Evaluation of Arterial Stiffness and Carotid Intima-Media Thickness in Children with Primary and Renal Hypertension

Hypertension is an increasing disease in children and the risk of endothelial damage and target organ damage increases in the presence of additional risk factors such as obesity. In our study, the effect of hypertension on early atherosclerotic changes and target organ damage in children was investigated. Twenty four-hour ambulatory pulse wave analysis was performed by oscillometric method in 71 children aged 8–18 years, 17 of whom were diagnosed with primary hypertension without obesity, 18 had both primary hypertension and obesity, and 16 had renal hypertension. Twenty healthy normotensive children were included as the control group. Carotid intima-media thickness (CIMT) and Left Ventricular Mass Index were measured. Central systolic blood pressure (cSBP), central diastolic blood pressure (cDBP), systolic blood pressure (SBP) and diastolic blood pressure (DBP) were higher in the primary hypertension group compared to controls (p = 0.001, p = 0.005, p = 0.001, p = 0.009, respectively), cSBP was higher in the renal hypertension group than the control group (p = 0.018). There was no difference between the groups in terms of pulse wave analysis parameters, CIMT, or left ventricular mass index (p > 0.05). Pulse wave velocity was positively correlated with SBP, DBP, cSBP, cDBP (p < 0.001). Augmentation index was positively correlated with DBP and cDBP (p = 0.01, p = 0.002, respectively). Our findings show that high blood pressure is associated with arterial stiffness and target organ damage beginning in childhood. The detection of early atherosclerotic vascular changes using pulse wave analysis allows to take necessary precautions such as lifestyle changes to prevent target organ damage in hypertensive children.


Introduction
Hypertension has an increasing prevalence in children and adolescents. Childhood hypertension is an important risk factor for cardiovascular system diseases, chronic renal failure, and cerebrovascular diseases in adulthood [1]. Although morbid cardiovascular events are rare in pediatric patients, it is important to identify hypertensive children at risk for complications later in life. Some target organ damage markers are used for this purpose, such as left ventricular hypertrophy and increased carotid intimamedia thickness (CIMT) [2].
Arterial stiffness is an indicator of atherosclerosis and occurs due to thickening and loss of elasticity of the arterial wall. It is a precursor to target organ damage and increased cardiovascular events [3]. The measurement of pulse wave velocity (PWV) was first proposed by the European Society of Hypertension-European Society of Cardiology (ESH-ESC) in 2003 to determine arterial stiffness [4]. Oscillometric and tonometric methods are the most commonly used non-invasive methods for in pulse wave analysis. Radial artery aplanation tonometry is performed by placing a hand-held tonometer (strain gauge pressure sensor) over the radial artery and applying mild pressure to partially flatten the artery [5]. The radial artery pressure is then transmitted from the vessel to the sensor (strain gauge) and is recorded digitally. In the oscillometric method, the measurement is made with the cuff placed on the upper arm. A small number of studies have shown that PWV is increased in children with obesity and hypertension [4,6].
Evaluation of subclinical end organ damage caused by hypertension in children will enable to identify future cardiovascular diseases and to reduce the possibility of end organ damage by taking lifestyle precautions in these individuals. Therefore, in our study, it was planned to measure PWV, augmentation index (AIx), central and peripheral blood pressure values by 24-h ambulatory pulse wave analysis with oscillometric method, CIMT by ultrasonographic method, and left ventricular mass index (LVmass index) by M-mode echocardiography (ECHO) in children with primary and renal hypertension and in the age-and-sex matched healthy control group.
Although there are many studies in the literature on the effects of primary hypertension and obesity on vessel wall damage, the number of studies comparing hypertension due to chronic renal diseases with primary hypertension is few. In our study, we aimed to investigate whether the effect of exposure to higher pressure load on arterial stiffness and end organ damage is different in the renal hypertension group compared to the primary hypertensive group, since hypertension is difficult to control in chronic renal diseases.

Materials and Method
The study included 51 patients who were previously diagnosed with hypertension and had been receiving antihypertensive medication for a certain period of time in the Departments of Pediatric Cardiology and Pediatric Nephrology, Faculty of Medicine, Eskişehir Osmangazi University and 20 age-and-sex matched healthy controls who applied to the Department of Pediatric Cardiology with nonspecific chest pain. The study population were divided into 4 groups as non-obese subjects with primary hypertension (PH group, n: 17) obese subjects with primary hypertension (OPH group, n: 18), subjects with renal hypertension (RH group, n: 16), and healthy subjects (control group, n: 20). At the first admission, three measurements were made at different times to these patients while they were in a rested and sitting position, with an appropriately sized cuff, and those whose blood pressure values were above the 95th percentile for their respective age, height and gender were diagnosed as hypertensive. Those whose secondary hypertension etiologies were clearly excluded were defined as primary hypertensive.
This cross-sectional study was conducted prospectively. The families of the children were informed about the purpose and method of the study and their written consent was obtained regarding their voluntary participation. The study protocol was approved by the Faculty of Medicine Ethics Committee at Eskisehir Osmangazi University with the decision numbered 80,558,721/59 and dated 27.06.2019.
The children's medical history, background, and family history were questioned and physical examinations were performed. Children with a history of congenital heart disease, diabetes, or any other chronic disease and those who had a condition that required medication other than antihypertensive treatment were excluded. Echocardiographic examinations were performed by an experienced pediatric cardiologist in the Department of Pediatric Cardiology using a Philips Epiq 3D ultrasound device. Left ventricular mass (LVmass) was calculated by Penn-Cube formula (LVmass = 1.04 [(LVEDd + IVSD + LVPWd)3− (LVEDd)3]− 13.6) [7]. LVmass index was calculated by dividing LVmass by the body surface area. Relative wall thickness (RWTh) was calculated using the formula RWTh = 2xPWd/LVEDd [8].
CIMT measurement was performed in all patients using VIVID I color Doppler ultrasonography (General Electric Ultrasound Systems, Mountain View, CA, USA) equipment (software version 1.36.19) with a 12-MHz linear probe using semi-automated measured method. The subjects were placed on supine position and a thin pillow was placed under their necks and then, their necks were turned to the opposite side. A 1 cm segment was determined within the first 2 cm distal region from the common carotid artery bulb and the images were transferred to the computer environment. Based on the far wall measurement method the mean, maximum, and minimum values of the segments were determined [9]. These measurements were repeated for both main carotid arteries and mean values were obtained.
The patients were weighed on a calibrated electronic scale (SECA digital scale, sensitive to 0.1 kg) with thin clothes and no shoes. Height measurements were made with a Harpenden stadiometer (sensitivity to 0.1 cm) in standing upright position, with bare feet and feet adjacent and parallel, and shoulders and the gluteal region in contact with the wall. Body mass index (BMI) was calculated using height and weight measurements [weight (kg)/height 2 (m 2 )]. According to the percentile curves determined by the World Health Organization, those with a body mass index in the 95 th percentile and above were considered obese [10].
Mobil-O-Graph 24 h Pulse Wave Analysis Monitor (IEM, Industrielle Entwicklung Medizintechnik und Vertriebsgesellschaft mbH, Stolberg, Germany) was used to automaticly measure 24-h pulse wave velocity by oscillometric method. Mobil-O-Graph 24 h Pulse Wave Analysis Monitor is a system which can measure both pulse wave velocity and the other pulse wave analysis parameters by oscillometric method. The children were recommended not to do heavy exercise before monitoring and not to take any caffeinated beverages or medical treatment the day before. A cuff of the appropriate size for the children's age and upper arm circumference was tied. The device was set at a protocol to measure every 20 min during daytime (when the children are awake) and once every 30 min during night time (when the children are asleep). The Hypertension Management Software Client Server (HMS-CS) version 2.0 was used for the transfer and analysis of the measurements.
Data analysis was employed with IBM SPSS 21. Descriptive statistics of quantitative variables were given as mean ± standard deviation or median (Q1-Q3) whereas categorical variables was presented as count and percentage. Normality of quantitative variables was evaluated with Shapiro Wilk test. For normally distributed variables, One Way ANOVA was used for overall comparison of groups, while pairwise comparisons were made with Tukey test. Kruskal Wallis test was performed for non-normal distributed group comparisons. Pairwise comparisons were carried out with the Dunn test for significant Kruskal Wallis test results. Chi square analysis were used to assess the relationship between categorical variables. Association between quantitative variables were evaulated with Pearson and Spearman correlation analysis. Regression models were created using the backward stepwise method. Variables with a variance influance factor (VIF) value below 10 were decided to be included in the final model. P values less than 0.05 were considered significant.

Results
Our study was carried out with a total of 71 children aged 8-18 years, 17 of whom had primary hypertension, 18 had both primary hypertension and obesity, and 16 had renal hypertension. Twenty healthy normotensive children were included as the control group. The demographic characteristics of the groups are given in Table 1.
At the beginning of the study, mean duration after the diagnosis of hypertension was 22.2 ± 21.8 months in the PH group, 17.5 ± 22.8 months in the OPH group, and 26.3 ± 27 months in the RH group, with no statistically significant difference (p > 0.05). Evaluating all study groups together, no significant correlation was found between the duration of hypertension and pulse wave analysis parameters or CIMT values (p > 0.05).
Renal function tests, renal ultrasonography, and renal Doppler ultrasonography findings were normal in the PH and OPH groups.
LVmass index, IVSDd, LVEDd, and LVPWd measurements were within normal limits according to body surface area in all study and control subjects. Two patients in the OPH group, 3 in the PH group, and 1 in the RH group had an RWT value above 0.42. Ejection fraction and shortening fraction values were above normal in 2 patients in the RH group and shortening fraction was above normal in 1 patient in the PH group. No cardiomyopathy was observed in any subject of the all groups.
Pulse wave analysis parameters were statistically compared between the study and control groups ( Table 2). Systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP) were statistically higher in the PH group than the control group (p = 0.001, p = 0.009, p = 0.002, respectively). MAP was statistically higher in the RH group compared to controls (p = 0.038). Central systolic blood pressure (cSBP) was found to be statistically higher in the PH and RH groups compared to the control group (p = 0.001, p = 0.018, respectively). Central diastolic blood pressure (cDBP) was statistically higher in the the PH group than the control group (p = 0.005). AIx was higher in the PH group than all the other groups, albeit not with statistical significance (p = 0.07). There was no statistically significant difference between the groups in terms of heart rate (HR), pulse pressure (PP), reflection magnitude (RM), peripheral resistance (PR), or PWV (p > 0.05).
CIMT and ECHO parameters were statistically compared between the study and control groups (Table 3). IVSDd was statistically higher in the OPH and RH groups than the control group (p = 0.02, p = 0.014, respectively). Mean LVmass was statistically higher in the OPH group compared to the controls (p = 0.016). There was no statistically significant difference between the groups in terms of the other echocardiographic parameters or CIMT values (p > 0.05).
The correlations of pulse wave analysis parameters and anthropometric measurements in all groups (n = 71) are given in Table 4.
The correlations between SBP, DBP, MAP, HR, PP, cSBP, cDBP values and AIx, PR, RM, and PWV are shown in Table 5. It was created by taking into account the significant correlations among these parameters backward stepwise regression model results for predicting AIx, PR and RM are given in Table 6. The significant independent variables explaining the change in AIx are HR, PP, cSBP, cDBP, the significant independent variables explaining the change in PR are MAP and PP, and the significant independent variables explaining the change in RM are DBP, PP, cSBP, cDBP.
Echocardiography parameters and correlations between CIMT measurements and anthropometric measurements are given in Table 7.   Correlations of pulse wave analysis parameters and biochemical parameters are given in Table 8.

Discussion
Childhood hypertension is an important risk factor for cardiovascular system diseases, chronic renal failure, and cerebrovascular diseases in adulthood [1]. Endothelial dysfunction and atherosclerosis are largely responsible for the development of hypertensive end organ damage. In cases who have hypertension associated with obesity and dyslipidemia, it is known that atherosclerosis findings in the aorta and coronary arteries begin in childhood. In our study, the presence of early atherosclerosis and arterial dysfunction findings in hypertensive pediatric patients was investigated by examining the parameters of arterial stiffness and CIMT.
In the present study, the BMI of obese hypertensive children was found to be higher compared to primary and renal hypertensive patients and healthy controls. However, peripheral and central blood pressure values were similar to other patients and healthy children. It has been reported that obese children have higher mean systolic and diastolic blood pressure levels in comparison to those with normal weight [13]. The lack of a significant difference in blood pressure values in obese hypertensive children despite high BMI values may be related to the low mean age of our patients and the low number of subjects. In addition, 44.4% of the patients in the OPH group were receiving antihypertensive treatment and strict diet control, resulting in an increased number of patients whose blood pressure values are under control. But, although the rate of antihypertensive drug use was higher in the PH and RH groups, peripheral MAP and cSBP values in both groups, and additionally peripheral SBP, DBP, and cDBP values in the PH group were higher compared to healthy children. Regarding the higher blood pressure values in the PH group, it was thought that their familial tendency to hypertension and less emphasis on dietary controls due to their normal BMI values may have played a role in the detection of higher blood pressure values in those children. However, in hypertension due to chronic renal diseases, blood pressure control is very difficult and often requires multiple antihypertensive therapy.
Pulse wave analysis by the oscillometric method is a very useful and reliable method in the evaluation of arterial stiffness especially in adults, and its use in children is becoming increasingly common. A systematic review and meta-analysis of prospective studies reported that elevated blood pressure in the young was associated with an intermediate marker of cardiovascular disease, high PWV, in adulthood [14]. In children with hypertension, especially    in the presence of obesity, the increase in PWV supports the increase in arterial stiffness [4,6,15]. Meng et al. [16] examined hypertensive and normotensive children and found PWV to be higher in the hypertensive group. Kulsum-Mecci et al. [15] compared normotensive obese, hypertensive, and hypertensive obese children with a healthy control group and found PWV to be significantly higher in all patient groups than the control group. They also stated that PWV increased with age and did not differ according to sex or race. Niboshi et al. [17] found that age, blood pressure, and heart rate were key determinants of PWV in healthy Japanese children, although obesity and PWV did not correlate. The augmentation index is another frequently used measure of arterial stiffness along with PWV. The fact that the shorter aortic length in young children causes the reflected wave to return early causes AIx to decrease gradually with age [18]. Wójtowicz et al. [19] compared obese/ overweight children with or without primery hypertension with a healthy control group and reported that obese children with arterial hypertension showed the highest values of PWV, with no difference in terms of AIx, RM, or PR. Tokgöz et al. [20] reported that PWV and AIx were higher in children with primary and white coat hypertension compared to healthy controls and this was associated with the duration of hypertension. In our study, there was no statistically significant difference between the groups in terms of arterial stiffness parameters such as AIx, RM, PR, and PWV (p > 0.05). No correlation was found between duration of hypertension and PWV, AIx, RM, or CIMT. However, in accordance with the literature, PWV was positively correlated with height, weight, BMI, SBP, DBP, MAP, PP, cSBP, and cDBP; AIx was negatively correlated with age, height, and weight and positively correlated with DBP, MAP, HR, and cDBP. Also, CIMT values were positively correlated with AIx and RM. According to our regression model the significant independent variables explaining the change in AIx are HR, PP, cSBP, cDBP, the significant independent variables explaining the change in PR are MAP and PP, and the significant independent variables explaining the change in RM are DBP, PP, cSBP, cDBP. This finding supports that central blood pressure values may be more significant than peripheral measurements for determining target organ damage due to hypertension in hypertensive pediatric patients. In our study, most of the patients were diagnosed with hypertension recently, which suggests that exposure to high blood pressure was short-lived. Besides, the low mean age of our study group and the low number of subjects might have been effective in the lack of difference in terms of endothelial damage findings and the lack of correlation between duration of hypertension and atherosclerosis markers.
Measurement of central (aortic) blood pressure in hypertensive patients has gained more importance as it gives a clearer idea about cardiovascular events than peripheral (arm) measurement. Central blood pressure is more sensitive in evaluating target organ damage and response to antihypertensive therapy, since central blood pressure directly affects the organs and peripheral blood pressure does not always reflect central blood pressure [21]. Litwin et al. [22] found central blood pressure to be normal in all patients with white coat hypertension and in 93% of the prehypertensive group. In the severe hypertension group, mean cSBP, PP, PWV, and the prevalence of left ventricular hypertrophy was found to be significantly higher than in children with white coat hypertension, prehypertensive, and hypertensive children. They also found that LVmass index and CIMT correlated with both cSBP and PP. In our study, there was no significant correlation between LVmass index and peripheral SBP, but a significant positive correlation with cSBP (p = 0.02). Moreover, a positive correlation was found between both peripheral and central DBP and LVmass index (p = 0.002). This finding supports that central blood pressure values may be more significant than peripheral measurements for determining target organ damage due to hypertension in hypertensive pediatric patients, as in adult patients.
Hypertension is an important risk factor for atherosclerosis. Especially in patients with additional risk factors such as diabetes, hyperlipidemia, and obesity, endothelial dysfunction caused by hypertension considerably increases the risk of developing atherosclerosis. Lande et al. [23] reported that CIMT was increased in hypertensive children and was correlated with SBP. Pandit et al. [24] found that CIMT did not differ between normotensive obese children and healthy controls. Similarly, we found no statistical difference between the groups in terms of CIMT, despite including non-obese primary hypertensive patients and those with additional risk factors such as obesity and chronic renal disease (p > 0.05). Also, CIMT measurements were not significantly correlated with blood pressure parameters or anthropometric measurements. This was again associated with the low mean age of our patients and the fact that hypertension was under control.
Cardiac remodeling caused by hypertension is the most prominent form of target organ damage and is associated with increased cardiovascular mortality and morbidity. It has been shown that increased left ventricular mass and concentric hypertrophy in hypertensive adult patients increase the risk of cardiovascular events [25,26]. Although hypertension-related left ventricular hypertrophy can be seen in children, it is more common, particularly in late adolescent and young adult age groups [27]. Celik et al. [28] found LVEDd, IVSDd, LVPWd, and LVmass indexes to be significantly higher in obese children compared to the controls. In our study, IVSDd and LVmass were statistically significantly increased in the OPH group compared to the control group (p = 0.02 and p = 0.016, respectively). There was no statistically significant difference between the groups in terms of LVmass index and RWT. The fact that IVSDd was statistically significantly higher (p = 0.014) in the RH group as well as OPH group than the control group suggests the potential role of the pressure overload effect of hypertension, along with hypertrophy caused by obesity, in the formation of cardiac target organ damage. Besides, despite the lack of a significant correlation between LVmass index and BMI in our study, high blood pressure was considered to be related with left ventricular hypertrophy, since LVmass index was positively correlated with peripheral diastolic and mean blood pressures, PP, or central systolic and diastolic blood pressures.
Various risk factors such as smoking, hypertension, and hyperlipidemia initiate an inflammatory response by disrupting the endothelial structure. C-reactive protein is a parameter used to assess inflammation and is a risk factor for cardiovascular events and increased arterial stiffness [29]. Rondò et al. [30] reported that CRP was positively correlated with SBP and waist circumference and negatively correlated with HDL-C levels. Mohamed et al. [31] found that CRP was positively correlated with BMI but not with blood pressure values. In our study, however, no correlation was found between CRP and blood pressure, pulse wave analysis parameters, or CIMT values, except for a weak negative correlation with peripheral and central DBP values.
In humans, uric acid is the main product of the catabolism of purine nucleotides (adenosine and guanosine). Hyperuricemia is associated with hypertension, vascular disease, renal disease, and cardiovascular events. The Bogalusa Heart Study [32], the Moscow Pediatric Hypertension Study [33], reported that serum uric acid levels are associated with hypertension in children. In our study, however, there was no statistically significant difference between the groups in terms of serum uric acid values. But, serum uric acid values were found to be positively correlated with peripheral SBP, PP, and PWV values (p = 0.009, p = 0.001, p = 0.02, respectively) and negatively correlated with AIx (p = 0.001). This finding suggests that, even when uric acid values are not high, they might be correlated with pulse wave analysis parameters.
Urinary albumin excretion reflects glomerular permeability and increased urinary albumin excretion is indicative of increased renal damage. Microalbuminuria is associated with early target organ damage, left ventricular hypertrophy, and increased CIMT in nondiabetic hypertensive patients [34]. Nguyen et al. [35] found a significant relationship between hypertension and microalbuminuria in obese adolescents. On the other hand, Göknar et al. [36] found no difference in terms of microalbuminuria between obese children and healthy controls. Similarly, we found no statistically significant difference between the groups in terms of microalbumin/creatinine in spot urine and microalbumin/creatinine ratio was not correlated with any of the study parameters. Considering that the urine samples were studied at different times of the day, the fact that we could not exclude factors like orthostatic proteinuria may have led to the lack of difference between the groups.
Dyslipidemias are known to play a role in cardiovascular events such as hypertension and obesity in adults. There is evidence indicates that high serum total cholesterol, LDL-C, glucose, and insulin levels in childhood are associated with atherosclerotic outcomes in young adults [37]. Agbaje et al. [38] showed that high PWV was associated with hyperinsulinemia and CIMT was associated with increased LDL, triglyceride and serum glucose in their long-term study with young people. Bjelakovic et al. [39] found that triglyceride/HDL-C ratio was positively correlated with eccentric left ventricular hypertrophy and both BMI and insulin levels were positively correlated with concentric left ventricular hypertrophy in obese children. Raitakari et al. [40] stated that adult CIMT was associated with childhood LDL-C, SBP, and BMI. In our study, serum fasting glucose levels were positively correlated with PP and CIMT right (p = 0.002, p = 0.03, respectively). Also, serum insulin levels were found to correlate with SBP, MAP, PP, cSBP, PWV, IVSDd, LVEDd, LVPWd, LVmass, and RWT (p < 0.001, p = 0.02, p < 0.001, p < 0.001, p = 0.007, p = 0.001, p = 0.05, p = 0.002, p = 0.01, p = 0.05, respectively). Moreover, homa-IR was positively correlated with SBP, MAP, PP, cSBP, PWV, IVSDd, LVPWd, LVmass, and RWT (p < 0.001, p = 0.02, p < 0.001, p < 0.001, p = 0.007, p = 0.001, p = 0.003, p = 0.002, p = 0.05, respectively). However, no correlation was found between serum HDL-C, LDL-C, and triglyceride levels with ECHO parameters or pulse wave analysis parameters. The correlations of PWV with homa-IR and insulin levels supports that insulin resistance is a predictor of subclinical vascular damage and future cardiovascular events.

Limitations
Our study has some limitations. One of them is the small sample size. The reason for this was the limited number of patients who were followed up in our outpatient clinic with a diagnosis of hypertension and met the inclusion criteria. The explanatory power of the regression models we established with variables other than blood pressures and arterial stiffness measurements was low (R2 < 0.25). For this reason, we could not comment on the regression results of them. Our study was designed prospectively, but the measurements were made cross-sectionally, more reliable results can be obtained in terms of end organ damage with measurements to be made after long-term follow-up.

Conclusions
In conclusion, the ambulatory 24-h pulse wave analysis by the oscillometric method revealed that children with primary hypertension had higher SBP, DBP, MAP cSBP, and cDBP values and those with renal hypertension had higher MAP and cSBP values compared to healthy controls. Especially, the positive correlations of central blood pressure values with PWV, AIx, and BMI supports that hypertension has a key role in the formation of arterial stiffness and that obesity is associated with blood pressure and arterial stiffness. LVmass index was found to have no significant correlation with peripheral SBP, although its significant positive correlation with cSBP and cDBP supports that central blood pressure is a better predictor of target organ damage. We believe that using pulse wave analysis and central blood pressure measurement to determine deterioration in arterial structure and functions due to hypertension, is an effective and reliable method in children. This can help prevent the development of cardiovascular disease and target organ damage by enabling necessary lifestyle measures to be taken.