DOI: 10.1111/eci.12331
ORIGINAL ARTICLE Uricaemia and left ventricular mass in hypertensive patients Cristiana Catena, GianLuca Colussi, Frine Capobianco, Gabriele Brosolo and Leonardo A. Sechi Hypertension Unit, Division of Internal Medicine, Department of Experimental and Clinical Medical Sciences, University of Udine, Udine, Italy
ABSTRACT Background Both hyperuricaemia and left ventricular (LV) hypertrophy are associated with the metabolic syndrome and increased cardiovascular risk. The relationship between uric acid levels and left ventricular mass in hypertension, however, is unclear. In this study, we have investigated this relationship in hypertensive patients without the metabolic syndrome. Materials and methods In a cross-sectional study, 367 nondiabetic, essential hypertensive patients (age 52 14; 194 males and 173 females) free of clinically relevant cardiovascular complications and without the metabolic syndrome were consecutively recruited at a university hypertension clinic. In these patients, we measured plasma levels of uric acid, lipids, glucose and insulin at fast and after an oral glucose load (OGTT), renal function and performed both conventional and tissue Doppler echocardiography. Results Hypertensive patients with LV hypertrophy had higher uric acid levels and greater prevalence of hyperuricemia than patients with normal left ventricular mass. Uric acid levels were directly related with fasting and post-OGTT plasma insulin and with the HOMA index and inversely with 24-h creatinine clearance. Uric acid was also significantly and directly related with the left ventricular mass and multivariate regression analysis showed that this relationship was independent from components of the metabolic syndrome and renal function in women, but not in men. Conclusions Elevated uric acid levels are independently related to the left ventricular mass in hypertensive women without the metabolic syndrome. In these patients with a low cardiovascular risk profile, uric acid might contribute to the development of subclinical cardiac damage. Keywords Hypertension, left ventricular hypertrophy, metabolic syndrome, tissue Doppler echocardiography, uric acid. Eur J Clin Invest 2014; 44 (10): 972–981
Introduction Left ventricular (LV) hypertrophy is an independent risk factor for cardiovascular morbidity and mortality in patients with hypertension and is associated with increased incidence of arrhythmia, myocardial infarction and stroke [1]. Increased blood pressure with the related increase in cardiac workload is not the only determinant of cardiac damage in hypertensive patients and the propensity of these patients to develop structural and functional abnormalities of the heart is under the influence of additional hormonal, haemorheologic and metabolic factors [2,3]. The metabolic syndrome with the related insulin resistance and hyperinsulinemia contributes to the cardiovascular risk and is associated with increased LV mass in the general population [3] and in patients with hypertension [4].
972
Hyperuricaemia is frequently detected in patients with the metabolic syndrome [5] and prospective studies have reported that elevated uric acid levels are associated with an increased risk of myocardial infarction and stroke [6] and with worse LV function in heart failure patients [7]. Studies conducted in the last decade have evaluated the relationships between uric acid levels and LV mass in hypertension reporting inconsistent findings [8–13], mostly because of interference of components of the metabolic syndrome. Moreover, different findings between men and women were reported in some studies [10,11]. Because of the important link between uric acid levels and the metabolic syndrome and its potential contribution to hypertensive cardiac damage, we sought to investigate the relationship between uric acid and cardiac structural and functional variables in a large group of hypertensive patients without this syndrome.
ª 2014 Stichting European Society for Clinical Investigation Journal Foundation
URIC ACID AND THE HEART
Materials and methods Patients Three hundred and sixty-seven patients with mild-to-moderate hypertension, who were consecutively recruited at our university clinic over a period of 8 years, were included in a crosssectional study. Blood pressure was measured by an automated device (Omron M6; OMRON Healthcare Co., Kyoto, Japan) using an appropriately sized cuff after each subject had been supine for 15 min, and the average of three readings was recorded [14]. Diagnosis of hypertension was based on at least 2 blood pressure measurements per visit in at least three visits or use of antihypertensive drugs, according to current guidelines [15]. All patients were white, lived in the north-east of Italy, and were representative of the hypertensive population in this area [16]. Patients younger than 18 years or older than 80 years and pregnant women were excluded, together with patients with history of alcohol abuse or acute illness, diabetes mellitus, metabolic syndrome, ischaemic heart, cardiac valve or other heart diseases, impaired systolic function (ejection fraction of < 50%), glomerular filtration rate (GFR) of < 60 mL/min/ 173 m2, use of diuretics, uric acid-lowering drugs or other drugs that could interfere with uric acid levels. Metabolic syndrome was diagnosed according to the American Heart Association criteria [17] when two or more of the following conditions were associated with hypertension: waist circumference of more than 102 cm in men or 88 cm in women; fasting plasma glucose of 100 mg/dL or more or use of hypoglycaemic drugs; triglycerides of 150 mg/dL or more or use of lipidlowering drugs; HDL cholesterol of < 40 mg/dL in men or 50 mg/dL in women. In all patients, secondary forms of hypertension were excluded on the basis of extensive investigations that included analysis of medical records, physical examination, urine analysis, blood biochemistries, 24-h creatinine clearance, plasma renin and aldosterone, urinary cortisol and catecholamines, ECG, echocardiography and renal ultrasound with measurement of renal resistance index. Renal angio-MRI/CT scan, and additional functional tests were performed when appropriate [18]. Two hundred and twenty (60%) of 367 patients had never been treated with antihypertensive drugs. The remaining 147 were treated with drugs (58% angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, 46% calcium channel blockers, 43% beta-blockers and 7% alpha-blockers) that were withdrawn for at least 2 weeks before the study. No patient was treated with losartan. Patients taking two or more antihypertensive agents were admitted to the hospital for the duration of the washout period and in those with blood pressure higher than 180/110 mm Hg (n = 15) alpha-blockers and/ or calcium channel blockers were given. Before the study,
patients ate a standard diet for 7 days to maintain a sodium intake of 100–150 mmol/day that was checked with measurement of 24-h urinary sodium excretion. Alcohol intake was estimated by a questionnaire [19]. Patients were classified as smokers if they had smoked for at least 5 years, and up to 1 year before the study. The study was approved by the local institutional review board and informed consent was obtained from all patients. Reporting of the study conforms to STROBE statement along with references to STROBE statement and the broader EQUATOR guidelines [20].
Laboratory measurements Venous blood was collected in the morning after an overnight fast, and plasma was frozen at -80°C until assay was performed, usually within 1 month from sampling. Uric acid levels were determined by the uricase–peroxidase system, and hyperuricaemia was defined by levels of more than 70 mg/dL in men and 60 mg/dL in women, according to the reference values of our laboratory [21]. Triglycerides and total and HDL cholesterol were assayed enzymatically and LDL cholesterol was calculated with the formula of Friedewald. Plasma glucose was measured using the glucose–oxidase method and plasma insulin by RIA. The Homoeostatic Model Assessment (HOMA) index was calculated as an index of insulin sensitivity from fasting glucose (mM) and insulin (lU/mL) using the formula: [(glucose 9 insulin)/225]. Glucose tolerance was evaluated by a 180-min oral glucose tolerance test (OGTT) and the area under the curve for glucose (AUC-G) and insulin (AUC-I) concentration during the OGTT was calculated by the trapezoidal rule [22]. GFR was measured by duplicate measurement of 24-h creatinine clearance and normalized for body surface area [23].
Echocardiography Echocardiography was performed as described previously by the same expert investigator who was unaware of the patients’ characteristics [24]. Measurements were obtained with a commercial machine (Aplio CV; Toshiba Medical System, Tokyo, Japan) and a 25 MHz transducer under bidimensional crosssectional control. LV mass index was calculated and adjusted for body height, with a cut-off value of 50 g/m27 for men and 47 g/m27 for women being used to define LV hypertrophy [25]. Concentric geometry was defined as relative wall thickness (RWT) ≥ 042 [25] and LV systolic function was estimated by the ejection fraction. Diastolic function was evaluated by pulsed Doppler that was recorded from apical four-chamber view at the level of mitral valve tips. Early- (E) and late-wave (A) diastolic velocities, their ratio (E/A) and the isovolumetric relaxation time (IVRT) were measured [26]. Tissue Doppler imaging (TDI) was also performed in a subset of 97 patients [22]. Early-diastolic and late-diastolic velocities of septal and lateral
European Journal of Clinical Investigation Vol 44
973
C. CATENA ET AL.
myocardial portions at the level of mitral valve annulus were measured, and the mean values (Em and Am) were calculated on 3 consecutive cycles together with the average Em/Am ratios. Diastolic dysfunction was defined on the basis of velocities lower than the age-specific cut-off value [22].
Statistical analysis Values are reported as mean SD for normally distributed variables and as median and interquartile ranges for variables with skewed distribution. Normality of distribution was assessed with the Kolmogorov–Smirnov test, and variables with skewed distribution were analysed after logarithmic transformation. Pearson’s chi-square test was used to compare frequency distributions. Student’s t-test was used for comparisons between two groups. Relationships between continuously distributed variables were examined through linear regression analysis, with correlation expressed by Pearson’s correlation coefficient. Multivariate regression analysis was done with inclusion of variables according to the level of correlation found in univariate analysis. A value of P of < 5% was considered to indicate statistical significance. Data analyses were performed using Stata 12.1 (StataCorp LP, College Station, TX, USA).
Results The clinical and biochemical characteristics of the study patients are summarized in Table 1 where patients are divided according to gender. In addition to older age, lower BMI and waist circumference and lower alcohol intake, women had lower uric acid and triglyceride levels, higher HDL cholesterol and lower fasting insulin that was associated with better insulin sensitivity, indicating a better overall metabolic profile than men. As expected, men had greater LV mass index than women, but frequency of LV hypertrophy was comparable between sexes. Among 83 patients with LV hypertrophy, 45 (12%; 24 males, 21 females) had concentric and 38 (11%; 18 males and 20 females) eccentric hypertrophy; concentric remodelling was present in 43 patients (12%; 23 males and 20 females) showing no significant differences in patterns of LV geometry between sexes. Ejection fraction was slightly higher in women who had also lower E/A ratio than men, but difference disappeared after correction by age. Variables of LV diastolic function obtained at TDI and prevalence of diastolic dysfunction did not differ between men and women. Fasting plasma glucose was higher in patients who had antihypertensive treatment washed out before the study than patients who had never received antihypertensive treatment, whereas the remaining glicometabolic and lipid variables, uric acid levels, left ventricular mass index and the other echocardiographic variables were comparable between the groups.
974
www.ejci-online.com
LV hypertrophy was found in 83 (226%; 42 males, 41 females) hypertensive patients and Table 2 shows the characteristics of patients with or without LV hypertrophy. Patients with LV hypertrophy were older had greater BMI and waist circumference, higher blood pressure, higher uric acid levels and more prevalent hyperuricaemia and greater glucose and insulin responses to OGTT than patients without LV hypertrophy. RWT was greater in patients with LV hypertrophy and was associated with worse diastolic function, whereas systolic function did not differ significantly from patients without LV hypertrophy. On univariate analysis, uric acid levels were significantly and directly related to BMI, waist circumference, alcohol consumption, serum triglyceride and creatinine levels, fasting and post-OGTT insulin, HOMA index and inversely to HDL cholesterol and GFR (Table 3). Moreover, uric acid levels were directly related with LV mass index in both men and women and with left atrial diameter, whereas no significant relationship was found with other variables of diastolic function. In addition to uric acid levels, LV mass index was significantly and directly related with age, BMI, waist circumference, systolic blood pressure and plasma creatinine (Table 4). Multivariate regression analysis that included demographic variables, BMI, blood pressure, duration of hypertension, alcohol consumption, plasma creatinine, fasting plasma insulin, HOMA index and components of the metabolic syndrome (waist circumference and fasting glucose, triglyceride and HDL cholesterol levels) showed that uric acid was independently related with waist circumference (b = 0206, P = 0003), plasma creatinine (b = 0435, P < 0001) and fasting insulin (b = 0185, P = 0006) and that LV mass index was independently related with age (b = 0198, P = 0001), waist circumference (b = 0133, P = 0048) and systolic blood pressure (b = 0316, P < 0001). Multivariate regression analysis was also performed separately in men and women (Table 5) including the same variables as above. In hypertensive men, LV mass index was independently related to age and systolic blood pressure and in hypertensive women with systolic blood pressure, AUC-G and uric acid levels.
Discussion Both hyperuricaemia and LV hypertrophy are frequently associated with the metabolic syndrome and contribute to the cardiovascular risk. Past studies that investigated the association between uric acid and LV mass in hypertensive patients reported inconsistent findings, most probably because results were affected by metabolic variables related to this syndrome. For these reasons, we excluded patients with the metabolic syndrome and conducted the study in nondiabetic hypertensive patients free of clinically relevant cardiovascular and renal complications. Findings demonstrate that frequency of hyperuricaemia in hypertensive patients with LV hypertrophy is
ª 2014 Stichting European Society for Clinical Investigation Journal Foundation
URIC ACID AND THE HEART
Table 1 Clinical, biochemical and echocardiographic characteristics of the study patients Variable
All patients (n = 367)
Men (n = 194)
Women (n = 173)
P
Clinical characteristics Age, years
52 14
50 14
55 15
0005
BMI, kg/m2
267 41
272 36
261 46
0010
91 11
96 9
86 12
< 0001
Waist circumference, cm
72 11
73 11
72 12
0405
Systolic blood pressure, mmHg
150 19
151 19
150 18
0732
Diastolic blood pressure, mmHg
92 12
93 13
91 11
0053
Heart rate, beats/min
Duration of hypertension, years
5 [1–11]
5 [1–11]
5 [1–10]
0554
Alcohol consumption, g/day
5 [0–20]
15 [0–39]
0 [0–10]
< 0001
Biochemical variables Total cholesterol, mg/dL
204 41
200 38
210 44
0043
HDL cholesterol, mg/dL
61 16
55 13
68 17
< 0001
LDL cholesterol, mg/dL
124 37
124 35
123 40
Triglycerides, mg/dL Uric acid, mg/dL Hyperuricemia, n (%) Creatinine, mg/dL GFR, mL/min 173 m2 Fasting glucose, mg/dL Fasting insulin, lUI/mL
91 [70–115]
98 [76–128]
83 [60–107]
0790 < 0001
515 146
584 121
437 132
51 (14)
31 (16)
20 (12)
097 026
108 026
085 018
92 25
93 25
90 24
0419
89 18
0271
90 16 760 [543–1030]
91 15 810 [598–1083]
< 0001 0238 < 0001
695 [513–1000]
0026
HOMA index
160 [114–227]
172 [117–252]
146 [107–210]
0015
AUC-G, mg/dL min
383 98
376 77
391 120
0294
AUC-I, lUI/mL min
153 [103–252]
147 [100–255]
161 [104–233]
0715
Echocardiographic variables LVMI, g/m27 LV hypertrophy, n (%)
413 134
429 142
394 122
0012
83 (23)
42 (21)
41 (24)
0599
038 008
038 008
038 009
0894
LV ejection fraction, %
69 7
68 7
71 6
0001
LAD, mm
38 7
41 7
36 6
< 0001
E/A ratio
113 040
118 041
106 038
0007
93 20
89 22
98 21
0291
RWT, %
IVRT, msec Em, cm/sec
97 26
96 24
98 31
0676
Em/Am ratio
114 045
108 038
119 050
0241
E/Em ratio
170 215
757 243
782 187
0585
48 (49)
23 (49)
25 (50)
0736
Diastolic dysfunction (TDI), n (%)
Values are expressed as mean SD. Interquartiles ranges are shown in square brackets [IQR] for variables with skewed distribution. To convert to international units, multiply creatinine by 884 (lM), cholesterol by 00259 (mM), triglycerides by 00113 (mM), uric acid by 59485 (lM), glucose by 005551 (mM) and insulin by 7175 (pM). Tissue Doppler imaging variables were measured in a subset of 97 patients. BMI, body mass index; GFR, glomerular filtration rate; HOMA, homeostatic model assessment; AUC-G, area under the curve of plasma glucose after oral glucose tolerance test; AUC-I, area under the curve of plasma insulin after oral glucose tolerance test; LVMI, left ventricular mass index; RWT, relative wall thickness; LAD, left atrial diameter; E, early-wave transmitral diastolic velocity; A, late-wave transmitral diastolic velocity; IVRT, isovolumetric relaxation time; Em, early-diastolic velocity of septal and lateral myocardial portions at tissue Doppler imaging; Am, late-diastolic velocity of septal and lateral myocardial portions at tissue Doppler imaging; TDI, tissue Doppler imaging.
European Journal of Clinical Investigation Vol 44
975
C. CATENA ET AL.
www.ejci-online.com
Table 2 Clinical, biochemical and echocardiographic variables of patients grouped according to the presence or absence of left ventricular (LV) hypertrophy LV hypertrophy No (n = 284)
Variable
LV hypertrophy Yes (n = 83)
P
Clinical characteristics Age, years
51 14
56 15
0004
Males, n (%)
152 (53)
42 (51)
0639
264 40
277 44
0016
Waist circumference, cm
90 11
95 11
< 0001
Heart rate, beats/min
72 11
71 12
0616
Systolic blood pressure, mmHg
148 17
158 23
< 0001
Diastolic blood pressure, mmHg
91 12
95 14
0054
2
BMI, kg/m
Duration of hypertension, years Previous treatment, n (%) Smokers, n (%) Alcohol consumption, g/day
4 [1–10]
8 [2–12]
0261
112 (39)
36 (43)
0520
47 (16)
13 (16)
0848
6 [0–20]
1 [0–20]
0363
Biochemical variables Total cholesterol, mg/dL
205 42
201 39
0470
HDL cholesterol, mg/dL
61 17
60 14
0541
LDL cholesterol, mg/dL
125 38
120 36
0297
Triglycerides, mg/dL
91 [68–116]
91 [75–110]
0760
506 140
546 161
0041
32 (11)
19 (23)
0007
Creatinine, mg/dL
096 020
102 038
0160
GFR, mL/min/173
101 29
95 34
0216
90 16
91 17
0566
Uric acid, mg/dL Hyperuricemia, n (%)
Fasting glucose, mg/dL Fasting insulin, lUI/mL
745 [550–1071]
780 [500–945]
0266
HOMA index
160 [115–237]
160 [106–220]
0976
AUC-G, mg/dL min
371 92
419 110
0004
AUC-I, lUI/mL min
146 [96–226]
189 [139–308]
0004
36 7
60 13
< 0001
0361 0070
0449 0091
< 0001
LV ejection fraction, %
70 7
68 8
0084
LAD, mm
38 6
41 10
0007
116 041
102 035
0011
IVRT, msec
88 17
100 27
0021
Em, cm/sec
101 28
86 24
0024
Echocardiographic variables LVMI, g/m27 RWT, %
E/A
976
ª 2014 Stichting European Society for Clinical Investigation Journal Foundation
URIC ACID AND THE HEART
Table 2 Continued Variable
LV hypertrophy No (n = 284)
LV hypertrophy Yes (n = 83)
P
Em/Am
117 044
101 045
0148
E/Em
722 180
928 247
0001
34 (45)
14 (64)
0507
Diastolic dysfunction (TDI), n (%)
Values are expressed as mean SD. Interquartiles ranges are shown in square brackets [IQR] for variables with skewed distribution. To convert to international units, multiply creatinine by 884 (lM), cholesterol by 00259 (mM), triglycerides by 00113 (mM), uric acid by 59485 (lM), glucose by 005551 (mM) and insulin by 7175 (pM). Tissue Doppler imaging variables were measured in a subset of 97 patients. BMI, body mass index; GFR, glomerular filtration rate; HOMA, homeostatic model assessment; AUC-G, area under the curve of plasma glucose after oral glucose tolerance test; AUC-I, area under the curve of plasma insulin after oral glucose tolerance test; LVMI, left ventricular mass index; RWT, relative wall thickness; LAD, left atrial diameter; E, early-wave transmitral diastolic velocity; A, late-wave transmitral diastolic velocity; IVRT, isovolumetric relaxation time; Em, early-diastolic velocity of septal and lateral myocardial portions at tissue Doppler imaging; Am, late-diastolic velocity of septal and lateral myocardial portions at tissue Doppler imaging; TDI, tissue Doppler imaging.
twofold that of patients without LV hypertrophy and that uric acid levels are significantly and directly related with LV mass. This relationship, however, is independent of possible confounders including renal function and sensitivity to insulin in women, but not in men. The frequent association of increased uric acid levels with insulin resistance has led to indicate hyperuricaemia as a possible additional component to the cluster of conditions that identify the metabolic syndrome. Initial observations reported a strong relationship between decreased insulin-mediated glucose disposal and uric acid levels in healthy subjects [27] and these observations were subsequently replicated in patients with the metabolic syndrome [5]. Also, healthy subjects with asymptomatic hyperuricaemia were reported to be hyperinsulinemic and dyslipidemic and to have higher blood pressure than normouricaemic subjects [28]. The suggested mechanism underlying the relationship between insulin resistance/hyperinsulinemia and uric acid is that elevated plasma insulin decreases renal clearance of uric acid [29]. In this study, we have found a direct relationship of uric acid levels with the HOMA index and fasting and post-OGTT plasma insulin in hypertensive patients without the metabolic syndrome. This finding, obtained in a different set of patients from previous studies, supports the view that elevated uric acid is another of the multiple metabolic abnormalities that comprise the insulin resistance syndrome. Increased LV mass is associated with insulin resistance in the context of the metabolic syndrome [3] and appears to be under the influence of gender, inasmuch as the effects of the metabolic syndrome on LV mass appear to be more pronounced in women [30]. In our hypertensive patients without the metabolic syndrome, the response of both plasma glucose and insulin levels to the OGTT was greater in patients with LV hypertrophy than without LV hypertrophy, but no significant differences were found in the HOMA index. These findings are in agreement with those reported by Sciacqua et al. [31] in 767 never treated
hypertensive subjects and suggest that postprandial glucose levels are a major determinant of LV mass in hypertensive patients without the metabolic syndrome. Because of the relevance that LV hypertrophy has on cardiovascular risk [1], the relationship of uric acid with LV mass of hypertensive patients was examined in previous studies. Viazzi et al. [8] reported an independent association between uricaemia and LV mass in 425 hypertensive patients, 21% of whom had the metabolic syndrome. The same association was observed in a study of 619 hypertensive patients who were followed for 34 months, showing that the combination of hyperuricaemia with LV hypertrophy was an independent predictor of cardiovascular events [9]. Subsequent observations confirmed the association between uric acid and LV mass, although this was found to be gender related [10,11] and a recent study conducted on 534 uncomplicated hypertensive patients has reported an association between circulating uric acid and LV mass index only in women [13]. At opposite, no independent association of uric acid levels with LV mass was reported in a study of 580 newly diagnosed, never treated, hypertensive patients with relatively low prevalence of hyperuricaemia [12] and in other large studies [32,33]. In addition to generic differences in ethnicity, age and gender distribution of patients, duration of hypertension, and prevalence of hyperuricemia or LV hypertrophy, important additional factors could explain the inconsistencies among these studies. These factors include the presence of the metabolic syndrome, impaired renal function and interference of antihypertensive drugs such as diuretics. For these reasons, we excluded patients with the metabolic syndrome, GFR of < 60 mL/min/173 m2, and previous use of diuretics. In these patients, we have found a significant correlation between uric acid levels and LV mass, although this correlation was independent of insulin sensitivity and renal function only in women. Relevant to this aspect, previous studies suggested a greater impact on LV structure of women than men of metabolic conditions such as glucose
European Journal of Clinical Investigation Vol 44
977
C. CATENA ET AL.
www.ejci-online.com
Table 3 Univariate correlation analysis with serum uric acid as the dependent variable All patients Variable
r
Men P
r
Women P
r
P
Age
0046
0375
0065
0369
0217
0004
BMI
0221
< 0001
0155
0031
0198
0009
Waist circumference
0448
< 0001
0154
0087
0321
< 0001
Systolic blood pressure
0048
0363
0125
0082
0035
0646
Diastolic blood pressure
0069
0188
0065
0371
0022
0776
Duration of hypertension
0065
0213
0109
0142
0159
0045
Alcohol consumption
0170
0002
0044
0625
0155
0053
Total cholesterol
0084
0109
0013
0854
0053
0488
HDL cholesterol
0325
< 0001
0015
0840
0315
< 0001
LDL cholesterol
0014
0790
0081
0284
0048
0542
Triglycerides
0305
< 0001
0212
0003
0224
0003
Creatinine
0495
< 0001
0246
0001
0509
< 0001
GFR
0142
0011
0209
0005
0180
0027
Fasting glucose
0072
0173
0076
0300
0026
0738
Fasting insulin
0253
< 0001
0217
0007
0216
0015
HOMA index
0256
< 0001
0214
0008
0212
0017
AUC-G
0061
0360
0107
0028
0122
0220
AUC-I
0140
0038
0198
0028
0042
0686
LVMI
0212
< 0001
0147
0041
0198
0009
LAD
0309
< 0001
0312
< 0001
0002
0982
E/A
0092
0127
0081
0355
0066
0459
IVRT
0042
0581
0039
0707
0073
0513
Em
0092
0372
0023
0878
0136
0345
Em/Am
0067
0512
0005
0972
0010
0946
E/Em
0121
0245
0359
0014
0078
0595
BMI, body mass index; GFR, glomerular filtration rate; HOMA, homeostatic model assessment; AUC-G, area under the curve of plasma glucose after oral glucose tolerance test; AUC-I, area under the curve of plasma insulin after oral glucose tolerance test; LVMI, left ventricular mass index; LAD, left atrial diameter; E, early-wave transmitral diastolic velocity; A, late-wave transmitral diastolic velocity; IVRT, isovolumetric relaxation time; Em, early-diastolic velocity of septal and lateral myocardial portions at tissue Doppler imaging; Am, late-diastolic velocity of septal and lateral myocardial portions at tissue Doppler imaging.
intolerance [34], obesity [35] and the metabolic syndrome [30,36], and therefore, the link between uric acid levels and insulin resistance might contribute to explain our findings. The mechanism(s) underlying the interaction of gender on the association between hyperuricaemia and left ventricular mass, however, remain hypothetical. Greater visceral fat accumulation, increased large artery stiffness and blunting of favourable cardiovascular effects of oestrogen [37] caused in women by insulin resistance [38] are possible explanations that should be tested in future studies.
978
Evidence on the association between hyperuricaemia and cardiovascular outcomes is growing [39], and this association might be of great relevance in patients with high blood pressure because uric acid might contribute to subclinical hypertensive organ damage. Circulating levels of uric acid reflect, among many other factors, the activity of xanthine oxidase, an enzyme crucially involved in purine metabolism. Pharmacologic blockade of xanthine oxidase resulting in decreased uric acid generation has been reported to improve endothelial function and decrease markers of inflammation [39]. Also, the
ª 2014 Stichting European Society for Clinical Investigation Journal Foundation
URIC ACID AND THE HEART
Table 4 Univariate correlation analysis with left ventricular mass index as the dependent variable All patients Variable
r
Men P
Women
r
P
r
P
Age
0218
< 0001
0204
0004
0291
< 0001
BMI
0219
< 0001
0143
0046
0276
< 0001
Waist circumference
0276
< 0001
0041
0651
0410
< 0001
Systolic blood pressure
0309
< 0001
0351
< 0001
0250
0001
Diastolic blood pressure
0125
0171
0178
0014
0018
0815
Duration of hypertension
0065
0236
0066
0388
0076
0347
Alcohol consumption
0077
0160
0065
0394
0070
0384
Total cholesterol
0002
0974
0132
0067
0179
0019
HDL cholesterol
0074
0171
0021
0782
0026
0742
LDL cholesterol
0034
0531
0173
0021
0177
0139
Triglycerides
0139
0072
0030
0685
0281
< 0001
Creatinine
0169
0014
0148
0041
0086
0259
GFR
0133
0540
0157
0038
0035
0669
Fasting glucose
0020
0706
0013
0862
0010
0894
Fasting insulin
0009
0881
0008
0921
0016
0857
HOMA index
0009
0881
0008
0923
0022
0812
AUC-G
0156
0145
0014
0875
0356
< 0001
AUC-I
0157
0140
0198
0028
0106
0300
Uric acid
0212
< 0001
0147
0041
0198
0009
BMI, body mass index; GFR, glomerular filtration rate; HOMA, homeostatic model assessment; AUC-G, area under the curve of plasma glucose after oral glucose tolerance test; AUC-I, area under the curve of plasma insulin after oral glucose tolerance test.
Table 5 Multivariate analysis with left ventricular mass index as the dependent variable in men and women Men Variable
b
P
Women Variable
b
P
Age
0044
0707
Systolic blood pressure
0357
0001
Age
0242
0008
Systolic blood pressure
0374