Original Investigation A Genetic Marker of Uric Acid Level, Carotid Atherosclerosis, and Arterial Stiffness: A Family-Based Study Francesca Mallamaci, MD,1,2 Alessandra Testa, DrBiol, PhD,1 Daniela Leonardis, MD,1 Rocco Tripepi,1 Anna Pisano, DrBiol,1 Belinda Spoto, DrBiol,1 Maria Cristina Sanguedolce, DrBiol,1 Rosa Maria Parlongo,1 Giovanni Tripepi, DrBiostat, PhD,1 and Carmine Zoccali, MD1,2 Background: Hyperuricemia associates with atherosclerosis complications, but it is uncertain whether this relationship is causal in nature. The urate transporter GLUT9 (encoded by the SLC2A9 gene) is a major genetic determinant of serum uric acid level in humans. Because polymorphisms are distributed randomly at mating (Mendelian randomization), studies based on GLUT9 polymorphisms may provide unconfounded assessment of the nature of the link between uric acid and atherosclerosis. Study Design: Cross-sectional study. Setting & Participants: Family-based study including 449 individuals in 107 families in a genetically homogeneous population in Southern Italy. Factor: Serum uric acid level, rs734553 allele, and age. Outcome: Ultrasound biomarkers of atherosclerosis (intima-media thickness [IMT] and internal diameter) and pulse wave velocity (PWV). Results: Serum uric acid level was dose-dependently associated with the T allele of rs734553, a polymorphism in SLC2A9 (P 5 8 3 10-6). Serum uric acid level was a strong modifier of the relationship between age and IMT in fully adjusted analyses (b 5 0.33; P 5 0.01), whereas no such relationship was found for internal diameter (b 5 20.15; P 5 0.3) or PWV (b 5 0.10; P 5 0.6). The T allele coherently associated with carotid IMT, internal diameter, and PWV and emerged as an even stronger modifier of the age-IMT and age–internal diameter relationships in both crude and fully adjusted (b 5 0.40 [P , 0.001] and b 5 0.48 [P 5 0.003], respectively) analyses. Limitations: This is a hypothesis-generating study. Conclusions: Results in this family-based study implicate uric acid as an important modifier of the agedependent risk for atherosclerosis. Trials testing uric acid–lowering interventions are needed to prove this hypothesis. Am J Kidney Dis. -(-):---. ª 2014 by the National Kidney Foundation, Inc. INDEX WORDS: Atherosclerosis; glucose transporter type 9 (GLUT9); hyperuricemia; SLC2A9; rs734553; uric acid; intima-media thickness (IMT); pulse-wave velocity (PWV); cardiovascular disease.
H
yperuricemia has long been suspected as a risk factor for atherosclerosis. The link between uric acid level and vascular health has been investigated intensively over the past 3 decades, but the question remains largely unsettled (reviewed in1). Although the biological interference of uric acid with oxidative stress and vascular integrity is complex and in many respects controversial,2 plausible biological mechanisms whereby hyperuricemia may cause atherosclerosis have been described.3 Case-control and cohort studies investigating the relationship between uric acid and cardiovascular disease have produced conflicting results, and it is the prevailing view that the hyperuricemia-atherosclerosis association described in these studies most likely depends on the confounding effect of comorbid conditions that go along with hyperuricemia, including hypertension, diabetes, dyslipidemia, and obesity.4 Whether high uric acid levels may cause atherosclerosis is a question with major clinical and public health implications. In the most recent National Health Am J Kidney Dis. 2014;-(-):---
and Nutrition Examination Survey, the prevalence of hyperuricemia in the general population was 21%, with trends over time suggesting that levels of this biomarker are continuing to increase.5 Variability in uric acid levels attributable to environmental factors is a possible explanation for why some observational studies have not succeeded in observing a link between uric acid level and cardiovascular events.4 Environmental factors such as hydration status, acid-base From the 1CNR-IFC/IBIM, Reggio Calabria; and 2Unità Operativa di Nefrologia, Ipertensione e Trapianto Renale Ospedali Riuniti, Reggio Cal, Italy. Received January 28, 2014. Accepted in revised form July 28, 2014. Address correspondence to Carmine Zoccali, CNR-IFC/IBIM, Reggio Calabria e Unità Operativa di Nefrologia, Ipertensione e Trapianto Renale Ospedali Riuniti, 89124 Reggio Cal, Italy. E-mail: [email protected] 2014 by the National Kidney Foundation, Inc. 0272-6386/$36.00 http://dx.doi.org/10.1053/j.ajkd.2014.07.021 1
Mallamaci et al
balance, nutrient intake, and use of diuretics and other drugs may substantially affect uric acid levels. With genes being transmitted randomly (Mendelian randomization),6 polymorphisms in genes that regulate serum uric acid concentration may represent an unbiased approach to further explore the link between uric acid and atherosclerosis. However, large-scale case-control studies across various populations and ethnicities have failed to show a link between uric acid–regulating genes and cardiovascular events.7-9 One potential reason for this failure is that, due to population admixture and population stratification, these studies tend to dilute any underlying link between a given gene and clinical outcomes. By contrast, studies based on families in homogeneous populations might yield key information about genetic associations.10 Although more difficult to perform, family-based studies have the advantage of controlling by design the issue of population stratification and minimizing the risk of false-positive and -negative genetic associations deriving from population admixture.11 Studies based on families in homogeneous populations also are advantageous because these communities are more homogeneous for exposure to environmental factors.10 With this background in mind and pending definitive proof (ie, a proper clinical trial) that the uric acid– atherosclerosis link is causal in nature, we investigated the relationship between a genetic marker of uric acid level and phenotypic markers of atherosclerosis in a family-based study. In particular, we looked at relevant intermediate phenotypes of the atherosclerosis process, such as intima-media thickness (IMT) or internal diameter in carotid arteries, or biomarkers of arterial rigidity, such as pulse wave velocity (PWV), which to our knowledge have not been studied for association with uric acid level. As the genetic marker of uric acid level, we used the single-nucleotide polymorphism (SNP) rs734553, which is located in an intron of SLC2A9, the gene encoding the urate transporter GLUT9. This SNP was reported to be the strongest genetic marker of uric acid level in a meta-analysis including 28,000 individuals.12 We conducted our study in a community with a shared genetic background13 and shared nutritional habits.14 Because previous observations in the ARIC (Atherosclerosis Risk in the Communities) Study showed that age not only is associated positively with carotid atherosclerosis, but also has a larger effect in individuals with stronger and/or longer exposures to risk factors for atherosclerosis,15 we tested the effect modification by rs734553 allele and uric acid level on the relationship between age and severity of carotid atherosclerosis. Of note, in our study, the age interaction was present for both serum uric acid level and the rs734553 allele. 2
METHODS Study Protocol The protocol conformed with the ethical guidelines of our institution, and informed consent was obtained from each participant.
Study Population We enrolled 449 individuals who participated in a health screening offered to 107 families. This screening was undertaken as a part of a large project promoted and supported by the Italian Society of Hypertension to investigate genetic factors of human hypertension.16 These families were part of a Southern Italy population, that is, a population with a shared genetic background13 and a peculiar Mediterranean nutrient intake characterized by relatively high intakes of vitamin E and monounsaturated fatty acids.14 The diagnosis of primary hypertension in these families was based on the diagnostic protocol routinely applied at our hypertension center, which includes routine biochemical tests; measurement of 24-hour catecholamines, supine and erect plasma renin activity, and plasma aldosterone; captopril renography and/ or digital subtraction angiography; and echo-color Doppler of renal arteries.
Blood Pressure Measurements Blood pressure (BP) was measured in a quiet well-lit room kept at a constant temperature. When a participant arrived at the research center, the individual rested in a semirecumbent position for 20 to 30 minutes in a comfortable armchair. BP then was measured 3 times by an automatic device with an interval of about 2 minutes between measurements, and the average of these 3 measurements was registered as the study BP.
Carotid Ultrasonography and PWV Studies In all patients, ultrasonographic studies of the common carotid arteries were performed with a Hewlett Packard Sonos 1500 (7.5-MHz high-resolution probe) by a single observer (R.T.). IMT, internal diameter of the carotid arteries, and plaque were assessed according to the protocol validated and systematically applied at our research unit, as described in detail in a previous study.17
Pulse Wave Analysis: Carotid-Radial The assessment of arterial wave reflection characteristics was performed noninvasively using the SphygmoCor Pulse Wave Analysis Px system and SCOR-2000, version 6.31, software (both AtCor Medical). Carotid-radial PWV was measured in triplicate from the left common carotid pulse to the left radial pulse using applanation tonometry, as described elsewhere.18 Carotid-radial PWV was estimated by dividing central transit distance by transit time (Dt), using the SphygmoCor Pulse Wave Velocity Vx system and SCOR-2000, version 6.31, software.
Laboratory Measurements Fasting blood sampling was undertaken and plasma was stored at 280 C until analysis. Serum uric acid, creatinine, lipids, and hemoglobin were measured by standard methods in the routine clinical laboratory. Estimated glomerular filtration rate (eGFR) was calculated using the 4-variable MDRD (Modification of Diet in Renal Disease) Study equation,19 and microalbuminuria, by 24-hour urine collection.
Genotyping of GLUT9 Gene Polymorphism Genomic DNA was extracted from peripheral-blood leukocytes by the standard salting-out technique. rs734553 genotype was determined with a validated TaqMan SNP Genotyping Assay, performed on an ABI PRISM 7900HT according to the Am J Kidney Dis. 2014;-(-):---
A Genetic Marker of Uric Acid and Atherosclerosis manufacturer’s recommendations (Life Technologies). The assay mix (including unlabeled polymerase chain reaction [PCR] primers, FAM [6-carboxyfluorescein] and VIC [4,7,20 -trichloro-70 phenyl-6-carboxyfluorescein] dye-labeled TaqMan MGB [minor groove binder] probes) was designed by Life Technologies. The reaction mix contained 1 to 5 ng of genomic DNA, 12.5 mL of TaqMan Universal PCR Master Mix, No AmpErase UNG (uracil N-glycosylase), and 1.25 mL of assay mix and was adjusted with water to a total volume of 25 mL. Alleles were scored using the allelic discrimination software Sequence Detection System, version 2.2 (Life Technologies). A random 5% of samples were repeated independently to confirm genotyping results. Genotype results for these samples were completely consistent.
Statistical Analysis Normally distributed data are summarized as mean 6 standard deviation, and non-normally distributed data, as median and interquartile range (IQR). Comparisons among groups were made by P for linear trend. The association between continuous variables was analyzed by the Pearson product moment correlation coefficient (r), Spearman rank correlation (r), and P value, as appropriate. As clinical outcomes in this study, we considered the indicators of carotid atherosclerosis, namely IMT, internal diameter of the carotid arteries, total number of plaques, and PWV. The key exposures were uric acid, rs734553 polymorphism, and age. As covariates, we considered the full list of variables listed in Table 1. The potential influence of family inter-relationships on the study outcomes (IMT and internal diameter of carotid arteries) was investigated preliminarily by calculating the intracluster correlation coefficient.20 In correlation analyses, shared variance between
variables considered simultaneously was calculated (as r2). Genotypic distribution was investigated by testing the HardyWeinberg equilibrium. Because serum uric acid and the rs734553 polymorphism were strong modifiers of the relationship between age, IMT, and the internal diameter of the carotid arteries, in multiple linear regression models, we always included age * uric acid level or age * rs734553 polymorphism interaction terms, as well as a series of potential confounders (sex, systolic BP, diabetes, waist circumference, low-density lipoprotein cholesterol level, triglyceride level, eGFR, diuretic use, and alcohol habit). We specifically looked at the effect modification by age because observations in the ARIC Study showed that age not only is associated with carotid atherosclerosis, but also has a larger effect on the same parameter in individuals with stronger and/or longer exposures to risk factors for this disease.15 An extensive effect modification analysis was carried out for investigating the interaction between uric acid levels and rs734553 polymorphisms with all variables listed in Table 1 for explaining the severity of carotid atherosclerosis in the study sample, and no significant effects were found. We had at least 37 patients for each variable in the model, emphasizing the stability of our analyses. In these models, variables were introduced in their original form and the normality assumption was tested by residuals plotting. Data were expressed as standardized regression coefficients (b) and P values. P # 0.05 was considered statistically significant. The estimated increase in IMT and the internal diameter of carotid arteries at predefined values of uric acid and rs734553 polymorphisms was expressed as unstandardized regression coefficients, 95% confidence interval, and P value. All calculations were done using a standard statistical package (SPSS for Windows, version 9.0.1; 1999).
Table 1. Clinical and Biochemical Data in the Study Population by rs734553 Genotype
Age (y) Male sex
GG (n 5 34)
GT (n 5 198)
TT (n 5 217)
P for Linear Trend
45 6 18
39 6 16
44 6 16
0.07
38%
51%
43%
0.5 0.06
Diabetes
3%
4%
9%
Smoker
24%
24%
20%
0.5
Alcohol usera
12%
22%
23%
0.3
Weight (kg)
77 6 22
73 6 15
74 6 15
0.6
BMI (kg/m2)
27.5 6 6.4
27.2 6 5.0
28.0 6 5.2
0.2
89 6 14 39%
89 6 14 25%
91 6 14 39%
0.09 0.07
Waist circumference (cm) Receiving antihypertensive treatment Use of diuretics Office systolic BP (mm Hg) Office diastolic BP (mm Hg) CRP (mg/L)
18%
3%
9%
0.9
122 6 19
125 6 18
129 6 20b
0.02
78 6 13
82 6 11
82 6 11
0.3
3.2 [1.6-3.8]
2.0 [0.9-3.3]
2.2 [1.2-3.3]
0.7
92 6 18
94 6 24
97 6 24
0.1
78 [48-104]
82 [56-115]
84 [60-114]
0.3
HDL cholesterol (mg/dL) LDL cholesterol (mg/dL)
48 6 15 102 6 33
48 6 14 106 6 33
47 6 12 112 6 32
0.4 0.06
Uric acid (mg/dL)
4.2 6 1.6
4.7 6 1.3
5.2 6 1.4
,0.001
Glucose (mg/dL) Triglycerides (mg/dL)
Note: Unless otherwise indicated, values for categorical variables are given as percentage; values for continuous variables, as mean 6 standard deviation or median [interquartile range]. Conversion factors for units: glucose in mg/dL to mmol/L, 30.05551; cholesterol in mg/dL to mmol/L, 30.02586; triglycerides in mg/dL to mmol/L, 30.01129; uric acid in mg/dL to mmol/L, 359.48. Abbreviations: BMI, body mass index; BP, blood pressure; CRP, C-reactive protein; HDL, high-density lipoprotein; LDL, low-density lipoprotein. a Median ethanol intake (% and range) for GG, GT, and TT: 30 (12-100), 22 (12-58), and 30 (12-50) mL/d, respectively. b P 5 0.02. Am J Kidney Dis. 2014;-(-):---
3
Mallamaci et al
RESULTS
Apart from systolic BP and as expected by Mendelian randomization, there were no significant differences among the 3 genotypes in terms of demographic and cardiovascular risk factors, although associations of number of T alleles with diabetes (P 5 0.06), serum glucose level (P 5 0.1), and low-density lipoprotein cholesterol level (P 5 0.06) were near the cutoff for statistical significance (Table 1). Age had an erratic trend across genotypes (Table 1). GFR (95 6 24, 102 6 21, and 94 6 19 mL/min/1.73 m2, respectively, for GG, GT, and TT genotypes) and albuminuria (medians of 11.9 [IQR, 9.3-18.6], 9.4 [IQR, 5.7-15.4], and 9.0 [IQR, 5.9-15.6] mg/24 h, respectively, for GG, GT, and TT genotypes) did not differ significantly between GG and TT homozygotes (P 5 0.8). Alcohol users (P 5 0.3) and the proportion of patients using diuretics (P 5 0.9) did not differ across genotypes (Table 1).
Clinical Characteristics of the Study Population Of 449 individuals in 107 families, 197 were hypertensive (43%; BP, 140 6 17/89 6 10 mm Hg) and the other 254 (57%) were normotensive (BP, 116 6 12/ 75 6 8 mm Hg). One-hundred forty-five (32%) patients were treated with various antihypertensive drugs, including angiotensin-converting enzyme inhibitors or angiotensin II blockers (n 5 89 [20%]), calcium antagonists (n 5 62 [14%]), sympathetic blockers (n 5 44 [10%]), thiazides (n 5 32 [7%]), and vasodilators (n 5 4 [1%]). No individual had frank proteinuria; average 24-hour albumin excretion was 19 mg/24 h, and most (86%) had albumin excretion , 20 mg/24 h. Prevalence rates of diabetes and smoking were 6% and 22%, respectively. Association of rs734553 Genotype With Uric Acid Level and Demographic and Clinical Variables
Associations of rs734553 Genotype With Uric Acid Levels and Biomarkers of Atherosclerosis
In the entire study population, rs734553 alleles were in Hardy-Weinberg equilibrium (c2 5 3.0; P 5 0.2), and this also was true in a separate analysis in hypertensive and normotensive individuals (c2 range, 0.14-3.0; P 5 0.2 and P 5 0.9, respectively). Serum uric acid levels cosegregated with the T allele (risk allele) of rs734553 (TT, GT, and GG genotypes had uric acid concentrations of 5.2 6 1.4, 4.7 6 1.3, and 4.2 6 1.6 mg/dL, respectively; P , 0.001), confirming in our population that this polymorphism is associated strongly with prevailing uric acid levels. Patients having the TT genotype showed higher office systolic BP (P 5 0.02) compared with those with GT and GG genotypes (Table 1). 1.4
IMT (mm)
Internal diameter (mm)
9.0 r=0.34 P
H
yperuricemia has long been suspected as a risk factor for atherosclerosis. The link between uric acid level and vascular health has been investigated intensively over the past 3 decades, but the question remains largely unsettled (reviewed in1). Although the biological interference of uric acid with oxidative stress and vascular integrity is complex and in many respects controversial,2 plausible biological mechanisms whereby hyperuricemia may cause atherosclerosis have been described.3 Case-control and cohort studies investigating the relationship between uric acid and cardiovascular disease have produced conflicting results, and it is the prevailing view that the hyperuricemia-atherosclerosis association described in these studies most likely depends on the confounding effect of comorbid conditions that go along with hyperuricemia, including hypertension, diabetes, dyslipidemia, and obesity.4 Whether high uric acid levels may cause atherosclerosis is a question with major clinical and public health implications. In the most recent National Health Am J Kidney Dis. 2014;-(-):---
and Nutrition Examination Survey, the prevalence of hyperuricemia in the general population was 21%, with trends over time suggesting that levels of this biomarker are continuing to increase.5 Variability in uric acid levels attributable to environmental factors is a possible explanation for why some observational studies have not succeeded in observing a link between uric acid level and cardiovascular events.4 Environmental factors such as hydration status, acid-base From the 1CNR-IFC/IBIM, Reggio Calabria; and 2Unità Operativa di Nefrologia, Ipertensione e Trapianto Renale Ospedali Riuniti, Reggio Cal, Italy. Received January 28, 2014. Accepted in revised form July 28, 2014. Address correspondence to Carmine Zoccali, CNR-IFC/IBIM, Reggio Calabria e Unità Operativa di Nefrologia, Ipertensione e Trapianto Renale Ospedali Riuniti, 89124 Reggio Cal, Italy. E-mail: [email protected] 2014 by the National Kidney Foundation, Inc. 0272-6386/$36.00 http://dx.doi.org/10.1053/j.ajkd.2014.07.021 1
Mallamaci et al
balance, nutrient intake, and use of diuretics and other drugs may substantially affect uric acid levels. With genes being transmitted randomly (Mendelian randomization),6 polymorphisms in genes that regulate serum uric acid concentration may represent an unbiased approach to further explore the link between uric acid and atherosclerosis. However, large-scale case-control studies across various populations and ethnicities have failed to show a link between uric acid–regulating genes and cardiovascular events.7-9 One potential reason for this failure is that, due to population admixture and population stratification, these studies tend to dilute any underlying link between a given gene and clinical outcomes. By contrast, studies based on families in homogeneous populations might yield key information about genetic associations.10 Although more difficult to perform, family-based studies have the advantage of controlling by design the issue of population stratification and minimizing the risk of false-positive and -negative genetic associations deriving from population admixture.11 Studies based on families in homogeneous populations also are advantageous because these communities are more homogeneous for exposure to environmental factors.10 With this background in mind and pending definitive proof (ie, a proper clinical trial) that the uric acid– atherosclerosis link is causal in nature, we investigated the relationship between a genetic marker of uric acid level and phenotypic markers of atherosclerosis in a family-based study. In particular, we looked at relevant intermediate phenotypes of the atherosclerosis process, such as intima-media thickness (IMT) or internal diameter in carotid arteries, or biomarkers of arterial rigidity, such as pulse wave velocity (PWV), which to our knowledge have not been studied for association with uric acid level. As the genetic marker of uric acid level, we used the single-nucleotide polymorphism (SNP) rs734553, which is located in an intron of SLC2A9, the gene encoding the urate transporter GLUT9. This SNP was reported to be the strongest genetic marker of uric acid level in a meta-analysis including 28,000 individuals.12 We conducted our study in a community with a shared genetic background13 and shared nutritional habits.14 Because previous observations in the ARIC (Atherosclerosis Risk in the Communities) Study showed that age not only is associated positively with carotid atherosclerosis, but also has a larger effect in individuals with stronger and/or longer exposures to risk factors for atherosclerosis,15 we tested the effect modification by rs734553 allele and uric acid level on the relationship between age and severity of carotid atherosclerosis. Of note, in our study, the age interaction was present for both serum uric acid level and the rs734553 allele. 2
METHODS Study Protocol The protocol conformed with the ethical guidelines of our institution, and informed consent was obtained from each participant.
Study Population We enrolled 449 individuals who participated in a health screening offered to 107 families. This screening was undertaken as a part of a large project promoted and supported by the Italian Society of Hypertension to investigate genetic factors of human hypertension.16 These families were part of a Southern Italy population, that is, a population with a shared genetic background13 and a peculiar Mediterranean nutrient intake characterized by relatively high intakes of vitamin E and monounsaturated fatty acids.14 The diagnosis of primary hypertension in these families was based on the diagnostic protocol routinely applied at our hypertension center, which includes routine biochemical tests; measurement of 24-hour catecholamines, supine and erect plasma renin activity, and plasma aldosterone; captopril renography and/ or digital subtraction angiography; and echo-color Doppler of renal arteries.
Blood Pressure Measurements Blood pressure (BP) was measured in a quiet well-lit room kept at a constant temperature. When a participant arrived at the research center, the individual rested in a semirecumbent position for 20 to 30 minutes in a comfortable armchair. BP then was measured 3 times by an automatic device with an interval of about 2 minutes between measurements, and the average of these 3 measurements was registered as the study BP.
Carotid Ultrasonography and PWV Studies In all patients, ultrasonographic studies of the common carotid arteries were performed with a Hewlett Packard Sonos 1500 (7.5-MHz high-resolution probe) by a single observer (R.T.). IMT, internal diameter of the carotid arteries, and plaque were assessed according to the protocol validated and systematically applied at our research unit, as described in detail in a previous study.17
Pulse Wave Analysis: Carotid-Radial The assessment of arterial wave reflection characteristics was performed noninvasively using the SphygmoCor Pulse Wave Analysis Px system and SCOR-2000, version 6.31, software (both AtCor Medical). Carotid-radial PWV was measured in triplicate from the left common carotid pulse to the left radial pulse using applanation tonometry, as described elsewhere.18 Carotid-radial PWV was estimated by dividing central transit distance by transit time (Dt), using the SphygmoCor Pulse Wave Velocity Vx system and SCOR-2000, version 6.31, software.
Laboratory Measurements Fasting blood sampling was undertaken and plasma was stored at 280 C until analysis. Serum uric acid, creatinine, lipids, and hemoglobin were measured by standard methods in the routine clinical laboratory. Estimated glomerular filtration rate (eGFR) was calculated using the 4-variable MDRD (Modification of Diet in Renal Disease) Study equation,19 and microalbuminuria, by 24-hour urine collection.
Genotyping of GLUT9 Gene Polymorphism Genomic DNA was extracted from peripheral-blood leukocytes by the standard salting-out technique. rs734553 genotype was determined with a validated TaqMan SNP Genotyping Assay, performed on an ABI PRISM 7900HT according to the Am J Kidney Dis. 2014;-(-):---
A Genetic Marker of Uric Acid and Atherosclerosis manufacturer’s recommendations (Life Technologies). The assay mix (including unlabeled polymerase chain reaction [PCR] primers, FAM [6-carboxyfluorescein] and VIC [4,7,20 -trichloro-70 phenyl-6-carboxyfluorescein] dye-labeled TaqMan MGB [minor groove binder] probes) was designed by Life Technologies. The reaction mix contained 1 to 5 ng of genomic DNA, 12.5 mL of TaqMan Universal PCR Master Mix, No AmpErase UNG (uracil N-glycosylase), and 1.25 mL of assay mix and was adjusted with water to a total volume of 25 mL. Alleles were scored using the allelic discrimination software Sequence Detection System, version 2.2 (Life Technologies). A random 5% of samples were repeated independently to confirm genotyping results. Genotype results for these samples were completely consistent.
Statistical Analysis Normally distributed data are summarized as mean 6 standard deviation, and non-normally distributed data, as median and interquartile range (IQR). Comparisons among groups were made by P for linear trend. The association between continuous variables was analyzed by the Pearson product moment correlation coefficient (r), Spearman rank correlation (r), and P value, as appropriate. As clinical outcomes in this study, we considered the indicators of carotid atherosclerosis, namely IMT, internal diameter of the carotid arteries, total number of plaques, and PWV. The key exposures were uric acid, rs734553 polymorphism, and age. As covariates, we considered the full list of variables listed in Table 1. The potential influence of family inter-relationships on the study outcomes (IMT and internal diameter of carotid arteries) was investigated preliminarily by calculating the intracluster correlation coefficient.20 In correlation analyses, shared variance between
variables considered simultaneously was calculated (as r2). Genotypic distribution was investigated by testing the HardyWeinberg equilibrium. Because serum uric acid and the rs734553 polymorphism were strong modifiers of the relationship between age, IMT, and the internal diameter of the carotid arteries, in multiple linear regression models, we always included age * uric acid level or age * rs734553 polymorphism interaction terms, as well as a series of potential confounders (sex, systolic BP, diabetes, waist circumference, low-density lipoprotein cholesterol level, triglyceride level, eGFR, diuretic use, and alcohol habit). We specifically looked at the effect modification by age because observations in the ARIC Study showed that age not only is associated with carotid atherosclerosis, but also has a larger effect on the same parameter in individuals with stronger and/or longer exposures to risk factors for this disease.15 An extensive effect modification analysis was carried out for investigating the interaction between uric acid levels and rs734553 polymorphisms with all variables listed in Table 1 for explaining the severity of carotid atherosclerosis in the study sample, and no significant effects were found. We had at least 37 patients for each variable in the model, emphasizing the stability of our analyses. In these models, variables were introduced in their original form and the normality assumption was tested by residuals plotting. Data were expressed as standardized regression coefficients (b) and P values. P # 0.05 was considered statistically significant. The estimated increase in IMT and the internal diameter of carotid arteries at predefined values of uric acid and rs734553 polymorphisms was expressed as unstandardized regression coefficients, 95% confidence interval, and P value. All calculations were done using a standard statistical package (SPSS for Windows, version 9.0.1; 1999).
Table 1. Clinical and Biochemical Data in the Study Population by rs734553 Genotype
Age (y) Male sex
GG (n 5 34)
GT (n 5 198)
TT (n 5 217)
P for Linear Trend
45 6 18
39 6 16
44 6 16
0.07
38%
51%
43%
0.5 0.06
Diabetes
3%
4%
9%
Smoker
24%
24%
20%
0.5
Alcohol usera
12%
22%
23%
0.3
Weight (kg)
77 6 22
73 6 15
74 6 15
0.6
BMI (kg/m2)
27.5 6 6.4
27.2 6 5.0
28.0 6 5.2
0.2
89 6 14 39%
89 6 14 25%
91 6 14 39%
0.09 0.07
Waist circumference (cm) Receiving antihypertensive treatment Use of diuretics Office systolic BP (mm Hg) Office diastolic BP (mm Hg) CRP (mg/L)
18%
3%
9%
0.9
122 6 19
125 6 18
129 6 20b
0.02
78 6 13
82 6 11
82 6 11
0.3
3.2 [1.6-3.8]
2.0 [0.9-3.3]
2.2 [1.2-3.3]
0.7
92 6 18
94 6 24
97 6 24
0.1
78 [48-104]
82 [56-115]
84 [60-114]
0.3
HDL cholesterol (mg/dL) LDL cholesterol (mg/dL)
48 6 15 102 6 33
48 6 14 106 6 33
47 6 12 112 6 32
0.4 0.06
Uric acid (mg/dL)
4.2 6 1.6
4.7 6 1.3
5.2 6 1.4
,0.001
Glucose (mg/dL) Triglycerides (mg/dL)
Note: Unless otherwise indicated, values for categorical variables are given as percentage; values for continuous variables, as mean 6 standard deviation or median [interquartile range]. Conversion factors for units: glucose in mg/dL to mmol/L, 30.05551; cholesterol in mg/dL to mmol/L, 30.02586; triglycerides in mg/dL to mmol/L, 30.01129; uric acid in mg/dL to mmol/L, 359.48. Abbreviations: BMI, body mass index; BP, blood pressure; CRP, C-reactive protein; HDL, high-density lipoprotein; LDL, low-density lipoprotein. a Median ethanol intake (% and range) for GG, GT, and TT: 30 (12-100), 22 (12-58), and 30 (12-50) mL/d, respectively. b P 5 0.02. Am J Kidney Dis. 2014;-(-):---
3
Mallamaci et al
RESULTS
Apart from systolic BP and as expected by Mendelian randomization, there were no significant differences among the 3 genotypes in terms of demographic and cardiovascular risk factors, although associations of number of T alleles with diabetes (P 5 0.06), serum glucose level (P 5 0.1), and low-density lipoprotein cholesterol level (P 5 0.06) were near the cutoff for statistical significance (Table 1). Age had an erratic trend across genotypes (Table 1). GFR (95 6 24, 102 6 21, and 94 6 19 mL/min/1.73 m2, respectively, for GG, GT, and TT genotypes) and albuminuria (medians of 11.9 [IQR, 9.3-18.6], 9.4 [IQR, 5.7-15.4], and 9.0 [IQR, 5.9-15.6] mg/24 h, respectively, for GG, GT, and TT genotypes) did not differ significantly between GG and TT homozygotes (P 5 0.8). Alcohol users (P 5 0.3) and the proportion of patients using diuretics (P 5 0.9) did not differ across genotypes (Table 1).
Clinical Characteristics of the Study Population Of 449 individuals in 107 families, 197 were hypertensive (43%; BP, 140 6 17/89 6 10 mm Hg) and the other 254 (57%) were normotensive (BP, 116 6 12/ 75 6 8 mm Hg). One-hundred forty-five (32%) patients were treated with various antihypertensive drugs, including angiotensin-converting enzyme inhibitors or angiotensin II blockers (n 5 89 [20%]), calcium antagonists (n 5 62 [14%]), sympathetic blockers (n 5 44 [10%]), thiazides (n 5 32 [7%]), and vasodilators (n 5 4 [1%]). No individual had frank proteinuria; average 24-hour albumin excretion was 19 mg/24 h, and most (86%) had albumin excretion , 20 mg/24 h. Prevalence rates of diabetes and smoking were 6% and 22%, respectively. Association of rs734553 Genotype With Uric Acid Level and Demographic and Clinical Variables
Associations of rs734553 Genotype With Uric Acid Levels and Biomarkers of Atherosclerosis
In the entire study population, rs734553 alleles were in Hardy-Weinberg equilibrium (c2 5 3.0; P 5 0.2), and this also was true in a separate analysis in hypertensive and normotensive individuals (c2 range, 0.14-3.0; P 5 0.2 and P 5 0.9, respectively). Serum uric acid levels cosegregated with the T allele (risk allele) of rs734553 (TT, GT, and GG genotypes had uric acid concentrations of 5.2 6 1.4, 4.7 6 1.3, and 4.2 6 1.6 mg/dL, respectively; P , 0.001), confirming in our population that this polymorphism is associated strongly with prevailing uric acid levels. Patients having the TT genotype showed higher office systolic BP (P 5 0.02) compared with those with GT and GG genotypes (Table 1). 1.4
IMT (mm)
Internal diameter (mm)
9.0 r=0.34 P
