Protein malnutrition and hypoalbuminemia as predictors of vascular events and mortality in ESRD

Protein malnutrition and hypoalbuminemia as predictors of vascular events and mortality in ESRD

Dialysis Therapies Protein Malnutrition and Hypoalbuminemia as Predictors of Vascular Events and Mortality in ESRD Bruce A. Cooper, MMed, Erik L. Pen...

49KB Sizes 0 Downloads 9 Views

Dialysis Therapies

Protein Malnutrition and Hypoalbuminemia as Predictors of Vascular Events and Mortality in ESRD Bruce A. Cooper, MMed, Erik L. Penne, Louise H. Bartlett, MND, and Carol A. Pollock, PhD ● Background: It is proposed that chronic inflammation is common to the pathogenesis of malnutrition and vascular disease, both frequently observed in patients with end-stage renal disease. However, previous studies were unable to differentiate between true protein malnutrition and hypoalbuminemia. Methods: This study was undertaken to determine the associations between malnutrition, measured by total-body nitrogen (TBN), and albumin, a marker of both nutritional status and chronic inflammation, with mortality and morbidity. One hundred nine patients starting dialysis therapy underwent nutritional assessment (TBN level and anthropometric measurements), vascular risk assessment (hypertension, hypercholesterolemia, diabetes mellitus, and smoking status), and serum albumin measurement. Subsequent patient mortality and new vascular events were recorded. Results: Survival was associated independently with both TBN (hazard ratio [HR], 1.6; 95% confidence interval [CI], 1.1 to 2.5; P ⴝ 0.02 for every 10% decrease in nitrogen index) and serum albumin levels (HR, 1.1; 95% CI, 1.0 to 1.2; P ⴝ 0.004 for every 0.1-g/dL (1-g/L) decrease in serum albumin level) adjusted for other variables. Only low serum albumin level predicted subsequent vascular morbidity (HR, 2.2; 95% CI, 1.0 to 4.9; P ⴝ 0.049), as did increasing age (HR, 2.0; 95% CI, 1.4 to 3.0; P ⴝ 0.0004 for every 10-year increase in age) adjusted for other important risk factors. Conclusion: This study dissociates the effect of protein malnutrition and hypoalbuminemia on morbidity and mortality in patients starting dialysis therapy. Protein malnutrition and hypoalbuminemia are independently predictive of mortality, whereas hypoalbuminemia is predictive of vascular morbidity. The study supports the hypothesis that hypoalbuminemia is pathogenically associated with vascular disease, but dissociates this effect from protein malnutrition. Am J Kidney Dis 43:61-66. © 2004 by the National Kidney Foundation, Inc. INDEX WORDS: Nutrition; albumin; total-body nitrogen (TBN); vascular disease; end-stage renal disease (ESRD).

I

NCREASINGLY, patients on renal replacement therapy are elderly and have such comorbid conditions as diabetes, vascular disease, and malnutrition. Each of these factors contributes in its own right to the excess morbidity and mortality seen in these patients.1-3 Often, several factors coexist in an individual patient and result in synergistic effects that have been well described.4,5 However, causative effects of these factors on morbidity are not well defined. Atherosclerosis recently has been recognized as an inflammatory process, characterized by elevations in levels of systemic markers of the acutephase response and inflammatory cytokines.6,7 It also has been proposed that inflammation is associated with malnutrition.8-11 Both inflammation and malnutrition have been implicated with vascular disease.12-17 However, it is unclear whether malnutrition is associated with vascular disease because of concurrent inflammation or, conversely, whether malnutrition in the dialysis population predisposes them to vascular events and thus accounts for the increased mortality in malnourished subjects. Although inflammation in patients with renal failure can be determined by using methods

largely applicable to patients with normal renal function, assessment of malnutrition in patients with renal disease is problematic, particularly in those with concurrent inflammatory disease. For example, serum albumin levels decrease in the presence of systemic inflammation18,19; thus, a low serum albumin level in the presence of vascular disease may not reflect nutritional state. Hence, a reference standard of nutritional measurement should be used to evaluate the presence of protein malnutrition in such patients.

From the Departments of Renal Medicine and Nutrition, and the Department of Medicine, Kolling Institute, Royal North Shore Hospital, University of Sydney, St Leonards, NSW, Australia. Received April 10, 2003; accepted in revised form August 28, 2003. Supported in part by the Baxter Extramural Grant Program, National Health and Medical Research Council of Australia, and Royal North Shore Hospital. Address reprint requests to Carol Pollock, PhD, Department of Medicine, Wallace Freeborn Bldg, Royal North Shore Hospital, St Leonards, NSW 2065, Australia. E-mail: [email protected] © 2004 by the National Kidney Foundation, Inc. 0272-6386/04/4301-0007$30.00/0 doi:10.1053/j.ajkd.2003.08.045

American Journal of Kidney Diseases, Vol 43, No 1 (January), 2004: pp 61-66

61

62

COOPER ET AL

The aim of this study is to assess the relationship between protein nutritional state (determined by direct measurement of total-body nitrogen [TBN]20,21) and serum albumin levels with morbidity (subsequent vascular disease) and mortality in patients starting renal replacement therapy. METHODS One hundred nine patients starting dialysis therapy at Royal North Shore Hospital (St Leonards, Australia) were included in the study. Patients were followed up for a median of 32 months (range, 26 to 74 months). All patients in the study received specific individualized dietary advice concerning protein, calorie, fat, and micronutrient intake. Dialysis regimens were tailored to ensure that patients achieved a weekly Kt/V greater than 3.6 for hemodialysis (HD) and greater than 2.0 for peritoneal dialysis (PD). The majority of patients on PD therapy used continuous ambulatory PD, although automated PD was used during the follow-up period if indicated by peritoneal transport studies. At the start of dialysis therapy, patients underwent an assessment of nutritional adequacy, including measurement of TBN by in vivo neutron activation analysis and anthropometric assessment. TBN is expressed as a percentage of expected TBN for the sex- and height-matched healthy population (nitrogen index [NI]).22 An NI less than 80% is defined as significant protein malnutrition. This technique has been validated previously by us and others and is considered a reference standard of protein nutritional assessment in patients with end-stage renal disease (ESRD).21,23,24 We also have shown that protein nutritional status, measured by TBN, does not significantly change during the first 3 months of dialysis therapy.25 Thus, TBN measured within 3 months of starting dialysis therapy is considered to reflect the baseline protein nutritional state. Serum albumin (bromocresol green method) also was measured at the time of TBN assessment. A serum albumin level less than 3.5 g/dL (35 g/L) is considered to represent significant hypoalbuminemia. Data relating to primary renal diagnoses, history of vascular disease, vascular risk factors, and comorbidity that existed on or before the date dialysis therapy was initiated were collected. Vascular conditions present at the beginning of the study or that subsequently developed during the study (termed new vascular events) were categorized into the following groups: ischemic heart disease, cerebrovascular disease, and peripheral vascular disease. Ischemic heart disease is defined as a history of myocardial infarction, angina pectoris, coronary artery bypass graft surgery, or angioplasty (with or without a stent) or angiographically proven severe atherosclerotic coronary disease. Cerebrovascular disease is defined as a history of a cerebrovascular accident, transient ischemic attack, or severe carotid stenosis or intracranial vascular disease proven by Doppler ultrasound studies or angiogram. Peripheral vascular disease is defined as a history of claudication, ischemic foot ulcer, amputation (toe, foot, or leg), peripheral revascularization procedure, radiologically proven

abdominal aortic aneurysm, or hemodynamically significant atherosclerotic peripheral vascular disease. Additional vascular risk factors are defined as follows: the presence of diabetes mellitus (type 1 or type 2), hypercholesterolemia (total serum cholesterol ⬎ 213 mg/dL [5.5 mmol/L] or treatment with lipid-lowering therapy), moderate hypertension (blood pressure persistently ⬎ 160/95 mm Hg and/or treatment with antihypertensive therapy), and a history of cigarette smoking (defined as current smoker or former smoker). During follow-up, all hospital admissions were reviewed by using medical records. Follow-up of all patients by using medical records was considered justified because the policy of our area health service is to manage all in-patients requiring dialysis therapy at the central hospital. Reason for admission (of at least 24 hours), new vascular events, change in dialysis modality, date of transplantation, hospital transfer or death, and cause of death were recorded. New vascular events were included if they were documented in the medical records or resulted in death. Time to first new vascular event or death was calculated. Relationships between TBN level and hypoalbuminemia with mortality and new vascular events were determined.

Statistical Analysis Statistical analyses were performed using Statview 5.0 (SAS Institute Inc, Cary, NC). Descriptive data are presented as mean ⫾ SEM. P less than 0.05 is taken to indicate statistical significance. Student’s unpaired t-tests were used to examine differences in means between groups. Chisquared test was used to examine differences in proportions between groups. Survival was followed up for a maximum of 3 years. A Cox proportional hazards model was used to assess survival adjusted for multiple covariates. Patients were censored if they underwent renal transplantation or were transferred to another center. When studying new vascular event–free survival, patients were censored as outlined and also for nonvascular death. For each model, all variables considered known risk factors were included in the initial model. Other variables that had a P less than 0.25 on univariate analysis also were included. TBN and albumin levels were tested as continuous and categorical variables, with results of the best interactions reported. Analysis was performed as a stepwise backward analysis. The final multivariate model for a given analysis include all a priori variables plus any other variable found to be significant (P ⬍ 0.05). For completeness, point estimates of study variables (albumin and TBN levels) not found to be significant in a given model were reported by adding such a variable to the final model.

RESULTS

Baseline characteristics of patients starting either PD or HD therapy at the time of nutritional assessment are listed in Table 1. Primary causes of renal failure are listed in Table 2. The percentage of patients with ESRD caused by renovascular disease is somewhat higher and that caused by diabetic nephropathy

MALNUTRITION IN ESRD Table 1.

63

Nutritional Characteristics at Baseline

Parameter

PD

HD

No. of patients 52 57 Sex (F/M) 25/27 22/35 Age (y) 63.9 ⫾ 1.8 57.7 ⫾ 2.1 Time on dialysis 1.6 ⫾ 0.2 2.9 ⫾ 0.3 (mo) Height (cm) 165 ⫾ 1.3 166 ⫾ 1.3 Weight (kg) 65.8 ⫾ 2.1 68.7 ⫾ 2.0 TBN (g) 1547 ⫾ 59 1709 ⫾ 69 NI (%) 90 ⫾ 2 95 ⫾ 2 Body fat (%) 28.0 ⫾ 1.0 27.8 ⫾ 0.9 Lean body mass 47.1 ⫾ 1.4 49.4 ⫾ 1.4 (kg) 24.8 ⫾ 0.5 Body mass index 24.1 ⫾ 0.6 (kg/m2) Serum albumin 3.5 ⫾ 0.08 3.97 ⫾ 0.07 (g/dL) Follow-up (mo) 33.3 ⫾ 2.8 41.1 ⫾ 3.3

Table 3. Prevalence of Vascular Disease and Vascular Risk Factors at Baseline

P

— 0.3 0.03 0.0003 0.4 0.3 0.08 0.09 0.9 0.3 0.5 ⬍0.0001 0.07

NOTE. Values expressed as mean ⫾ SEM or number of patients. To convert albumin in g/dL to g/L, multiply by 10.

is somewhat lower than those observed in the overall dialysis population of Australia and New Zealand.26 The prevalence of extrarenal vascular disease and associated vascular risk factors are listed in Table 3. Median patient follow-up period was 32 months. During the study period, 40 patients died, 21 patients underwent transplantation, 10 patients were transferred to other renal units, and 38 patients were still alive at the end of 3 years. Causes of death were cardiac disease (20 patients), withdrawal from dialysis therapy (7 patients), infection (4 patients), cerebrovascular accident (2 patients), malignancy (2 patients), and miscellaneous (5 patients). Patient survival was found to be independently associated with TBN (hazards ratio [HR], 1.6; 95% confidence

Variable

Count

%

Ischemic heart disease Peripheral vascular disease Cerebrovascular disease Hypertension Smokers Diabetes

45 23 13 103 59 28

41 21 12 94 54 26

interval [CI], 1.1 to 2.5; P ⫽ 0.02 for every 10% decrease in NI) and serum albumin level (HR, 1.1; 95% CI, 1.0 to 1.2; P ⫽ 0.004 for every 0.1-g/dL (1-g/L) decrease in serum albumin level) adjusted for other variables (age, smoking, dialysis modality, sex, and diabetes; Table 4). New vascular events that occurred during the follow-up period included acute myocardial infarction (6 patients), unstable angina pectoris (14 patients), coronary artery bypass graft surgery (3 patients), cardiac death (7 patients), peripheral vascular disease (9 patients), cerebrovascular accident or transient ischemic accident (9 patients), and ischemic colitis (1 patient). Sixty patients had no new vascular event recorded during the follow-up period. A low serum albumin level was associated with this vascular morbidity (HR, 2.2; 95% CI, 1.0 to 4.9; P ⫽ 0.049), along with increasing age (HR, 2.0; 95% CI, 1.4 to 3.0; P ⫽ 0.0004 for every 10-year increase in age) adjusted for other important risk factors (age, smoking, dialysis modality, hypertension, sex, and diabetes; Table 5). No association was found between TBN level and the development of new vascular disease. Table 4.

Table 2.

Primary Renal Diagnosis

Diagnosis

No. of Patients

%

Glomerulonephritis Renovascular/hypertension Diabetic nephropathy Miscellaneous Analgesic nephropathy Polycystic kidney disease Reflux nephropathy Unknown Total

31 20 19 15 7 6 6 5 109

28 18 17 14 6 6 6 5 100

Predictors of Mortality

Variable

HR

CI

P

Age per 10 y Smoking status PD Female Diabetes Serum albumin, 0.1-g/dL (1 g/L) decrease NI per 10% decrease

1.5 1.0 1.6 0.4 0.6 1.1

0.9–2.5 0.4–2.8 0.6–4.5 0.2–1.2 0.2–1.6 1.0–1.2

0.11 1.0 0.4 0.12 0.3 0.004

1.6

1.1–2.5

0.02

NOTE. Model was adjusted for all a priori variables considered important.

64

COOPER ET AL Table 5.

Predictors of New Vascular Events

Variable

HR

CI

P

Age per 10 y Smoking status PD Hypertension Female Diabetes Albumin ⬍3.5 g/dL (⬍35 g/L) NI ⬍80%*

2.0 1.0 1.2 0.6 0.9 1.1 2.2 1.0

1.4–3.0 0.5–2.1 0.6–2.4 0.1–4.8 0.4–1.9 0.5–2.2 1.0–4.9 1.0–1.0

0.0004 0.9 0.7 0.6 0.8 0.8 0.049 0.9

NOTE. Model was adjusted for all a priori variables considered important. *Multivariate point estimate for this variable was reported after it was added to the final model containing age, smoking status, dialysis modality, hypertension, sex, diabetes, and serum albumin level.

DISCUSSION

Hypoalbuminemia and low TBN level were found to be associated with increased mortality. Protein malnutrition independently impacted on survival, with a 10% decrease in nitrogen level associated with a 60% increase in death during the next 3 years. Hypoalbuminemia also independently impacted on survival, with a 0.1-g/dL (1-g/L) decrease in albumin level associated with a 10% increase in death during the study period. This analysis was adjusted for other variables considered a priori to be important predictors of survival (age, smoking status, sex, and diabetes). However, none of these variables was found to be an independent predictor of outcome (Table 4). Although age was not found to be a significant predictor of mortality in this model, a trend existed. Despite the association of both hypoalbuminemia and protein malnutrition with increased mortality, only the former correlated with the development of new vascular events. Specifically, a serum albumin level less than 3.5 g/dL (35 g/L) was independently associated with a 2.2 increase in the likelihood of a new vascular event occurring in the next 3 years. Protein malnutrition, determined by a low TBN level, was not significantly associated with the development of subsequent vascular disease. However, as expected, increasing age was associated with new vascular events. These findings suggest that serum albumin level is reflective of factors other than protein nutrition in patients with ESRD because hypoalbuminemia and protein malnutri-

tion were dissociated with respect to their complications. Associations between cardiovascular events and hypoalbuminemia in patients with renal disease have been well documented by Foley et al.27 Their studies showed that hypoalbuminemia is associated with the development of cardiac failure, ischemic heart disease, and cardiac mortality in HD patients and cardiac failure in PD patients. Another large observational study of dialysisdependent patients28 showed that both hypoalbuminemia and clinically determined malnutrition were associated with increased cardiovascular mortality. However, the diagnosis of malnutrition was based on medical record documentation of either undernourishment or cachexia. Hence, a reference standard of nutritional assessment appears not to have been used and probably was based on subjective clinical evaluation. Heimbu¨ rger et al29 showed that even a formalized method of subjective nutritional assessment could not discriminate between nutritional state determined by serum albumin level, and we recently showed that it does not accurately determine nutritional state compared with TBN.30 Hence, subjective nutritional assessment scores probably reflect comorbid illness, as well as nutritional state, in patients with renal disease and may not be the best method of isolated nutritional assessment. Clearly, a decrease in serum albumin level can occur in chronic inflammatory conditions.18,19 In a group of HD patients,31 hypoalbuminemia was found to be caused by a reduced albumin synthesis rate, considered a consequence of inflammation, rather than a result of protein malnutrition. It is well shown that systemic inflammation is associated with a poor outcome.12-17 However, the present study provides no support to the hypothesis that malnutrition arises because of vascular inflammation inherent in atherogenesis, the so-called malnutrition, inflammation, and atherosclerosis syndrome. Conversely, because atherogenesis has been postulated to be an inflammatory process,32-35 it may primarily be responsible for the low albumin level through the associated inflammation. Although it is known that serum albumin level correlates negatively with C-reactive protein (CRP) level in patients with renal disease,34 we did not prospectively measure CRP in our population because our initial

MALNUTRITION IN ESRD

aims were to compare albumin and TBN levels with respect to outcome. In this study, patients starting PD therapy were nutritionally assessed slightly earlier than those starting HD therapy (Table 1). However, because overall mean time on dialysis therapy before nutritional assessment was less than 3 months, we believe these data reflect baseline protein nutritional state.25 Patients starting PD therapy were significantly older and had lower serum albumin levels than those on HD therapy. However, there were no significant differences in anthropometric or TBN measurements between patients using these 2 dialysis modalities. Because the current study is observational in design, it can only determine disease associations, rather than causation. However, no human study could ever be designed to directly study the effect of protein malnutrition and inflammation on vascular disease and mortality in a prospective randomized fashion. Although the overall number of patients studied was relatively small, this would be considered a relatively large study of patients with ESRD using a reference standard measure of protein nutrition. The present study provides important information relating to vascular risk assessment that extends to the understudied population with renal failure. A randomized controlled trial of dialysis start time currently is being performed that will study the effects of dialysis therapy initiation on protein nutrition, inflammation measured by CRP level, cardiovascular disease, and mortality.36 In conclusion, the present study shows that morbidity and mortality are associated with both hypoalbuminemia and protein malnutrition. However, only hypoalbuminemia is associated with the subsequent development of new vascular disease. This suggests that hypoalbuminemia reflects factors other than a poor nutritional state and dissociates malnutrition as a precursor of vascular disease. REFERENCES 1. Khan IH, Catto GR, Edward N, Fleming LW, Henderson IS, MacLeod AM: Influence of coexisting disease on survival on renal-replacement therapy. Lancet 341:415-418, 1993 2. Mailloux LU, Napolitano B, Bellucci AG, Mossey RT, Vernace MA, Wilkes BM: The impact of co-morbid risk factors at the start of dialysis upon the survival of ESRD patients. ASAIO J 42:164-169, 1996

65

3. Collins AJ, Li S, Ma JZ, Herzog C: Cardiovascular disease in end-stage renal disease patients. Am J Kidney Dis 38:S26-S29, 2001 (suppl 1) 4. Hazzard WR: Atherosclerosis and aging: A scenario in flux. Am J Cardiol 63:20H-24H, 1989 5. Kannel WB, McGee DL: Diabetes and cardiovascular risk factors: The Framingham study. Circulation 59:8-13, 1979 6. Ross R: Atherosclerosis—An inflammatory disease. N Engl J Med 340:115-126, 1999 7. Ross R: Atherosclerosis is an inflammatory disease. Am Heart J 138:S419-S420, 1999 (suppl) 8. Klasing KC: Nutritional aspects of leukocytic cytokines. J Nutr 118:1436-1446, 1988 9. Grimble RF: Cytokines: Their relevance to nutrition. Eur J Clin Nutr 43:217-230, 1989 10. Espat NJ, Copeland EM, Moldawer LL: Tumor necrosis factor and cachexia: A current perspective. Surg Oncol 3:255-262, 1994 11. Beisel WR: Herman Award Lecture, 1995: Infectioninduced malnutrition—From cholera to cytokines. Am J Clin Nutr 62:813-819, 1995 12. Bergstrom J, Lindholm B: Malnutrition, cardiac disease, and mortality: An integrated point of view. Am J Kidney Dis 32:834-841, 1998 13. Stenvinkel P, Heimburger O, Paultre F, et al: Strong association between malnutrition, inflammation, and atherosclerosis in chronic renal failure. Kidney Int 55:1899-1911, 1999 14. Kim SB, Yang WS, Park JS: Role of hypoalbuminemia in the genesis of cardiovascular disease in dialysis patients. Perit Dial Int 19:S144-S149, 1999 (suppl 2) 15. Bergstrom J, Lindholm B: Malnutrition, cardiac disease, and mortality. Perit Dial Int 19:S309-S314, 1999 (suppl 2) 16. Stenvinkel P, Heimburger O, Lindholm B, Kaysen GA, Bergstrom J: Are there two types of malnutrition in chronic renal failure? Evidence for relationships between malnutrition, inflammation and atherosclerosis (MIA syndrome). Nephrol Dial Transplant 15:953-960, 2000 17. Stenvinkel P: Malnutrition and chronic inflammation as risk factors for cardiovascular disease in chronic renal failure. Blood Purif 19:143-151, 2001 18. Moshage HJ, Janssen JA, Franssen JH, Hafkenscheid JC, Yap SH: Study of the molecular mechanism of decreased liver synthesis of albumin in inflammation. J Clin Invest 79:1635-1641, 1987 19. Pannen BH, Robotham JL: The acute-phase response. New Horiz 3:183-197, 1995 20. Allen BJ, Blagojevic N, Delaney I, et al: The role of body protein studies in clinical trials. Basic Life Sci 55:155169, 1990 21. Pollock CA, Ibels LS, Allen BJ, et al: Total body nitrogen as a prognostic marker in maintenance dialysis. J Am Soc Nephrol 6:82-88, 1995 22. Harrison JE, McNeill KG, Strauss AL: A nitrogen index—Total body protein normalized for body size—For diagnosis of protein status in health and disease. Nutr Res 4:209-224, 1984 23. Allman MA, Allen BJ, Stewart PM, et al: Body

66

protein of patients undergoing haemodialysis. Eur J Clin Nutr 44:123-131, 1990 24. Kerr PG, Strauss BJ, Atkins RC: Assessment of the nutritional state of dialysis patients. Blood Purif 14:382-387, 1996 25. Pollock CA, Ibels LS, Zhu FY, et al: Protein intake in renal disease. J Am Soc Nephrol 8:777-783, 1997 26. Disney APS: ANZDATA Registry Report 2002, Adelaide, South Australia, Australian and New Zealand Dialysis and Transplant Registry, 2002 27. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE: Hypoalbuminemia, cardiac morbidity, and mortality in end-stage renal disease. J Am Soc Nephrol 7:728-736, 1996 28. Fung F, Sherrard DJ, Gillen DL, et al: Increased risk for cardiovascular mortality among malnourished end-stage renal disease patients. Am J Kidney Dis 40:307-314, 2002 29. Heimbu¨ rger O, Qureshi AR, Blaner WS, Berglund L, Stenvinkel P: Hand-grip muscle strength, lean body mass, and plasma proteins as markers of nutritional status in patients with chronic renal failure close to start of dialysis therapy. Am J Kidney Dis 36:1213-1225, 2000 30. Cooper BA, Bartlett LH, Aslani A, Allen BJ, Ibels LS, Pollock CA: Validity of subjective global assessment as

COOPER ET AL

a nutritional marker in end-stage renal disease. Am J Kidney Dis 40:126-132, 2002 31. Kaysen GA, Rathore V, Shearer GC, Depner TA: Mechanisms of hypoalbuminemia in hemodialysis patients. Kidney Int 48:510-516, 1995 32. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH: Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 336:973-979, 1997 33. Ridker PM, Hennekens CH, Buring JE, Rifai N: C-Reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med 342:836-843, 2000 34. Zimmermann J, Herrlinger S, Pruy A, Metzger T, Wanner C: Inflammation enhances cardiovascular risk and mortality in hemodialysis patients. Kidney Int 55:648-658, 1999 35. Yeun JY, Levine RA, Mantadilok V, Kaysen GA: C-Reactive protein predicts all-cause and cardiovascular mortality in hemodialysis patients. Am J Kidney Dis 35:469476, 2000 36. Cooper BA, Branley P, Bulfone L, et al: The Initiating Dialysis Early and Late (IDEAL) Study: Study rationale and design. Perit Dial Int (in press)