Hypoalbuminaemia--a marker of cardiovascular disease in patients with chronic kidney disease stages II-IV.

UNLABELLED: Cardiovascular disease (CVD) is a major cause of morbidity and mortality in patients with chronic kidney disease (CKD) patients. Serum albumin, a negative acute-phase reactant and marker for underlying inflammation and/or malnutrition, is an independent predictor of CVD and mortality in CKD VI patients. Such an association in patients with less severe CKD is not well established. We conducted a cross sectional study of all CKD II-IV patients attending the nephrology clinic (N=376; mean age: 57+/-17 years; GFR: 47+/-20 mL/min/1.73 m(2); females 48%; blacks 15%; diabetics 27%; hypertensive 79%). Laboratory and clinical data including risk factors and evidence of CVD were obtained at the point of the most recent visit. The association between risk factors and CVD was evaluated by logistic regression. In the simple logistic regression model, age (p<0.0001), sex (P= 0.02), hypertension (P<0.0001), diabetes (P<.0001), dyslipidemia (p=.01), and serum albumin (p<.0001) were found to be statistically significant. Serum albumin was found to be an independent predictor (p=0.04) of CVD by multiple logistic regression analysis using the above risk factor variables. IN
CONCLUSION: a) hypoalbuminaemia is an independent predictor of CVD in early CKD stages; b) hypoalbuminaemia may be used to identify the population at higher risk for CVD.

INTRODUCTION

400,000 Americans have ESRD and over 300,000 of these patients are on maintenance dialysis 1. CVD is the leading cause of morbidity and mortality in these patients accounting for more than 40% of hospitalizations and almost 50% of deaths 1, 2. This death rate attributed to CVD is 10-20 times that in the general population, stratified for age, race and gender 3.

An estimated 8 million patients have chronic kidney disease of at least stage III (as defined by an estimated glomerular filtration rate [GFR] of less than 60 ml per minute per 1.73 m2 of body surface area) 4. These patients are not on dialysis. However, the prevalence of CVD in these patients has been shown to be significantly higher than the general population 5, 6.

The high burden of CVD in these patients can not be explained just by the high prevalence of traditional risk factors like hypertension (HTN), diabetes (DM), Dyslipidemia (DLP) and advanced age. Of late, novel risk factors like malnutrition and inflammatory state have been implicated in maintenance dialysis patients (CKD stage VI). This association is not well established in patients with less severe CKD. Here we evaluated association between serum albumin, a negative acute-phase reactant and marker of inflammation and/or malnutrition and CVD in patients with CKD stages II to IV.

METHODS

STUDY DESIGN

This is a cross sectional study of all CKD stage II-IV (n= 376) patients attending nephrology clinic of a community hospital. Excluded from initial sample of 583 patients were those with CKD stage I, V and those with missing albumin values or CVD data. Patients with a previous history of dialysis and/or renal transplant were also excluded.

DATA COLLECTION

Data of age, sex, race, DM, HTN, serum chemistries and CVD were collected from office charts and from the most recent visit. All names and identifiers were removed before any analysis of data was performed. Serum chemistries included blood urea nitrogen (BUN), serum creatinine, serum albumin, electrolytes, calcium (Ca), phosphorus (PO4), uric acid, lipids and hemoglobin. CVD included angina pectoris, myocardial infarction (MI), coronary artery disease (CAD), left ventricular hypertrophy (LVH) and heart failure (HF) as per history.

RENAL FUNCTI0N

We used the abbreviated Modification of diet in Renal Disease (MDRD) equation to estimate the GFR 7, 8. Creatinine value on last visit was used to calculate MDRD eGFR. The formula is as below:

OUTCOMES AND COVARIATES

Outcomes were measured in form of prevalence of CVD as defined above. Potential confounders selected based on prior studies and clinical relevance, including age, sex, HTN, DM and DLP were used in final model.

STATISTICAL ANALYSIS

Results were expressed as mean ± SD for continuous variables, and as percentages for categorical data. The association between potential risk factors and CVD was evaluated by logistic regression model. Simple logistic regression models were used to evaluate associations of traditional and novel risk factors for CVD. To evaluate the independent effect of serum albumin on CVD, a multiple logistic regression model was used. All variables known to be associated with CVD involving all traditional risk factors like sex, age, DM, HTN, DLP were put in final multiple logistic regression model to remove confounding effects. Results were reported as p values and Odds Ratios with 95% confidence intervals. All analyses were conducted with the use of STATVIEW software.

RESULTS

The patients included 52% male and 48% female, 85% white and 15% black and mean age was 57 years. The majority of patients (79%) had HTN and 27% had DM. CVD was prevalent in 35% of patients. A total of 62% were on angiotensin converting enzyme inhibitors (ACE-i), 30% on beta blockers and 53% on statins (Table 1). Laboratory data is shown in Table 2. All variables evaluated using simple logistic regression were significant except serum cholesterol (Table 3). The effects of BUN, serum creatinine, calcium, phosphorus, uric acid, and hemoglobin values on CVD were accounted by the GFR status. Multivariate analysis showed that GFR, serum albumin concentration, age, male sex and presence of DM were the independent predictors of CVD in the earlier stages of CKD (Table 4).

Table 1

Characteristics of patients with CKD stage II to IV excluding patient with history of renal transplant or maintenance dialysis.

Variable	Value
Age (years)	57±17
Sex	48% Female, 52% Male
Race	15% Black, 85% White
Diabetes	27%
Hypertension	79%
Cardiovascular disease	35%
ACEi/ ARB	62%
Beta blockers	40%
Statins	53%

Table 2

Laboratory data of the study patients with CKD stage II to IV including serum chemistries.

Parameter	Mean ± SD
MDRD GFR (mL/min/1.73m2):	47 ± 20
BUN (mg/dL):	28 ± 16
S. Creatinine (mg/dL):	1.6 ± 0.7
S. Calcium (mg/dL):	9.4 ± 0.5
S. Phosphorus (mg/dL):	3.6 ± 0.6
S. Albumin (g/dL):	4.1 ± 0.5
S. Uric Acid (mg/dL):	7.0 ± 2.3
S. Cholesterol (mg/dL):	188 ± 43
S. Triglycerides (mg/dL):	169 ± 136
S. hemoglobin (g/dL):	13.2±1.8

Table 3

Univariate Logistic Regression Analysis evaluating association of risk factors with CVD.

Variable	Exp (coef.)	P Value
Age:	1.065	0.0100
Sex (M):	1.689	0.0170
Race (W):	0.489	0.0140
HTN:	5.422	0.0001
DM:	3.849	0.0001
S. Phosphorus	1.609	0.0063
S. Uric Acid:	1.287	0.0001
GFR:	0.952	0.0001
S. Hemoglobin:	0.740	0.0001
S. Calcium	0.539	0.0041
S. Albumin:	0.376	0.0001
S. Cholesterol:	0.997	0.3377

Table 4

Multivariate Logistic Regression Analysis evaluating independent effects of risk factors with CVD.

Variable	P Value	Exp (Coef.)	OR (95% CI)
Age:	0.0001	1.043	(1.023-1.063)
Sex (M):	0.0280	1.788	(1.065-3.002)
HTN:	0.0714	2.127	(0.936-4.833)
DM:	0.0012	2.472	(1.430-4.273)
GFR:	0.0004	0.976	(0.961-0.992)
S. Albumin:	0.0365	0.554	(0.391-0.962)

DISCUSSION

The majority of patients with CKD have severe manifestations of CVD by the time they need maintenance dialysis. This suggests that the damage to the cardiovascular system starts quite early in the time course of progressive chronic kidney disease. Indeed over the last few years, it has been well recognized that CKD patients in the predialysis stage are at increased risk of CVD and its complications. This has led to a rapidly growing interest in the relation between kidney disease and the risk of CVD and is the focus of several recent studies. These studies have shown that the association of CKD to CVD is independent of any traditional risk factors 9-12. The National Kidney Foundation, American Heart Association and the Seventh Joint National Committee on Prevention, Detection, Evaluation & Treatment of High Blood Pressure have classified the presence of CKD as a cardiovascular risk factor 9, 13, 14. These findings led to the evaluation of novel cardiovascular risk factors like chronic inflammation and malnutrition as predictors of the CVD in chronic kidney disease. This multifactorial disease introduces new challenges in predicting and treating patients early in course of CKD to positively alter patient outcome.

There is definitely a high prevalence of traditional risk factors like HTN, DM and DLP in chronic kidney disease patients 15-20, but this alone cannot explain the existing high burden of CVD in this population 21-24. There appears to be a close link between CVD, malnutrion and inflammation in ESRD patients 25, 26. Hypoalbuminemia, a marker of malnutrition and underlying inflammation has come up as a powerful predictor of mortality in patients with ESRD 27-30 and also a significant predictor for the occurrence of de novo vascular events in this population 27. In a recent study C-reactive protein and low albumin has been shown to be the predictor of morbidity and mortality in CKD 3-5 patients in Spain 31. Our study extends this finding to patients with more preserved renal function in American population.

Our hypothesis that serum albumin is a significant risk factor for cardiovascular disease in CKD patients is highlighted by our results on 376 patients with CKD II to IV in whom low serum albumin was significantly associated with CVD irrespective of traditional risk factors like age, sex, HTN, DM and DLP in multivariate analysis.

Beddhu et al 32 showed association between serum albumin level and CVD in chronic hemodialysis patients. An association between serum albumin and cardiovascular mortality has been reported by several studies. Owen et al 33 demonstrated that hypoalbuminemia was a strong predictor of mortality in dialysis patients. Kalantar-Zadeh et al 34 also showed higher mortality in dialysis patients with lower albumin. Many recent studies showed serial measurement of serum albumin can even better predict chronic inflammation and clinical events 35-37. Looking at the results of all these studies it is clear that hypoalbuminemia is adversely associated with CVD in ESRD. Stenvinkel et al. 38 were first to demonstrate that patients in predialysis chronic renal failure with carotid plaque has lower serum albumin level. Nobuhiko et al demonstrated that even in predialytic phase of chronic renal failure, hypoalbuminemia is an excellent reflection of CVD 39. Our study concludes that this is true even in patients with less severe kidney dysfunction. So serum albumin can be a helpful predictor of CVD at early stage of CKD and this patient population needs focused attention because early detection and intervention can provide better outcome.

Available data suggests interrelationship between hypoalbuminemia, inflammation, malnutrition and atherosclerosis in patients with kidney failure 38, 40. In some studies the relation between hypoalbuminemia and CVD is the reflection of inflammation induced malnutrition. The underlying mechanism behind this includes appetite suppression and increased catabolism by inflammatory cytokines 38. Cai and colleagues have implied serum albumin as potential scavenger of free radicals. Decrease in serum albumin level would lead to decrease antioxidant capacity and favor the noxious effects of oxidative stress on a variety of tissues, including the arterial vessel wall 41. These data suggest that hypoalbuminaemia can be more appropriately viewed as a composite marker which reflects malnutrition as well as increased acute phase inflammation, considering that albumin is also a negative acute phase reactant 38, 42-45.

Our study has several limitations. This was a retrospective chart review. Patients were not followed over time, so causal relationships cannot be established. We have not used serum albumin as a time dependent covariate which may have led to an even more reliable prediction of CVD. In addition, many of our patients were on medications that have protective effects on cardiovascular disease. Another important limitation is the lack of consistent measurements of other inflammatory markers such as CRP, whose importance as a predictor of CVD has been shown by others 46. Hypoalbuminemia is a non specific marker of a micro inflammatory state, and is seen in other diseases such as systemic lupus, rheumatoid arthritis, other connective tissue diseases, liver disease, malnutrition from other causes, and does not just represent cardiovascular disease. In addition, data on food intake, basal energy expenditure (BEE), and total daily energy expenditure (TEE) were not available for analysis that may have provided a better understanding of nutritional status in these patients.

Our study concludes that hypoalbuminemia is a strong risk factor for CVD, probably in context of the complex syndrome of malnutrition, inflammation and oxidative stress in patients with CKD. In order to reduce cardiovascular mortality, nutritional, anti-inflammatory and antioxidant intervention will need to be assessed in more randomized control trials. Statins and ACE inhibitors have been shown to have anti-inflammatory effects 47, 48. Recent studies have shown beneficial effects of statins on CVD outcome in CKD patients 49-51. But it is unclear whether the benefit is due to lipid lowering effect, an anti inflammatory effect or both.