| |
Metabolic Syndrome, Proteinuria, and the Risk of Progressive CKD in Hypertensive African Americans
|
| |
| |
American Journal of Kidney Diseases
May 2008
Janice Lea, MD, MSc1, Deanna Cheek, MD2, Denise Thornley-Brown, MD3, Lawrence Appel, MD4, Lawrence Agodoa, MD5, Gabriel Contreras, MD6, Jennifer Gassman, PhD7, Jim Lash, MD8, Edgar R. Miller III, MD, PhD4, Otelio Randall, MD9, Xuelei Wang, MS7, William McClellan, MD, MPH1, AASK Study Investigators
"In summary, metabolic syndrome is associated with proteinuria in hypertensive African Americans with CKD, but is not independently associated with CKD progression. Moreover, proteinuria predicts CKD progression in those with and without metabolic syndrome. Because of the well-described association of proteinuria and CKD progression, our finding of increased proteinuria in subjects classified to have metabolic syndrome warrants additional studies of possible putative mechanisms of this association and for establishing temporality between metabolic syndrome and proteinuria. Certainly, future investigations of the effect of metabolic syndrome on proteinuria and CKD progression should include more specific measures of insulin resistance. Our data provide a possible mechanism to explain, at least in part, some of the variability observed within and between races in the progression of CKD. Moreover, metabolic syndrome is a modifiable risk factor and thus strategies to decrease the prevalence and/or severity of metabolic syndrome while also focusing on strategies to decrease proteinuria may be a new and early target in treating patients with CKD.....
.....Our report of the prevalence of metabolic syndrome is higher than the nationally based estimates in the African-American population without CKD in that 41.7% of AASK participants met criteria for metabolic syndrome. We report that the presence of metabolic syndrome predicts a greater level of proteinuria in a high-risk group of African Americans with hypertensive renal disease. This finding is of vast importance because of the escalation of CKD and ESRD rates, as well as the increasing prevalence of obesity/insulin resistance and impaired glucose tolerance.10 We also show that the association of metabolic syndrome and CKD progression is confounded by proteinuria and there is no interaction between metabolic syndrome and proteinuria. These findings suggest that proteinuria is equally predictive of CKD progression in subjects with and without metabolic syndrome, and metabolic syndrome is not an independent predictor of hypertensive renal disease progression....
....Dyslipidemias are reported to be an important factor in progression of CKD. Observational data and a recent meta-analysis suggest that increased triglyceride and low HDL cholesterol levels are independent risk factors for the acceleration of CKD,18, 19 and use of statins may slow the progression of CKD.20 However, in our Cox multivariable model, there was no significant association observed between any of the individual components of metabolic syndrome on progression of CKD, ESRD, or death....
....metabolic syndrome was linked historically to hyperuricemia,27 which was confirmed in our study..."
ABSTRACT
Background
Chronic kidney disease (CKD) is more likely to progress to kidney failure (end-stage renal disease) in African Americans, although the reasons for this are unclear. Metabolic syndrome is a risk factor for the development of diabetes and cardiovascular disease and recently was linked to incident CKD. The purpose of this study is to examine whether metabolic syndrome is associated with kidney disease progression in hypertensive African Americans.
Design & Participants
The current study design is a secondary analysis of the African-American Study of Hypertension and Kidney Disease, a randomized controlled trial of blood pressure goal and agents in hypertensive African Americans with CKD.
Predictors
Metabolic syndrome was defined according to the modified National Cholesterol Education Program guidelines.
Outcomes
Decrease in glomerular filtration rate of 50% or 25 mL/min/1.73 m2, end-stage renal disease (initiation of dialysis therapy or transplantation), death, or a composite outcome of all 3.
Results
842 subjects were included in this analysis, and 41.7% met criteria for metabolic syndrome. Subjects meeting criteria for metabolic syndrome had greater levels of proteinuria. Cox regression analyses adjusted for age, sex, glomerular filtration rate, and other significant covariates except for proteinuria indicated a 31% increased risk, with a 95% confidence interval of 1.03 to 1.7 (P = 0.03) for time to reach the composite outcome in those with metabolic syndrome. Adjusting for proteinuria, the effect was abated to 16% (95% confidence interval, 0.9 to 1.5), no longer remained significant (P = 0.2), and was unchanged by adjusting randomized treatment group (blood pressure goal or antihypertensive drug).
Limitations
Lack of waist circumference as a better surrogate of abdominal obesity.
Conclusions
In summary, metabolic syndrome is associated with proteinuria in hypertensive African Americans, but is not independently associated with CKD progression.
Hypertension is the second leading cause of end-stage renal disease (ESRD) in hemodialysis patients in the United States and accounts for 30% of the ESRD population.1, 2 African Americans are disproportionately affected by hypertensive renal disease, and recent data from the United States Renal Data System showed incidence rates of hypertensive ESRD in African Americans about 7-fold greater than corresponding rates in white Americans of European ancestry.2 Known risk factors for progression of hypertensive renal disease include level of blood pressure (BP) control, male sex, lower socioeconomic status, greater serum creatinine level, and proteinuria.3, 4 The rate of progression of hypertensive ESRD is greater in African Americans compared with whites.5 However, factors that contribute to more rapid progression from chronic kidney disease (CKD) to ESRD in African Americans were not clearly defined.
Metabolic syndrome is a risk factor for the development of diabetes and cardiovascular disease and recently was linked to CKD, mostly in cross-sectional analyses,6, 7 with 1 recent longitudinal study reporting increased risk of incident CKD in persons without diabetes and with metabolic syndrome.8 However, there has been no report of an association between metabolic syndrome and progression of established CKD. The purpose of this study is to determine whether there are differences in the progression of kidney disease in African Americans with hypertensive renal disease according to the presence of metabolic syndrome at baseline entry into the African-American Study of Kidney Disease and Hypertension (AASK) trial and whether any differences were dependent on the level of BP control and/or type of antihypertensive treatment.
Discussion
Our report of the prevalence of metabolic syndrome is higher than the nationally based estimates in the African-American population without CKD in that 41.7% of AASK participants met criteria for metabolic syndrome. We report that the presence of metabolic syndrome predicts a greater level of proteinuria in a high-risk group of African Americans with hypertensive renal disease. This finding is of vast importance because of the escalation of CKD and ESRD rates, as well as the increasing prevalence of obesity/insulin resistance and impaired glucose tolerance.10 We also show that the association of metabolic syndrome and CKD progression is confounded by proteinuria and there is no interaction between metabolic syndrome and proteinuria. These findings suggest that proteinuria is equally predictive of CKD progression in subjects with and without metabolic syndrome, and metabolic syndrome is not an independent predictor of hypertensive renal disease progression.
Individual components of metabolic syndrome were linked to CKD.11, 12, 13, 14, 15, 16, 17, 18, 19 Diabetes and hypertension are well-established risk factors for progression of CKD.11, 12, 13 Obesity results in focal segmental glomerulosclerosis and glomerulomegaly,14 whereas greater BMI is associated with higher GFR and hyperfiltration, possibly leading to increased susceptibility to kidney damage.15 Others showed that greater BMI is associated with lower GFR and CKD.5, 16 Whereas 46.6% of the AASK cohort had BMI greater than 30 kg/m2 and thus were obese, there was no association observed with BMI alone and the renal outcome, which differs from other reports.17 Dyslipidemias are reported to be an important factor in progression of CKD. Observational data and a recent meta-analysis suggest that increased triglyceride and low HDL cholesterol levels are independent risk factors for the acceleration of CKD,18, 19 and use of statins may slow the progression of CKD.20 However, in our Cox multivariable model, there was no significant association observed between any of the individual components of metabolic syndrome on progression of CKD, ESRD, or death.
Perhaps our lack of finding a significant effect of the individual components of metabolic syndrome on the composite of clinical renal events and death was caused by the comprehensive multivariable adjustments. It also is possible that the lack of increased risk of renal events and death in subjects with the presence of various cardiovascular risk factors is related to the large decrease in BP and other associated cardiovascular risk factors experienced by a majority of AASK participants. In other words, it is possible that the greater cardiovascular risk factor burden at baseline entry identified subjects who benefited the most from participation in the trial, thus decreasing their risks of adverse events.
Insulin resistance was associated with microalbuminuria in the Third National Health and Nutrition Examination Survey (NHANES III).7 Furthermore, accumulating data indicate that microalbuminuria clusters with several metabolic abnormalities of cardiometabolic syndrome and is a part of and an early marker for this syndrome.21 Chen et al6 reported that the presence of metabolic syndrome was a strong and independent risk factor for CKD and microalbuminuria, then later reported in a cross-sectional analysis a 2.6-fold increased prevalence of CKD (defined as estimated GFR < 60 mL/min/1.73 m2 [<1.00 mL/s/1.73 m2]) in 6,217 adults in NHANES III with metabolic syndrome.7 Kurella et al8 most recently reported that metabolic syndrome predicted incident CKD in a cohort of 10,096 subjects without diabetes in the Atherosclerosis Risk in Communities Study. Incident CKD was defined as estimated GFR less than 60 mL/min/1.73 m2 (<1.00 mL/s/1.73 m2) at study entry. The multivariable adjusted odds ratio of developing CKD in participants with metabolic syndrome was 1.43 (95% CI, 1.18 to 1.73). After adjusting for the subsequent development of diabetes and hypertension during the 9-year follow-up, the odds ratio for incident CKD in participants with metabolic syndrome was 1.24 (95% CI, 1.01 to 1.51).8 Nonetheless, our report confirms an association of metabolic syndrome and proteinuria, but does not establish metabolic syndrome as an independent predictor of CKD progression.
Results from the main AASK finding showed a significant benefit of initial antihypertensive therapy with an ACE inhibitor, ramipril, in comparison to a dihydropyridine calcium channel blocker, amlodipine, or a β-blocker, metoprolol, for reducing clinical composite outcome (50% or 25 mL/min/1.73 m2 [0.42 mL/s/1.73 m2] decrease in GFR, ESRD or death.9 Despite the decrease in the renal composite outcome observed with ACE-inhibitor-based therapy, there was still progression to ESRD that was not influenced by level of BP control. Our analyses were unchanged by controlling for drug or BP intervention groups, thus negating any differential effect of randomized treatment on proteinuria or surrogates for insulin resistance. Despite the known effects of ACE inhibitors, β-blockers, and calcium channel blockers on insulin sensitivity,22, 23, 24 our findings do not suggest that the antihypertensive agents used in the study influenced our outcome.
One limitation of this study is the absence of waist circumference data to use as a surrogate for abdominal obesity as defined by NCEP. However, we used an alternative measure of obesity in our modified NCEP criteria, BMI, which is an acceptable alternative used in the World Health Organization classification of metabolic syndrome.25 Another limitation is the 242 missing values for triglycerides, which resulted in a smaller data set, likely underestimating prevalent cases of metabolic syndrome. However, there was no difference in subject characteristics between those with missing values and those included in our analyses that would impact on metabolic syndrome status or the renal outcome. Strengths of our study include a well-characterized study population of African Americans with CKD and the most accurate measure of renal function measured every 6 months throughout the study, GFR iodine-125-iothalamate clearances.
This report identifies proteinuria as a predictor of CKD progression in those with and without metabolic syndrome. This is not surprising because microalbuminuria was shown to cluster with metabolic syndrome, and proteinuria is a well-established predictor of CKD progression.22, 23 In the World Health Organization classification of metabolic syndrome, microalbuminuria is included as one of the criteria.26 In our present analyses, we show that subjects defined as having metabolic syndrome according to the modified NCEP criteria had more proteinuria, and no interaction was observed between proteinuria and the metabolic syndrome composite. In a previous post hoc analysis of the AASK trial, we reported that proteinuria level both at baseline and after 6 months of antihypertensive treatment predicted the rate of GFR decrease and risk of ESRD. For every 2-fold increase in baseline proteinuria, there was a 0.54 mL/min/y greater decrease in GFR and 1.8-fold increased risk of ESRD.4 This report also shows that the association of baseline GFR and renal outcome was confounded by proteinuria,4 whereas proteinuria remained a strong and independent predictor of adverse renal outcomes. Finally, metabolic syndrome was linked historically to hyperuricemia,27 which was confirmed in our study. However, uric acid level did not change the relationship between metabolic syndrome and the composite outcome.
In summary, metabolic syndrome is associated with proteinuria in hypertensive African Americans with CKD, but is not independently associated with CKD progression. Moreover, proteinuria predicts CKD progression in those with and without metabolic syndrome. Because of the well-described association of proteinuria and CKD progression, our finding of increased proteinuria in subjects classified to have metabolic syndrome warrants additional studies of possible putative mechanisms of this association and for establishing temporality between metabolic syndrome and proteinuria. Certainly, future investigations of the effect of metabolic syndrome on proteinuria and CKD progression should include more specific measures of insulin resistance. Our data provide a possible mechanism to explain, at least in part, some of the variability observed within and between races in the progression of CKD. Moreover, metabolic syndrome is a modifiable risk factor and thus strategies to decrease the prevalence and/or severity of metabolic syndrome while also focusing on strategies to decrease proteinuria may be a new and early target in treating patients with CKD.
Results
Baseline Clinical Characteristics
Eight hundred forty-two subjects with information about metabolic syndrome status were included in this analysis; 252 of the 1,094 subjects enrolled in AASK were excluded because of missing values for triglyceride level. Mean age was 54.7 ± 10.6 years, and 521 (61.8%) were men. Mean BMI at baseline was 30.6 ± 6.6 kg/m2, with 46.6% greater than 30 kg/m2, and baseline SBP and DBP were 150 ± 24 and 96 ± 14 mm Hg, respectively. Mean glucose level was 94.9 ± 18.5 mg/dL (5.3 ± 1.0 mmol/L; only 13% > 110 mg/dL [>6.1 mmol/L]), mean HDL cholesterol level was 48.3 ± 16 mg/dL (1.24 ± 0.41 mmol/L; 42.6% < 40 mg/dL [<1 mmol/L]), and triglyceride level was 140.5 ± 80.9 mg/dL (1.59 ± 0.91 mmol/L; 33.2% > 150 mg/dL [>1.69 mmol/L]). Serum creatinine level was 2.0 ± 0.7 mg/dL (176.8 ± 62 _mol/L), with a mean GFR of 45.7 ± 13 mL/min/1.73 m2 (76 ± 0.22 mL/s/1.73 m2). Metabolic syndrome was present in 41.7% of participants (351 participants). There were no differences in characteristics of the 252 subjects with missing values compared with the 842 subjects included in this analysis.
Associations between metabolic syndrome and baseline continuous and categorical variables, including individual components of metabolic syndrome, are listed in Table 1. Participants with metabolic syndrome had higher urinary protein excretion (UP/Cr ratio) and serum uric acid level in addition to the expected higher values for glucose, triglycerides, and BMI and lower HDL cholesterol level that define metabolic syndrome (Table 1). Also as listed in Table 1, there were no differences according to metabolic syndrome status between randomized treatment groups stratified on BP goal or antihypertensive therapy.
Cox Proportional Hazards Analysis of Covariates and Individual Components of Metabolic Syndrome for the Composite Outcome (time to GFR event, ESRD, or death)
The number of subjects who reached the composite outcome, as well as the separate events of GFR decrease, ESRD, or death, stratified by metabolic syndrome status is listed in Table 2.
Table 3 lists univariate and multivariable analyses for all covariates grouped according to demographics (age, sex, and alcohol and tobacco use), components of metabolic syndrome (BMI, glucose level, triglyceride level, HDL level, SBP, and DBP), renal function (BUN level, GFR, and UP/Cr ratio), and other laboratory values (hematocrit, uric acid level, and phosphorus level). Significant inverse associations with the composite outcome were observed in men and subjects with increases in hematocrit after multivariable adjustments. Current smokers and subjects with increases in serum BUN level and UP/Cr ratio were associated positively with the outcome after multivariable adjustments. None of the individual components of metabolic syndrome predicted the outcome after multivariable analyses, and SBP and serum glucose level were associated inversely in univariate analyses only. In addition, uric acid level predicted the outcome in univariate analyses only.
As listed in Table 4, risk of reaching the composite outcome (GFR, ESRD, or death) in subjects with metabolic syndrome increased by 31% and was statistically significant without covariate adjustment. Controlling for all covariates except proteinuria, the relationship remained strong and significant. However, adjusting for all covariates plus UP/Cr ratio, the effect of metabolic syndrome on the composite outcome was decreased to only 16% and was no longer statistically significant (Table 4). When treatment assignment (angiotensin-converting enzyme [ACE] inhibitor versus β-blocker versus calcium channel blocker) and BP goal group assignment (low versus usual) were controlled for in the full model, the relationship with the outcome was unchanged.
Interactions for each covariate and the outcome were tested and were not significant and thus not included in the full model. In addition, the interaction term for metabolic syndrome and proteinuria was not statistically significant. Furthermore, we observed that the association of metabolic syndrome with the outcome of ESRD alone and with ESRD and death was stronger than for the composite outcome, but was not statistically significant when adjusted for proteinuria with hazard ratios of 1.21 (95% confidence interval [CI], 0.83 to 1.76) and 1.21 (95% CI, 0.89 to 1.6), respectively.
Correlation analyses did not show significant correlations greater than 0.4 between any of the covariates used in the full model, including proteinuria and the metabolic syndrome composite, and thus there was no evidence of collinearity. Furthermore, there were no strong and significant correlations between proteinuria and any of the individual components of metabolic syndrome.
Survival curves stratified on metabolic syndrome status (Fig 1) show early separation of curves with statistically significant differences in survival favoring the group without metabolic syndrome (P = 0.03). The proportional hazards assumption was met for the metabolic syndrome composite with proteinuria, as well as for time of follow-up. In addition, all covariates met the proportional hazards assumption with follow-up time.
Methods
Participants and Study Design
The AASK trial was a multicenter randomized clinical trial comparing the effect of strict BP control, mean arterial pressure (MAP; <92 mm Hg) or less than 125/75 mm Hg versus usual BP control (MAP, 102 to 107 mm Hg) or 140/90 mm Hg and 3 antihypertensive regimens. Patients were randomly assigned to receive either ramipril, metoprolol, or amlodipine, and other drugs were added stepwise to achieve BP goals.9 Patients were self-identified African Americans who had hypertension (n = 1,094), were aged 18 to 70 years, and had a glomerular filtration rate (GFR) of 20 to 65 mL/min/1.73 m2 (0.33 to 1.08 mL/s/1.73 m2) and no other identified causes of renal insufficiency. Exclusion criteria were: (1) diastolic BP (DBP) less than 95 mm Hg, (2) known history of diabetes mellitus (fasting blood glucose > 140 mg/dL [>7.8 mmol/L] or random glucose > 200 mg/dL [>11.1 mmol/L]), (3) urinary protein-creatinine (UP/Cr) ratio greater than 2.5, (4) accelerated or malignant hypertension within 6 months, (5) secondary hypertension, (6) serious systemic disease, (7) clinical congestive heart failure, or (8) specific indication for or contraindication to a study drug or study procedure.
Main trial results observed were that the ramipril group compared with the amlodipine group had a 38% decreased risk of clinical end points, 36% slower mean decrease in GFR after 3 months, and less proteinuria. The amlodipine treatment arm was terminated at the recommendation of the Data Safety and Monitoring Board based on interim analyses showing a slower mean rate of GFR decrease and reduced rate of clinical end points (rapid decrease in renal function, ESRD, or death) in the ramipril and metoprolol groups. Follow-up to the end of the study ranged from 3 to 6.4 years, with a mean of 4.1 years. The institutional review board at each center approved the protocol and procedures. The present study is a secondary analysis of the randomized controlled trial as a cohort study with the following predictor and outcome variables.
Measurement of Baseline and Follow-up Demographic, Laboratory, and Clinical Data
Baseline information was collected for sociodemographics and lifestyle habits. A central laboratory measured baseline serum and urinary chemistry levels, and GFR was assessed by using iodine-125-iothalamate clearance twice during baseline and at 3 and 6 months and every 6 months thereafter during follow-up. GFR is expressed after standardization for body surface area in units of milliliters per minute per 1.73 m2 (milliliters per second per 1.73 m2). Urinary protein excretion is expressed as UP/Cr ratio from a 24-hour urine collection. Three consecutive seated BPs were measured using a Hawksley random zero sphygmomanometer (Hawksley, Lancing, West Sussex) after at least 5 minutes of rest, with the mean of the 2 last readings recorded.9 Baseline BPs were those obtained at the screening visits before randomization.
Predictor Variable: Definition of Metabolic Syndrome
All participants enrolled in the AASK trial had their baseline data analyzed to determine whether criteria for metabolic syndrome were met by using a modified National Cholesterol Education Program (NCEP) definition of metabolic syndrome, defined as any 3 of the following: body mass index (BMI) greater than 30 kg/m2, fasting plasma glucose level of 110 mg/dL or greater (≥6.1 mmol/L), triglyceride level of 150 mg/dL or greater (≥1.69 mmol/L), high-density lipoprotein (HDL) cholesterol level less than 50 mg/dL (<1 mmol/L) in women and less than 40 mg/dL (<1 mmol/L) in men, and hypertension with BP greater than 135/85 mm Hg or use of antihypertensives. All participants in AASK had a diagnosis of hypertension and thus only 2 additional criteria were needed to define metabolic syndrome from either BMI, glucose, HDL cholesterol, or triglyceride measurements.
Other Predictor Variables
Other covariates examined and adjusted for that are known to influence renal function were age, sex, smoking status, alcohol use, blood urea nitrogen (BUN) level, GFR, UP/Cr ratio, serum phosphorus level, uric acid level, and hematocrit.
Outcome Variables
The outcome variables in this report are clinical events given by the time from randomization to either: (1) a GFR event, defined as confirmed decrease in GFR by either 50% or 25 mL/min/1.73 m2 (42 mL/s/1.73 m2) from the mean of 2 baseline GFR measurements; (2) occurrence of ESRD; (3) death; or (4) a composite of all 3.
Statistical Analyses
Before statistical analyses, we selected the factors listed in Table 1 as potential risk factors for investigation. Continuous variables are expressed as mean ± SD and compared by using Student t-test, whereas categorical variables are expressed as a percentage and compared by using _2 analysis. Cox regression analyses were used to assess the impact of metabolic syndrome on the time to each event in addition to the composite outcome to account for differences in follow-up time. Time-to-event analyses of the combined clinical composite of GFR decrease, ESRD, or death for metabolic syndrome status and for each covariate were examined in univariate analyses. Multivariable analyses for each covariate were adjusted for age, sex, tobacco and alcohol use, GFR, BUN level, UP/Cr ratio, phosphorus level, uric acid level, hematocrit, and for each component of metabolic syndrome: BMI, glucose level, HDL cholesterol level, triglyceride level, systolic BP (SBP), and DBP cumulatively. Multivariable analyses for the composite of metabolic syndrome with the composite outcome were adjusted for the other covariates not included in the definition of metabolic syndrome.
|
|
| |
| |
|
|
|
