;104:125–7.
      OpenUrlCrossRefPubMed

    Footnotes

    • Contributors LB, EL, OF, KT and KS participated in the original conception and design of the study. EL, AM, DH-K, LB, KT and KS participated in the interpretation of the data, drafting and critical revision of the paper and approved the final manuscript submitted.

    • Funding Laboratory investigations were funded through the Medical Research Council of South Africa and the University of the Witwatersrand.

    • Competing interests None.

    • Provenance and peer review Not commissioned; internally peer reviewed.

    Linked Articles