Search tips
Search criteria 


Logo of ndtLink to Publisher's site
Nephrol Dial Transplant. 2009 December; 24(12): 3677–3686.
Published online 2009 September 10. doi:  10.1093/ndt/gfp471
PMCID: PMC2790952

Adrenergic beta-1 receptor genetic variation predicts longitudinal rate of GFR decline in hypertensive nephrosclerosis


Background. End-stage renal disease (ESRD) due to hypertension is common and displays familial aggregation in African Americans, suggesting genetic risk factors, including adrenergic activity alterations which are noted in both hypertension and ESRD.

Methods. We analysed 554 hypertensive nephrosclerosis participants (without clinically significant proteinuria) from the longitudinal National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) African American Study of Kidney Disease and Hypertension (AASK) cohort to determine whether decline in glomerular filtration rate (GFR) over ~3.8 years was predicted by common genetic variation within the adrenergic beta-1 (ADRB1) receptor at non-synonymous positions Ser49Gly and Arg389Gly.

Results. The polymorphism at Ser49Gly (though not Arg389Gly, in only partial linkage disequilibrium at r2 = 0.18) predicted the chronic rate of GFR decline, with minimal decline in Gly49/Gly49 (minor allele) homozygotes compared to Ser49 carriers; concordant results were observed for haplotypes and diploid haplotype pairs at the locus. An independent replication study in 1244 subjects from the San Diego Veterans Affairs Hypertension Cohort confirmed that Gly49/Gly49 homozygotes displayed the least rapid decline of eGFR over ~3.6 years.

Conclusion. We conclude that GFR decline rate in hypertensive renal disease is controlled in part by genetic variation within the adrenergic pathway, particularly at ADRB1. The results suggest novel strategies to approach the role of the adrenergic system in the risk and treatment of progressive renal disease.

Keywords: AASK, African Americans, glomerular filtration rate, kidney disease, sympathetic nervous system


End-stage renal disease (ESRD) resulting from hypertension aggregates in families [1,2] and is approximately six times more likely in African Americans than white subjects [3]. Recently, genetic studies have linked hypertensive ESRD to polymorphisms in myosin heavy chain-9 (MYH9) [4] and chromogranin A (CHGA)[5], which has a pivotal role in formation of catecholamine storage vesicles [6]. The adrenergic system, implicated in a number of cardiovascular disease processes [7], contributes to the progression of renal failure [8]. Though activation of the adrenergic receptors is vital for physiologic processes, prolonged stimulation likely contributes to cardiovascular disease states, and thus the sensitization and down-regulation of these receptors may play a role in such chronic diseases [9,10].

Many reports have noted genetic contributions from adrenergic beta-1 receptors (ADRB1) to hypertension [11,12], heart failure [13] and coronary artery disease [14]. Given the relationship between cardiovascular and kidney disease [15], we elected to study subjects from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) African American Study of Kidney Disease and the Hypertension (AASK) trial [16] to evaluate common genetic variants of ADRB1 receptors as predictors of glomerular filtration rate (GFR) decline in hypertensive nephrosclerosis. The AASK trial is ideally suited to study such genetic variants, since it is a large-scale longitudinal cohort evaluating the impact of blood pressure (BP) goals (lower versus usual) and antihypertensive medications (ramipril, metoprolol, or amlodipine) on the progression of hypertensive nephropathy [17,18]. Subjects were carefully monitored for progressive hypertensive nephropathy [19], with GFR measured regularly through iothalamate clearance for a median of ~3.8 years [17].

Previous reports from the AASK trial indicate that subjects with minimal proteinuria (defined as a urine protein-to-creatinine ratio (UPCR) <0.22 g/g, approximating the ~300 mg/day threshold defining clinically significant proteinuria) display the most dynamic acute changes in GFR in response to drugs [18,20]. Such subjects without elevated proteinuria also have a slower rate of decline of renal function (−1.35 mL/min/1.73 m2/year when compared to −4.09 mL/min/1.73 m2/year in subjects with overt proteinuria) [20]. Since accelerated decline in GFR in the overt proteinuria group may be further confounded by proteinuria as an independent risk factor for renal function decline [20], we evaluated AASK study subjects without clinically significant proteinuria to determine whether risk of accelerated loss of kidney function is conferred by common genetic variation at ADRB1, focusing on the only two common (MAF>5%) non-synonymous (amino acid replacement) variants at ADRB1 which have also been shown to affect functional down-regulation of the receptor, Ser49Gly and Arg389Gly [21], shown in Figure Figure1,1, assessed as single nucleotide polymorphisms (SNPs), haplotypes and diploid haplotypes.

Fig. 1
Pictoral depiction of human adrenergic beta-1 receptor (ADRB1) gene. The illustration shows the only two common (minor allele frequency >5%) non-synonymous polymorphisms found in humans: Ser49Gly and Arg389Gly. Based on <>. ...



AASK study subjects

Subjects were from the AASK study, a 21-centre randomized, controlled prospective trial that has been previously described [16,17]. Briefly, participants were 18- to 70-year-old self-identified African Americans with hypertension (n = 1094) with a clinical diagnosis of hypertensive renal disease, documented by an initial GFR (by [125I]-iothalomate clearance or Cockroft-Gault equation) between 20 and 65 ml/min/1.73 m2, with UPCR <2.5 g/g, and no other identifiable causes of renal insufficiency. In particular, subjects with diabetes mellitus were excluded. Based on a 3 × 2 factorial design, participants were randomized to one of two goal BP ranges (‘usual’ mean arterial pressure goal of 102– 107 mmHg, or a lower mean arterial pressure goal of ≤92 mmHg), and to double-blinded treatment with one of three antihypertensive drug classes (40% to beta-blockade with metoprolol, 50–200 mg/day; 40% to ACE inhibition with ramipril, 2.5–10 mg/day; or 20% to calcium channel blockade with amlodipine, 5–10 mg/day). GFR was assessed by renal clearance of [125I]-iothalamate at baseline twice, at 3 and 6 months and then every 6 months thereafter, for a median of 3.8 years (range 3.0–6.4 years) [20]. The primary endpoint in this study was chronic GFR slope, defined as that beginning 3 months after randomization, in order to better reflect long-term disease progression, given that the medical interventions (particularly calcium channel blockade) were noted to have acute effects on GFR that may differ from their long-term effects on disease progression [18,22].

Genomic DNA from blood leucocytes was ascertained and prepared from 830 of the original 1094 participants in the AASK study who consented for genetic evaluation. Each subject gave informed, written consent to the local institutional review boards. Of those, the 554 that did not have clinically significant proteinuria, defined as a UPCR <0.22 and were genotyped at the ADRB1 loci, were included in this study and are described in Table Table1.1. UPCR of <0.22 corresponds to a urine protein excretion of ~300 mg/day, and separates out the ~2/3 of patients with the lowest (‘normal’ range) proteinuria from the 1/3 with highest proteinuria (the UPCR parameter is skewed towards higher values and is inversely associated with initial GFR). The 328 males and 226 females had a mean age of 55.6 ± 0.4 years and a baseline GFR of 51.2 ± 0.5 mL/min/1.73 m2. The genotyped AASK subset (n = 830) did not differ from the complete AASK cohort (n = 1094) [17,19,23] in age, sex, BMI, duration of hypertension, SBP, DBP, serum creatinine, baseline GFR, baseline proteinuria or overall chronic rate of the progression of renal disease (by normal and high proteinuria groups).

Table 1
Demographics of NIDDK AASK genome study participants

Replication study: San Diego Veterans Affairs Hypertension Cohort (VAHC)

The San Diego VAHC is a multiethnic, non-interventional/ observational study of veterans recruited from San Diego Veterans Affairs Healthcare System primary care internal medicine clinics in 2003–2004 [24]. Subjects had been diagnosed by their primary care physicians with essential hypertension or were on BP-lowering medications, excluding subjects with known secondary causes of hypertension. The cohort utilizes the comprehensive VA complete electronic medical record (EMR) known as VISTA (Veterans Health Information Systems and Technology Architecture; [25], which includes all vital signs, laboratory data, medical diagnoses, pharmacy records and procedure codes in digital format. EMR data were extracted in Microsoft SQL-Server tables for study subjects from October 2000 to November 2007. Subjects were excluded if they had prevalent or incident ESRD during this period (n = 25). eGFR was estimated by the Modification in Diet and Renal Disease (MDRD) equation [26].

A total of 1244 subjects were genotyped at the ADRB1 locus and included in this replication study as described in Table Table2.2. Of those, 95.6% were men, reflecting the veteran population, with 68.2% self-identified as white and 15.5% as black. At entry, 15.7% had evidence of chronic kidney disease (CKD), defined by eGFR <60 mL/min/1.73 m2 [27]. By the end of the study, 30.1% had a diagnosis of type-2 diabetes with <2% of them coded for renal complications of diabetes. The mean age at the start of the study was 61.9 ± 0.4 years (range 24.5–89.6) with a mean follow-up of 3.64 ± 0.05 years (range < 1 to 6.9 years). The primary endpoint in this study was chronic eGFR slope (change over time) in order to better reflect long-term disease progression. For subjects who were determined to have an episode of acute kidney injury (AKI), defined as an acute increase in plasma creatinine >0.3 mg/dL [28], only the eGFR measurements prior to the AKI were used in longitudinal analysis.

Table 2
Subject demographics of San Diego Veterans Affairs Hypertension Cohort replication study


SNP genotyping and haplotypes at ADRB1

Two common polymorphisms, illustrated in Figure Figure1,1, at ADRB1 were genotyped (Ser49Gly and Arg389Gly) in the AASK study, giving rise to three common haplotypes and three diploid haplotype combinations, as listed in Table Table33.

Table 3
Effects of adrenergic beta-1 receptor (ADRB1) genetic variants on chronic glomerular filtration rate (GFR) slope in NIDDK AASK subjects

SNPs were obtained from the public SNP database (http://www.ncbi. Genomic DNA was typed by an immobilized probe approach [29], in which each DNA sample was amplified in a multiplex polymerase chain reaction (PCR) using biotinylated primers. Each PCR product pool was then hybridized to a panel of sequence-specific oligonucleotide probes immobilized in a linear array. The colorimetric detection method was based on the use of streptavidin–horseradish peroxidase conjugate with hydrogen peroxide and 3,3′,5,5′-tetramethylbenzidine as substrates. Reproducibility of genotyping was verified with 50 blinded replicate samples. Both SNPs evaluated in this study were in Hardy–Weinberg equilibrium (P > 0.05) and were in incomplete linkage disequilibrium as determined by the Hill algorithm [30] (r2 = 0.18).

Statistical analyses

Progression of GFR loss in AASK

Statistical analyses for the AASK study were performed using Statistical Package for the Social Sciences (SPSS v11.0; Chicago, IL, USA). ANOVA was used to determine the significance of genotype and haplotype effects on chronic decline of GFR slope (mL/min/1.73 m2/year). Mixed-effects models were utilized to determine the slope and were estimated by restricted maximum likelihood in each treatment group and considered covariates including treatment centre, proteinuria, sex, age and mean arterial pressure [18]. Various models included covariates in the analysis. Initial P-value is unadjusted, without any covariates, Model 1 includes baseline age and sex, Model 2 includes baseline GFR and UPCR (urine protein: creatinine ratio) and Model 3 includes baseline GFR, baseline UPCR, age at the start of the study, sex and randomized drug group (ramipril, metoprolol or amlodipine) and BP goal (low or usual). A conservative Bonferroni correction was applied for the two ADRB1 SNPs studied in this analysis, for an adjusted significance level of P = 0.025 (= 0.05/2) despite evidence of linkage disequilibrium between the two variants [10], with r2 = 0.18. All analyses were performed by the study investigators and not by the AASK data coordinating centre.

Population admixture

African Americans represent an admixed population with genetic contributions from both African and European biogeographic origins [31]. To confirm that AASK individuals with or without trait-associated genotypes were of comparable overall genetic background, and the observed associations are not simply an artefact of differential admixture between faster and slower GFR decline rates, generalized analysis of molecular variance (GAMOVA) [32] was used to test for and quantify the relationship between the overall genetic background of the subjects and quantitative phenotype GFR decline rate (chronic GFR slope), with an IBS (identity-by-state) distance matrix based on genotypes at 126 bi-allelic markers. The admixture analysis was done in this study cohort without clinically significant proteinuria (N = 554), as well as in the entire genomic AASK cohort (N = 830).

Replication study in VAHC

Mixed-effects linear models (PROC MIXED) in SAS 9.2 (Cary, NC, USA) [33] were utilized to assess the influence of ADRB1 Ser49Gly polymorphism on longitudinal ambulatory eGFR considering sex, age at start of study and race [18,34]. Subject-specific intercept and slope parameters were estimated and regressed against the predictor of interest simultaneously using mixed-effects models [33,35,36]. This approach can be applied to unbalanced longitudinal data, particularly to the repeated measurements taken at unequal time intervals of the VAHC study, which utilizes observational information in the EMR and yields information on both genotype and genotype-by-time effects [37]. Because only one polymorphism was studied in this cohort, P = 0.05 was considered statistically significant in this replication study.

To illustrate the influence of the ADRB1 Ser49Gly variant on longitudinal GFR decline, chronic longitudinal GFR was plotted by genotype over time (Figure (Figure3).3). The longitudinal GFR profile by genotype was generated using an adaptive regression cubic spline approach [38], with 95% confidence intervals for fitted spline function of the adjusted GFR values (from mixed model results). To address the potential confounding effects of loss to follow-up, joint modelling of longitudinal and time to event data was implemented [33]. This joint model combines longitudinal and time to event analysis, allowing the covariance structure to be adjusted for censoring (loss to follow-up) and event (death) data [33,39].

Fig. 3
Replication of adrenergic beta-1 receptor (ADRB1) genotype: Effects on estimated glomerular filtration (eGFR) decline rate in the San Diego Veterans Affairs Hypertension Cohort. ADRB1 Ser49Gly polymorphism predicts the longitudinal decline in eGFR (P ...

Power analyses

Statistical power was determined using the online instrument G*Power 3 ( aap/gpower3/literature) [40]. More than 90% power to detect an effect size of 0.15 would be achieved with a sample size of 554.

Haplotype inference

Two common variants were genotyped at ADRB1, allowing the inference of haplotypes from unphased diploid genotypes by the HAP imperfect phylogeny method (version 3.0; [41].


Primary studies in NIDDK AASK subjects with hypertensive nephrosclerosis

ADRB1 polymorphisms and haplotypes contribute to GFR slope

This genomic study cohort of 554 AASK study participants genotyped at the ADRB1 locus (Table (Table1)1) had a mean age of 55.6 ± 0.4 years and initial GFR of 51.2 ± 0.5 mL/min/1.73 m2. Table Table33 reports the results of association testing of ADRB1 genetic variants (polymorphisms, haplotypes and diploid haplotypes) on chronic GFR decline.

ADRB1 Ser49Gly affected GFR slope: Gly49/Gly49 minor allele homozygosity predicted slower decline of GFR (P = 0.003, Figure Figure2A),2A), which maintained significance when considering covariates in alternative Models 1, 2 and 3, reported in Table Table3.3. Arg389Gly alone did not show an independent effect (P = 0.53), but haplotypes composed of Ser49Gly and Arg389Gly predicted GFR decline: subjects with two copies of haplotype-3 (Gly49Arg389) displayed a slower GFR decline rate than those with zero or one copy (P = 0.002; Figure Figure2B),2B), again maintaining significance when considering several covariates. In analyses of diploid haplotypes, haplotype-3/haplotype-3 homozygotes had a decreased rate of GFR decline compared to those with copies of haplotype-1 (P = 0.001, Figure Figure2C)2C) or haplotype-2 (P = 0.014, Figure Figure2D).2D). No significant changes were noted when covariates were entered into the models, seen in Table Table33.

Fig. 2
Adrenergic beta-1 receptor (ADRB1) genetic variants: Effects on chronic glomerular filtration (GFR) decline rate in the NIDDK AASK cohort. AASK subjects without clinically significant proteinuria (urine protein/creatinine ratio <0.22 g/g) were ...

No gene-by-drug interactions (based upon trial randomization groups) were noted with the ADRB1 variants. Nor did the Ser49Gly polymorphism associate with drug or BP-goal randomization groups, age, GFR, UPCR or BMI at the start of the study.

Genetic admixture in NIDDK AASK

One hundred and twenty-six bi-allelic markers were genotyped in the genomic AASK cohort. Generalized analysis of molecular variance (GAMOVA) indicated that people with similar predictor variables (GFR slope, mL/min/1.73 m2/year) were not genetically more related to each other than expected by chance alone in this group of AASK subjects without clinically significant proteinuria (P = 0.87), nor in the entire genomic AASK cohort (n = 830, P = 0.45). Thus, the adrenergic allele and haplotype influences on this chronic GFR decline trait cannot be attributed to differential admixture between higher and lower GFR decline rate strata.

Replication study: San Diego Veterans Affairs Hypertension Cohort (VAHC)

This replication study consisted of 1244 veterans genotyped at the ADRB1 Ser49Gly locus (minor allele frequency 14.8%), as described in Table Table2.2. Of those, 95.6% were male and 68.2% were white. At the start of the study the mean age was 61.9 ± 0.35 years, serum creatinine was 1.04 ± 0.01 mg/dL, eGFR (by MDRD equation) was 85.6 ± 0.7 mL/min/1.73 m2 and body mass index was 30.1 ± 0.16 kg/m2.

As noted in the AASK subjects, ADRB1 Ser49Gly predicted decline in eGFR over a mean follow-up of 3.6 ± 0.05 years in the VAHC cohort, adjusted by sex, age at start of study and race, shown in Figure Figure3.3. ADRB1 Ser49Gly displayed a significant genotype-by-time interaction on eGFR, at P = 0.043. Subjects homozygous for the minor allele (Gly49/Gly49) had slower decline in the slope of eGFR decline compared to heterozygotes (Ser49/Gly49) or wild-type homozygotes (Ser49/Ser49). A joint model, which combines longitudinal and time to event analysis adjusting for censoring (loss to follow-up) and death, yielded consistent results (data not shown, P = 0.045).

When restricting the analysis only to subjects self-identified as African American (n = 191, minor allele frequency = 14.8%), statistical significance was maintained in this model with sex and age at start of study as covariates (P = 0.037), again with Gly49/Gly49 homozygotes having slower decline in eGFR. The findings in other ethnic groups, including whites and Asian/Pacific Islanders, showed a similar trend (with Gly49/Gly49 variants having slower eGFR decline), but the difference did not reach statistical significance. Similarly, when evaluating subjects with CKD (eGFR at the start of the study ≤60 mL/min/1.73 m2, n = 195), a significant genotype-by-time interaction on eGFR was noted (P = 0.047) when considering sex, race and age at the start of the study as covariates. A similar trend was noted for the gene-by-time interaction in subjects that started the study without CKD (eGFR > 60 mL/min/1.73 m2), but did not reach statistical significance (P = 0.19).

The Ser49Gly variant was not associated with sex, self-identified ethnicity, BMI, serum creatinine or eGFR at the start of the study (results not shown) in the VAHC study. The participants differed in age at baseline (P = 0.025), with Gly49/Gly49 homozygotes slightly younger (57.5 ± 2.3 years) than Ser49/Ser49 (62.3 ± 0.41 years) or Ser49/Gly49 (61.2 ± 0.71 years) subjects. Subjects with CKD at the start of the study were noted to be older (67.7 ± 0.78 years) compared with those without CKD (61.0 ± 0.37, P < 0.001).



The autonomic nervous system is a key determinant of blood pressure and renal function. Previous studies showed that the adrenergic response to stress contributes to hypertension, proteinuria and renal pathologic changes in rats [42], and catecholamine infusion worsened GFR decline and proteinuria in humans [43]. African Americans, who are at an increased risk for hypertension and nephrosclerosis particularly within families, display exaggerated autonomic responses to environmental stressors [44]. Here, we found that common non-synonymous variation of ADRB1 predicted the rate of GFR decline in the NIDDK AASK cohort of hypertensive nephrosclerosis (Table (Table3)3) and then replicated the finding in an independent population of hypertensive veterans.

Adrenergic hereditary mechanisms

The adrenergic system, through adrenergic receptors, regulates multiple aspects of renal function [45] as well as systemic haemodynamics [46] and is under substantial hereditary control [47]. In this study, we found that ADRB1 variants affected GFR slope (Table (Table3).3). ADRB1 receptors are expressed throughout the body in many regulatory pathways, but in the kidney, ADRB1 is expressed in renal arterial smooth muscle cells as well as in the mesangium [48]. The ADRB1 GPCR is typically stimulatory, coupling through Gs to increase adenylyl cyclase activity and hence cyclic AMP (cAMP). cAMP dilates renal arterioles and mesangial cells, thereby regulating not only renal blood flow but also filtration surface, consequently altering glomerular capillary hydrostatic pressure, which plays a role in progressive renal disease [49]. Adrenergic genetic variants have been found to be functional due to a number of mechanisms, such as altering functional receptor number, binding affinity, agonist coupling to effectors, or desensitization [50,51], each with the potential to affect GFR decline.

ADRB1 Ser49Gly lies within the extra-cellular amino terminus of the receptor, shown in Figure Figure1,1, and influences desensitization of the receptor after repeated agonist exposure [52]; the wild-type Gly49 variant exhibits heightened desensitization (agonist-promoted down-regulation of signalling), particularly with prolonged agonist stimulation [52–54]. This mechanism appears to limit ADRB1 signalling in response to chronic catecholamine excess, and thus Gly49 has been shown to be beneficial in subjects with heart failure [54].

The other ADRB1 variant included in this study, Arg389Gly, which was not independently correlated with decline in renal function in our study, has been evaluated extensively due to its location within the intra-cellular carboxy-terminal tail (Figure (Figure1)1) that constitutes the Gs coupling domain of the receptor. In recombinant systems, Arg389 exhibited increased coupling to Gs with higher adenylyl cyclase activity but greater short-term desensitization than Gly389 [21].

ADRB1 haplotypic variation (at Ser49Gly→Arg389Gly) also has functional consequences. Sandilands et al. [51] found that the β-adrenergic agonist isoprenaline/ isoproterenol increased cAMP production most prominently in the Gly49Arg389 haplotype, which also displayed the greatest isoprenaline-induced desensitization. Our findings were consistent in that subjects with two copies of the Gly49Arg389 haplotype (Figure (Figure2B)2B) had a decreased rate of GFR decline. When evaluating diploid haplotypes, the subjects with diploid copies of the haplotype also had a decreased decline in their renal decline compared with subjects with one or two copies of either Ser49Arg389 (Figure (Figure2C)2C) or Ser49Gly389 (Figure (Figure2D),2D), suggesting a dominant action of this haplotype on the GFR decline trait.

These ADRB1 variants have been evaluated for roles in cardiac disease [55–57], hypertension [11,58], resting heart rate [58] and mortality [12], including pharmacogenetic studies [11,12], with mixed results. The studies with significant findings were consistent with our results in that those with the Gly49 variant have improved cardiac outcomes [55,56], and also improved mortality [12], particularly when treated with a beta-blocker. The cardiac studies that were unable to determine associations with ADRB1 Ser49Gly had null findings and may have lacked statistical power [57]. This variant alone has not been shown to be independently associated with hypertension or response to beta-blocker therapy [11,59]. Thus, our findings here of improved renal decline are unlikely to be related simply to BP control or response to therapy, but whether they may be associated with improved cardiac function was not assessed here. The Arg389Gly polymorphism has been linked to hypertension [59], which likely contributes to ADRB1 haplotypes that have previously been associated with hypertension and response to beta-blocker therapy [11,60].

Given that ADRB1 stimulation increases renin release [61], further study is necessary to determine whether plasma renin activity is affected by ADRB1 genetic variants. If the Gly49 allele results in desensitization and down-regulation, decreased renin may be released, as has been noted in beta-1/beta-2-adrenergic receptor-deficient mice [62].

ADRB1 genetic associations with chronic GFR slope in the NIDDK AASK cohort

ADRB1 genetic associations with GFR slope were found in AASK subjects without clinically significant proteinuria. Previous reports from the AASK trial indicate that subjects with minimal proteinuria display the greatest acute dynamic changes in GFR in response to drugs, while elevated proteinuria itself is a strong predictor of GFR decline [63]. Minimal proteinuria likely corresponds to less advanced pathological findings, such as fewer initial structural and fibrotic changes of nephrosclerosis; at this earlier stage, the regulation of GFR may be more sensitive to haemodynamic changes, with loss of such sensitivity as fibrosis progresses. The SNPs may also require active dynamic regulation of glomerular vessels, which may be absent when UPCR ≥0.22 g/g as previously seen in the AASK cohort acutely when BP meds were started [20]. A limitation to this study is that such conclusions may not be readily extended to subjects with more advanced proteinuria.

Replication study

An exact replication of the AASK trial is unlikely, given the scope of its duration, size, intensity, complexity and expense. Though randomized clinical trials are often considered the gold standard for clinical research, detailed comparison with observational studies via meta-analysis provides evidence that well-designed observational studies can produce valid and similar results [64,65]. The VAHC has many characteristics which make it an ideal study population: primary care ascertainment basis, detailed and comprehensive EMR, nearly complete picture of medical use from a single provider, standard drug formulary and the ability to track subjects longitudinally. Thus, we used the VAHC study as a replication study. Though the VISTA–EMR system provides a vast amount of medical information about the study subjects, one limitation is that accuracy and completeness are reliant upon the EMR. Advanced statistical methods (see the Method section) were employed to best utilize and transform the ambulatory setting data into models that may emulate a clinical trial.

These results are from veteran subjects, who are primarily male and with heterogeneous biogeographic ancestry. This population was generally elderly (at 61.9 ± 0.35 years) with a corresponding modest decrease in eGFR (at 85.6 ± 0.7 mL/min/1.73 m2), though not so low as GFR in subjects with nephrosclerosis from the AASK study (Tables (Tables11 and and2).2). Because of the relatively small number of African Americans (n = 191) in VAHC, we opted to study all of the subjects using ethnicity as a covariate. Nonetheless, even when confining the replication to only the African American subjects, the association of ADRB1 Ser49Gly genotype-by-time decline in eGFR maintained statistical significance (P = 0.037).

Because the VAHC cohort is based upon the EMR, iothalomate clearance results, which were utilized in the AASK study, were not available. Rather, eGFR was determined by the MDRD equation [26], using ambulatory serum creatinine measurements all performed at the same San Diego Veterans Affairs clinical laboratory. Such eGFR values are most accurate in subjects with GFR <60 mL/min/1.73 m2, tending to underestimate GFR in healthy individuals [26]. Though our sample included healthy subjects as well as those with CKD stages 1–4 [27], imprecision of eGFR determination by underestimation in healthy individuals would be expected to bias the results towards the null, in contrast to the significant findings in our analyses (Figure (Figure3).3). Despite the small subject number (n = 195) in the subset of subjects with CKD, statistical significance for an ADRB1 gene-by-time interaction with eGFR (P = 0.047) was maintained.

Statistical confidence

In a comprehensive study of a trait involving multiple genotypes, the possibility of false-positive (type 1) statistical conclusions must be considered. We approached this issue in four ways: reducing the target α (P value) by Bonferroni correction based on the number of SNPs evaluated, by haplotyping (i.e., analysing multiple variants at a locus simultaneously), and by admixture analyses. These approaches continued to yield significant GFR predictions, with results that were congruent across approaches. Nonetheless, the ultimate guardian against false-positive errors is replication in an independent sample, which we tested in the VAHC cohort.


The adrenergic system, including its cardiorenal effector the ADRB1 receptor, is implicated in the genesis and the progression of hypertension and cardiovascular disease. Our study utilized the large longitudinal AASK genomic cohort and found that the rate of GFR decline in nephrosclerosis is controlled in part by common genetic variation within the ADRB1 receptor, perhaps influencing renal function as a consequence of enhanced receptor desensitization. The results suggest novel pathophysiological links between the adrenergic system and chronic renal injury and suggest new strategies for probing the role and actions of the pathway in this setting. These findings may generate hypotheses for more rigorous testing of the role of adrenergic receptors in complex renal traits, which may assist in further understanding pathophysiological principles and designing potential therapeutic interventions that might be implemented after early identification of genetic risk.


We appreciate the support of the Department of Veterans Affairs, the NIH/NCMHD-sponsored (MD000220) EXPORT Minority Health Center, as well as the NIH/NCRR-sponsored (RR00827) General Clinical Research Center. NJS is supported in part by the Scripps Translational Sciences Institute Clinical Translational Science Award (U54 RR0252204-01).

Conflict of interest statement. VHB is employed by Roche Molecular Systems, Inc. that does not have a business interest in this field.

(See related article by F. C. Luft. Not so free associations. Nephrol Dial Transplant 2009; 24: 3576–3577.)


1. Freedman BI, Spray BJ, Tuttle AB, et al. The familial risk of end-stage renal disease in African Americans. Am J Kidney Dis. 1993;21:387–393. [PubMed]
2. Ferguson R, Grim CE, Opgenorth TJ. A familial risk of chronic renal failure among blacks on dialysis? J Clin Epidemiol. 1988;41:1189–1196. [PubMed]
3. Rahman M, Douglas JG, Wright JT., Jr Pathophysiology and treatment implications of hypertension in the African-American population. Endocrinol Metab Clin North Am. 1997;26:125–144. [PubMed]
4. Freedman BI, Hicks PJ, Bostrom MA, et al. Polymorphisms in the non-muscle myosin heavy chain 9 gene (MYH9) are strongly associated with end-stage renal disease historically attributed to hypertension in African Americans. Kidney Int. 2009;75:736–745. [PMC free article] [PubMed]
5. Salem RM, Cadman PE, Chen Y, et al. Chromogranin A polymorphisms are associated with hypertensive renal disease. J Am Soc Nephrol. 2008;19:600–614. [PubMed]
6. Taupenot L, Harper KL, O’Connor DT. The chromogranin-secretogranin family. N Engl J Med. 2003;348:1134–1149. [PubMed]
7. Mancia G, Grassi G, Giannattasio C, et al. Sympathetic activation in the pathogenesis of hypertension and progression of organ damage. Hypertension. 1999;34(Pt 2):724–728. [PubMed]
8. Converse RL, Jacobsen TN, Jr, Toto RD, et al. Sympathetic overactivity in patients with chronic renal failure. N Engl J Med. 1992;327:1912–1918. [PubMed]
9. Bristow MR, Ginsburg R, Minobe W, et al. Decreased catecholamine sensitivity and beta-adrenergic-receptor density in failing human hearts. N Engl J Med. 1982;307:205–211. [PubMed]
10. Sandilands AJ, O’Shaughnessy KM. The functional significance of genetic variation within the beta-adrenoceptor. Br J Clin Pharmacol. 2005;60:235–243. [PMC free article] [PubMed]
11. Johnson JA, Zineh I, Puckett BJ, et al. Beta 1-adrenergic receptor polymorphisms and antihypertensive response to metoprolol. Clin Pharmacol Ther. 2003;74:44–52. [PubMed]
12. Pacanowski MA, Gong Y, Cooper-Dehoff RM, et al. beta-adrenergic receptor gene polymorphisms and beta-blocker treatment outcomes in hypertension. Clin Pharmacol Ther. 2008;84:715–721. [PMC free article] [PubMed]
13. Forleo C, Resta N, Sorrentino S, et al. Association of beta-adrenergic receptor polymorphisms and progression to heart failure in patients with idiopathic dilated cardiomyopathy. Am J Med. 2004;117:451–458. [PubMed]
14. Hassan M, York KM, Li H, et al. Association of beta1-adrenergic receptor genetic polymorphism with mental stress-induced myocardial ischemia in patients with coronary artery disease. Arch Intern Med. 2008;168:763–770. [PubMed]
15. Mann JF, Gerstein HC, Pogue J, et al. Renal insufficiency as a predictor of cardiovascular outcomes and the impact of ramipril: the HOPE randomized trial. Ann Intern Med. 2001;134:629–636. [PubMed]
16. Sika M, Lewis J, Douglas J, et al. Baseline characteristics of participants in the African American Study of Kidney Disease and Hypertension (AASK) Clinical Trial and Cohort Study. Am J Kidney Dis. 2007;50:78–89. [PubMed]
17. Wright JT, Jr, Kusek JW, Toto RD, et al. Design and baseline characteristics of participants in the African American Study of Kidney Disease and Hypertension (AASK) Pilot Study. Control Clin Trials. 1996;17(4 Suppl):3S–16S. [PubMed]
18. Agodoa LY, Appel L, Bakris GL, et al. Effect of ramipril versus amlodipine on renal outcomes in hypertensive nephrosclerosis: a randomized controlled trial. JAMA. 2001;285:2719–2728. [PubMed]
19. Fogo A, Breyer JA, Smith MC, et al. AASK Pilot Study Investigators. Accuracy of the diagnosis of hypertensive nephrosclerosis in African Americans: a report from the African American Study of Kidney Disease (AASK) Trial. Kidney Int. 1997;51:244–252. [PubMed]
20. Wright JT, Bakris G, Greene T, et al. Effect of blood pressure lowering and antihypertensive drug class on progression of hypertensive kidney disease: results from the AASK trial. JAMA. 2002;288:2421–2431. [PubMed]
21. Rathz DA, Gregory KN, Fang Y, et al. Hierarchy of polymorphic variation and desensitization permutations relative to beta 1- and beta 2-adrenergic receptor signaling. J Biol Chem. 2003;278:10784–10789. [PubMed]
22. Hall WD, Kusek JW, Kirk KA, et al. Short-term effects of blood pressure control and antihypertensive drug regimen on glomerular filtration rate: the African-American Study of Kidney Disease and Hypertension Pilot Study. Am J Kidney Dis. 1997;29:720–728. [PubMed]
23. Wright JT, Jr, Agodoa L, Contreras G, et al. Successful blood pressure control in the African American Study of Kidney Disease and Hypertension. Arch Intern Med. 2002;162:1636–1643. [PubMed]
24. Salem RM, Fung MM, Pandey B, et al. Electronic Medical Records in the Post-Genomic Era: The VA Hypertension Primary Care Longitudinal Cohort. Genetic Epidemiology. . Forthcoming 2009. [PMC free article] [PubMed]
25. Brown SH, Lincoln MJ, Groen PJ, et al. VistA—U.S. Department of Veterans Affairs National-Scale HIS. Int J Med Inform. 2003;69:135–156. [PubMed]
26. Levey AS, Bosch JP, Lewis JB, et al. Modification of Diet in Renal Disease Study Group. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med. 1999;130:461–470. [PubMed]
27. Levey AS, Stevens LA, Coresh J. Conceptual model of CKD: applications and implications. Am J Kidney Dis. 2009;53(Suppl 3):S4–S16. [PubMed]
28. Mehta RL, Kellum JA, Shah SV, et al. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care. 2007;11:R31. [PMC free article] [PubMed]
29. Cheng S, Grow MA, Pallaud C, et al. A multilocus genotyping assay for candidate markers of cardiovascular disease risk. Genome Res. 1999;9:936–949. [PubMed]
30. Hill W. Estimation of linkage disequilibrium in randomly mating populations. Heredity. 1974;33:229–239. [PubMed]
31. McKeigue PM, Carpenter JR, Parra EJ, et al. Estimation of admixture and detection of linkage in admixed populations by a Bayesian approach: application to African-American populations. Ann Hum Genet. 2000;64(Pt 2):171–186. [PubMed]
32. Nievergelt CM, Libiger O, Schork NJ. Generalized analysis of molecular variance. PLoS Genet. 2007;3:e51. [PubMed]
33. Littell RC, Milliken GA, Stroup WW, et al. 2nd edn. Cary, NC: SAS Institute; 2006. SAS for Mixed Models.
34. Salem RM, Pandey B, Richard E, et al. Analysis of longitudinal clinical information in genetics studies: structured and unstructured measurement patterns. J Health Services Res Policy. . Forthcoming 2009.
35. Wolfinger RD. An example of using mixed models and PROC MIXED for longitudinal data. J Biopharm Stat. 1997;7:481–500. [PubMed]
36. Littell RC, Pendergast J, Natarajan R. Modelling covariance structure in the analysis of repeated measures data. Stat Med. 2000;19:1793–1819. [PubMed]
37. Holden JE, Kelley K, Agarwal R. Analyzing change: a primer on multilevel models with applications to nephrology. Am J Nephrol. 2008;28:792–801. [PMC free article] [PubMed]
38. Denison DGT, Mallick BK, Smith AFM. Bayesian MARS. Stat Comput. 1998;V8:337–346.
39. Henderson R, Diggle P, Dobson A. Joint modelling of longitudinal measurements and event time data. Biostatistics. 2000;1:465–480. [PubMed]
40. Faul F, Erdfelder E, Lang AG, et al. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39:175–191. [PubMed]
41. Halperin E, Eskin E. Haplotype reconstruction from genotype data using Imperfect Phylogeny. Bioinformatics. 2004;20:1842–1849. [PubMed]
42. Kanayama N, Tsujimura R, She L, et al. Cold-induced stress stimulates the sympathetic nervous system, causing hypertension and proteinuria in rats. J Hypertens. 1997;15:383–389. [PubMed]
43. King SE, Baldwin DS. Production of renal ischemia and proteinuria in man by the adrenal medullary hormones. Am J Med. 1956;20:217–224. [PubMed]
44. Light KC, Obrist PA, Sherwood A, et al. Effects of race and marginally elevated blood pressure on responses to stress. Hypertension. 1987;10:555–563. [PubMed]
45. Insel PA, Snavely MD. Catecholamines and the kidney: receptors and renal function. Annu Rev Physiol. 1981;43:625–636. [PubMed]
46. Muhlbauer B, Spohr F, Schmidt R, et al. Role of renal nerves and endogenous dopamine in amino acid-induced glomerular hyperfiltration. Am J Physiol. 1997;273(Pt 2):F144–F149. [PubMed]
47. O’Connor DT, Tyrell EA, Kailasam MT, et al. Early alteration in glomerular reserve in humans at genetic risk of essential hypertension: mechanisms and consequences. Hypertension. 2001;37:898–906. [PubMed]
48. Boivin V, Jahns R, Gambaryan S, et al. Immunofluorescent imaging of beta 1- and beta 2-adrenergic receptors in rat kidney. Kidney Int. 2001;59:515–531. [PubMed]
49. Anderson S, Brenner BM. Intraglomerular hypertension: implications and drug treatment. Annu Rev Med. 1988;39:243–253. [PubMed]
50. Green SA, Turki J, Innis M, et al. Amino-terminal polymorphisms of the human beta 2-adrenergic receptor impart distinct agonist-promoted regulatory properties. Biochemistry. 1994;33:9414–9419. [PubMed]
51. Sandilands A, Yeo G, Brown MJ, et al. Functional responses of human beta-1 adrenoceptors with defined haplotypes for the common 389R>G and 49S>G polymorphisms. Pharmacogenetics. 2004;14:343–349. [PubMed]
52. Rathz DA, Brown KM, Kramer LA, et al. Amino acid 49 polymorphisms of the human beta1-adrenergic receptor affect agonist-promoted trafficking. J Cardiovasc Pharmacol. 2002;39:155–160. [PubMed]
53. Hausdorff WP, Caron MG, Lefkowitz RJ. Turning off the signal: desensitization of beta-adrenergic receptor function. FASEB J. 1990;4:2881–2889. [PubMed]
54. Levin MC, Marullo S, Muntaner O, et al. The myocardium-protective Gly-49 variant of the beta 1-adrenergic receptor exhibits constitutive activity and increased desensitization and down-regulation. J Biol Chem. 2002;277:30429–30435. [PubMed]
55. Terra SG, Hamilton KK, Pauly DF, et al. Beta1-adrenergic receptor polymorphisms and left ventricular remodeling changes in response to beta-blocker therapy. Pharmacogenet Genomics. 2005;15:227–234. [PubMed]
56. Magnusson Y, Levin MC, Eggertsen R, et al. Ser49Gly of beta1-adrenergic receptor is associated with effective beta-blocker dose in dilated cardiomyopathy. Clin Pharmacol Ther. 2005;78:221–231. [PubMed]
57. Sehnert AJ, Daniels SE, Elashoff M, et al. Lack of association between adrenergic receptor genotypes and survival in heart failure patients treated with carvedilol or metoprolol. J Am Coll Cardiol. 2008;52:644–651. [PubMed]
58. Ranade K, Jorgenson E, Sheu WH, et al. A polymorphism in the beta1 adrenergic receptor is associated with resting heart rate. Am J Hum Genet. 2002;70:935–942. [PubMed]
59. Bengtsson K, Melander O, Orho-Melander M, et al. Polymorphism in the beta(1)-adrenergic receptor gene and hypertension. Circulation. 2001;104:187–190. [PubMed]
60. Liu J, Liu ZQ, Yu BN, et al. beta1-Adrenergic receptor polymorphisms influence the response to metoprolol monotherapy in patients with essential hypertension. Clin Pharmacol Ther. 2006;80:23–32. [PubMed]
61. DiBona GF, Kopp UC. Neural control of renal function. Physiol Rev. 1997;77:75–197. [PubMed]
62. Kim SM, Chen L, Faulhaber-Walter R, et al. Regulation of renin secretion and expression in mice deficient in beta1- and beta2-adrenergic receptors. Hypertension. 2007;50:103–109. [PubMed]
63. Peterson JC, Adler S, Burkart JM, et al. The Modification of Diet in Renal Disease Study. Blood pressure control, proteinuria, and the progression of renal disease. Ann Intern Med. 1995;123:754–762. [PubMed]
64. Concato J, Shah N, Horwitz RI. Randomized, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med. 2000;342:1887–1892. [PMC free article] [PubMed]
65. Kunz R. Randomized trials and observational studies: still mostly similar results, still crucial differences. J Clin Epidemiol. 2008;61:207–208. [PubMed]

Articles from Nephrology Dialysis Transplantation are provided here courtesy of Oxford University Press