Ultrasonography to detect cardiovascular damage in children with essential hypertension

Introduction

Essential hypertension (HTN) in adults may begin inchildhood (the so-called trajectory phenomenon [1]. Temortality and morbidity of hypertension in children andadolescents are closely related to cardiovascular diseasein adults. However, hypertension has a low incidence ofcardiovascular events in children and adolescents. Terefore, it is of great signifcance to study whether hypertension causes structural and functional damage to the heartand blood vessels in childhood and adolescence, so thatearly target organ damages can amply proves to cardiovascular events in adulthood.

Te diagnostic criteria for HTN in children are not uniform and are generally considered higher than percentilevalues for blood pressure in this age group. In 2017, JieMi et al. [2] updated the blood pressure standards forChinese children issued in 2010. Tey also developedblood pressure reference standards for children basedon sex, age, and height [2] In recent years, the prevalence of hypertension among Chinese children and adolescents has been on the rise. Hypertension usually hasno obvious clinical manifestations. Tus, people oftenlack vigilance, leading to delayed treatment. Many children and adolescents with abnormal blood pressure continue into adulthood. Tus, hypertension in children and adolescents develops into adult hypertension, causingheart, brain, kidney and other target organ damage, aswell as atherosclerosis.

Over time, essential hypertension leads from structuralchanges in heart to systolic and diastolic dysfunctionseventually culminate in heart failure. Although heart failure is uncommon in children and adolescents with hypertension, it is important to assess the presence of earlychanges in diastolic and systolic function. Hypertensionin adults is often associated with the heart and other target organs. However, studies on target organ damage inchildren with essential hypertension are lacking.

Hypertension can also lead to changes in vascularstructure and function. Tere are multiple models available to assess vascular structure and function, which aredivided into three categories: vascular structure, arterialstifness, and endothelial function. Pulse wave velocity(PWV) is the most commonly measured non-imagingparameter to assess arterial stifness. Carotid intimal–medial thickness (cIMT) is the primary index for measuring vascular structure.

Tis study aimed to noninvasively evaluate changes incardiovascular structure and function in children withhypertension by ultrasonography to verifed the targetorgan damages.

Materials and methods

Subjects

We continously retrospectively studied 45 children (34simple hypertensive in group 1, 11 hypertensive co-existing with obesity in group 2) with newly diagnosed essential hypertension admitted in the cardiology departmentin the Children’s Hospital of Soochow University fromMarch 2020 to May 2021, and 32 healthy childrenmatched with age and sex were recruited from the community-based population as the control healthy group3. Clinical parameters (Table 1), including age, gender,body mass index (BMI; kg/m2), blood pressure (BP) andbiochemical data on lipids were obtained in all children.Hypertension was diagnosed on systolic and/or diastolicpressure≥95th percentile for sex, age, and height according to the reference value of the Chinese Child BloodPressure References Collaborative Group。[2] Obesitywas defned as BMI>95th percentile for age, sex, andheight.

Among the 45 hypertension patients, 8 patients wereadmitted with obese and 5 children were admitted withdizziness and chest tightness as the chief complaint. 31patients were admitted with high blood pressure on routine physical examination, and 1 patient was admitted tohospital for digestive tract foreign body. All the childrenwere newly diagnosed with essential hypertension andhad not been treated with drugs. Tey were excluded secondary hypertension, cardiomyopathy, and valvular heartdisease at the time of admission to hospital, meanwhilesubjects with hyperthyroidism or diabetes mellitus werealso excluded.

Method

Blood pressure measurement

Irritant drugs and food were not permitted before measurement. Patients were required to sit for 5–10 min in aquiet environment, with the cubital fossa and heart at thesame level. A standard clinical cuf sphygmomanometer

was used to measure blood pressure in the right upper arm.Te chest part of the membrane stethoscope was placed onthe medial side of the cubital fossa for brachial artery pulsation (2 cm above the elbow fossa). We took the K1 toneas the standard for systolic blood pressure and the K5 toneas the standard for diastolic blood pressure. We performedmeasurements twice continuously and averaged the twomeasurements. Te interval between each measurementwas 3 min. If the diference between the frst two readingswas>5 mmHg, we used the average value after obtaining athird reading. Te frst diagnosis of hypertension was verifed by a third examination.

Echocardiography

Measurement of cardiac structure

Echocardiography was performed using a Philips EPIQ7Ccolor Doppler ultrasound machine ( Koninklijke PhilipsUltrasound Inc., Netherland), equipped with S8-3, S5-1and L12-5 probes ranging from low to high frequency(3–6 MHz, 1.6–3.2 MHz and 4.4–8.8 MHz, respectively).According to American Society of Echocardiographyguidelines for pediatric echocardiogram from 2006 [3],we measured diastolic left ventricular internal diameter (LVIDd), diastolic interventricular septum thickness(IVSd), and diastolic left ventricular posterior wall thickness (LVPWd), diameter of aortic root (AO) and left atrialdiameter (LAD).

Measurement of cardiac diastolical function

Te Doppler spectrum of the mitral valve pulse wasrecorded in an apical four-chamber view. Te peak velocityof the flling peak in the early diastolic period (E) and thepeak velocity of the flling peak in the late diastolic period(A) were measured, and the E/A ratio was calculated. Treecardiac cycles were measured, and the average value wasused. We used the Tissue Doppler Imaging technique tomeasure the velocity of the mitral annulus (E’) and calculated the E/E’ value by combining this with the velocity ofearly diastolic blood fow (E) through the mitral valve orifce. Tis study defned left ventricular diastolic dysfunction as an E/E’>15 or an E/A<1.0.

Left ventricular remodeling

Te calculations of left ventricular mass (LVM), LVM index(LVMI), and relative wall thickness (RWT) were based onDevereux formulas:

LVM(g)=0.80× [1.04×(VSd+LVPWd+LVIDd³−LVIDd³)] +0.6.

LVMI(g/m^2.7)=LVM/height^2.7.

RWT=(IVSd+LVPWd)/LVIDd.

Tree cardiac cycles were measured, and the averagevalue was taken. Te criterion for left ventricular hypertrophy (LVH) was a LVMI>38.6 g/m2.7in children. AnRWT>0.41 was considered abnormal [4]. According toechocardiography, there are four possible left ventricularconfgurations: concentric hypertrophy, eccentric hypertrophy, concentric remodeling, and a normal confguration (as judged by LVMI and RWT).

Ultrasonography

Measurement of arterial stifness

Carotid–femoral PWV (cfPWV) was calculated as theratio of the travel distance (D) to carotid–femoral pulsetransit time (T). Te path length of the direct carotidfemoral distance segment (D) was estimated based on thesurface distance (Ds) from the carotid arteral measurement point to the femoral measuremental point using theformula: D=Ds×0.8 (a correction of the distance for afxed factor of 0.8 was applicated jn order to encount thesimultaneously traveling of the pulse wave from aorta tocarotid and to femoral arteries) [5]. Te so-called transit time (T) was the time of travel from carotid arteryto femoral artery of the wave over the distance. cfPWVwas calculated as follows: cfPWV=D (meters) / T (seconds). Each participant was examined in supine positionand the head slightly back while electrocardiography wasconnected synchronously. Te carotid artery point waslocated at a distance of 1.0–2.0 cm to the carotid bifurcation and marked on the body surface. Te process wasrepeated at the femoral artery point in the groin positionwith three repeated measurements. We use a caliper tomeasure the time from the ECG R wave peak to the startof the waveform as the wave transmission time (Fig. 1).Two measurements were obtained in each patient andthe mean was used for the analysis. Te acquisitions wereall performed by the same operator in a double-blindcondition to exclude inter-group diferences.

Measurement of vascular structure

Patients assumed a supine position to expose the neck.Ultrasonography of the carotid artery lumen was performed, and cIMT of the anterior and posterior walls wasmeasured. At 1–2 cm below the bifurcation plane, thethickness of the internal–medial carotid membrane were automatically measured by the built-in software package.Carotid distensibility (CD) can be measured by acquiring carotid artery diameter in end diastole and the

systo-diastolic change also called distension or the diameter changes of carotid artery (∆D). Te systolic and diastolic diameter of carotid artery were manually measured,then calculated ∆D. After that CD is calculated accordingto the reneman formula:

△D=Dd−Ds

△D%=(Dd−Ds) /Dd×100%

CD= [(2△D×Dd)+△D²]/PP×Dd².

where Dd is the end diastolic diameter of the vessel,Ds is the systolic diameter of the carotid artery. PP is thecentral pulse pressure [6].

Statistical analysis

We used SPSS 25.0 software (SPSS Inc., Chicago, IL) toperform statistical processing of the data. Measurementdata are expressed as mean±standard deviation. As wewere primarily interested in comparing both diseasegroups (hypertension and hypertension+obesity) withthe healthy control group, we performed one-way analysis of variance (ANOVA), where the disease groups wereconsidered mutually exclusive. All results are presentedfrom the one-way ANOVA. APvalue of<0.05 was considered statistically signifcant.

Te clinical characteristics and biochemical parametersof hypertension and control groups are shown in Table 1.Tere was no signifcant diference in age, sex, height,triglycerides (TG), total cholesterol (TC), high-densitylipoprotein cholesterol (HDL) and low-density lipoprotein cholesterol (LDL) among the three groups (P>0.05).Tere were signifcant diferences in body weight, BMI,systolic blood pressure (SBP), diastolic blood pressure (DBP) and pulse pressure (PP) in group 1 and group 2compared with group 3 (P<0.05).

Tere were signifcant diferences in LVM, LVMI, RWT,LVIDd, IVSd, LVPWd, LAD, A peak, E’ peak, A’ peakand E/E’ among the three groups (P<0.05). LVM, RWT,LVIDd, LVPWd, LAD and E/E’ in group 1 and group 2were signifcantly higher than the healthy control group.Te two hypertension groups were not signifcantly different from each other. IVSd and LVMI were signifcantly diferent among the three groups by S–N-K test.A peak and A’ peak were signifcantly higher in group 1compared with group 2 and group 3 in a post-hoc analysis. Moreover, E’ peak was signifcantly lower in group 2compared with group 1 and group 3. However, there wasno diference in AO, E peak and E/A ratio among thethree groups (Table 2).

A 0.31 cut-of value of RWT had 85.7% sensitivity and77.4% specifcity for hypertension patients. A cut-ofvalue of cfPWV is 4.55 m/s, and this value had 88.9% sensitivity and 53.6% specifcity for hypertension patients.2 of 11 children in HTN+Obese group were concentric hypertrophy (2/11, 18.1%), 2 children were eccentrichypertrophy (2/11, 18.1%) and 1 child were concentricremodeling (1/11, 9.0%), respectively. Of the 34 HTNpatients, 2, 6 and 5 children had concentric hypertrophy(2/34, 5.9%), eccentric hypertrophy(6/34, 17.6%) andconcentric remodeling(5/34, 14.7%), respectively.

Tere were signifcant diferences in cfPWV amongthe three groups. cfPWV was signifcantly higher in bothhypertension groups in comparison with the healthy control group (Fig. 2). Tere were no signifcant diferencesin cIMT, ∆D% and CD among three groups as shown inTable 3.

Discussion

In 2017, with the blood pressure standards for Chinesechildren by sex and age, and height developed, hypertension in children and adolescents aroused great concern inpediatrics. It is the most widely used diagnostic criteria.Our defnition of hypertension is based on this criterion,and our study found the structural changes in heart, cardiac diastolic dysfunction and vascular stifness.

There were signifcant diferences in LVM, LVMI, RWT,LVIDd, IVSd, LVPWd, LAD in hypertensive children. Anincrease in afterload caused by high blood pressure leadsto an increase in ventricular wall stress and eventuallychanges in myocardial structure, which is called left ventricular hypertrophy or remodeling. In patients with continuous hypertension in the systemic circulation, cardiacafterload increases, resulting in an increase in left ventricular end-systolic residual blood volume and diastolicvolume. Tis in turn results in left ventricular compensatory hypertrophy, a decrease in myocardial elasticity, and an increase in left ventricular volume. Ten an increasingin left atrial residual blood volume at the end of diastoleand gradual left atrial enlargement maintain an atrioventricular pressure diference and excessive cardiac output [7]. Left ventricular remodeling refers to a series ofchanges in the shape, structure, and function of the heartas hypertension progresses.

High blood pressure leads to an increase in myocardialmass, which is refected by an increase in LVIDd, IVSd,LVPWd of cardic morphology and structure. An increasein LVM is a manifestation of left ventricular hypertrophy(LVH) and an independent risk factor for LVH [8]. Usingthe Devereux formula to calculate LVM and correctingLVM for height to obtain LVMI is a common method tojudge whether or not LVH exists. Our study confrmedthat LVMI is signifcantly higher in children with hypertension especially obesity than healthy children just likethe other published litertures.

RWT is also an important indicator of cardiovascular remodeling in hypertension. Although the RWT ofhypertensive groups were higher than that of the controlgroup, their values were all within the normal range inour study. A 0.31 cut-of value had 85.7% sensitivity and77.4% specifcity for hypertension patients while a 0.41cut-of value only had 21.4% sensitivity and 100% specifcity. We should establish a new cut-of value for RWT inchildren with hypertension.

Among children with 45 hypertension, 12 (12/45,26.7%) ones had LVH and 6 children(13.3%) had leftventricular remodeling. Te incidence of LVH was consistent with the 27% reported by Sorof [9], so LVMI andRWT can be used as clinical evidence of target organdamage in children with hypertension. Although 60%of patients in this study showed a normal confguration,the rate of LVH was signifcantly higher compared withhealthy controls. Echocardiography to diagnose LVHand remodeling, to follow up in screening and to managehypertension is a viable strategy.

Studies had shown a positive correlation between highblood pressure and obesity in children and adolescents[10,11]. Te prevalence of hypertension in children withobesity is>10-times that of children with a normal BMI.Studies had confrmed that the earlier obesity occurs,the longer its duration, and the greater the likelihood ofhypertension. In our study, obese children with hypertension had higher diastolic blood pressure and moreimpaired left ventricular diastolic function. And otherindicators were consistent with those of children withhypertension. Te ratio of left ventricular hypertrophyand remodeling in group 2 was higher than that in group1. Both hypertension and obesity are considered independent risk factors for LVMI. To avoid errors in estimating the efects of being overweight LVMI was better than LVM in distinguishing the damage to the heart caused byhypertension.

In addition to changes in left ventricular structure,changes in left atrial structure and function have beenconfrmed as early pathological changes in patientswith hypertension [12]. In our study, there was signifcant diference in LAD between the three groups. Teresult shows that children of hypertension groups demonstrates left atrial enlargement compared with healthychildren, which is consistent with the results of Tsai [13].Keller et al.[14] also found that even if echocardiography showed no myocardial hypertrophy, LAD wouldincrease signifcantly. It is possible that left atrial enlargement exists before LVH, because the atrial muscle fbersare smaller and shorter than ventricular muscle fbers, sothey are also more sensitive to pressure.

Previous study defned left ventricular diastolic dysfunction as an E/E’ of>15 or an E/A of<1.0. We compared E/E’ ratio between the three groups and found thatalthough E/E’ in children with hypertension was<15,children in these groups had a signifcantly increasedE/E’ compared with the healthy group. Tis indicatedthat left ventricular diastolic function decreases in thegroups with hypertension. We found that the E’ peakof the ventricular septum in group 2 was signifcantlydecreased, and A’ peak of the ventricular septum and Apeak of mitral valve were higher in group 1 than the othertwo groups. Tese results proved that increased left atrialpressure and impaired left atrial function were commonin children with hypertension. Tere was no signifcantdiference in E/A ratio between the three groups but theE/E’ in children with hypertension whatever obese or notwas signifcantly increased. E/A probably is less sensitivethan E/E’ as previously reported literatures [15,16]...

In addition to the above cardiac damage, hypertensioncan also cause changes in vascular structure and function.Arterial stifness or arterial compliance can be measuredas PWV which was the speed of pulse wave transmitted across the length of the arterial tree. A higher PWVindicates a stifer blood vessel, contributing to increasedafterload and subsequent cardiac remodeling [17]. Inadults, a higher PWV predicts the risk of cardiovascular disease events, such as stroke, ischemic heart disease,and hypertension [18,19]. A meta-analysis of>15,000subjects confrmed that an increase in PWV of 1 m/sresults in an increase of 14% in the risk of cardiovascular events after adjustment for age, sex, and cardiovascular risk factors, and an increase of 15% in cardiovascularmortality [20]. Tere are a few studies on PWV on adolescent HTN. Kulsum-Mecci N et al. found adolescentaged 4 to 18 years of obesity and HTN both signifcantlyand independently increased PWV, while PWV increasedwith age but did not difer by race or sex [21].

Te predictors of PWV were age, SBP, heart rate, BMI,antihypertensive treatment and drug classes, not relatedto race, gender, smoking, dyslipidaemia, diabetes, kidneydisease, or genetic factors. Two factors were identifedas being responsible for accelerated progression of PWVafter correction for HR: age and BP values [22]. However, in our study, there was no diference in biochemical data on blood lipids, no use of antihypertensive drugs,and no diference in age and sex. Terefore, most of therelated predictors could be excluded and the infuence ofBP could be considered importantly[23]. Meanwhile, inour study, a relative increasing in cfPWV in HTN with orwithout obesity indicated that HTN had a greater impacton vascular stifness than obesity. Studies by some groupshad shown an increase in PWV (a decrease in arterialcompliance) with obesity alone [24,25]. But one limitation of our study was missing parameters of simple obesity group.

Doppler methods to assess cfPWV is feasible. It is notthe preferred one compared to the automatic one usingtonometey or pressure sensor, but it’s a good replacementto assess cfPWV when you don’t have the automaticmachine. Although Doppler-PWV measurements determined more evaluation error than ones of applanationtonometry, piezoelectric mechano-transducer or cufbased oscillometry, Doppler-PWV exactly showed closecorrelation with invasive assessment as well. StyczynskiG. et al. [26] found that mean invasive PWV was 9.38 m/sec and mean echo-PWV was 9.51 m/sec(P=0.78), thePearson’s correlation coefcient between methods was0.93 (P<0.0001), a Bland–Altman plot revealed a meandiference between invasive PWV and echo-PWV of0.13±0.79 m/sec. Doppler-PWV is a reliable method ofPWV measurement. Wider implementation of the Doppler-PWV method for the evaluation of aortic wall stifness can further expand the clinical and scientifc utilityof echocardiography [27].

In our study, absolute cfPWV values in the threegroups were much lower than the threshold for severecardiovascular events in adults (10 m/s), but both HTNgroups had signifcantly higher values than the healthychildren. We found that 4.55 m/s cut-of value of cfPWVhad 88.9% sensitivity and 53.6% specifcity for hypertension patients. Although 10 m/s cut-of value had gainedindependent predictive value for fatal and nonfatal cardiovascular events in hypertensive adult patients, thisthreshold did not distinguish HTN let alone stratify themin child and adolescents. We need more studies of vascular PWV in children to further determine the cut-ofvalue and reclassify intermediate risk patients into higheror lower target organ damage risk.

Tere were no signifcant diference in carotid IMTand CD between the three groups of children, probablybecause our sample size was too small to stratifed analysis based on age, weight and height. Carotid ultrasound measurement of IMT is the most commonlyused assessment of vascular structure, since it is considered to refect overall atherosclerotic burden. Currently, many machines had built-in software that canautomatically measure IMT to reduce manual measurement error. Kollias et al. [28] studied the relationshipbetween ambulatory BP and cIMT. Te mean cIMTin children with a higher BP was 0.03 mm larger thanthose with a normal BP in a meta-analysis. Day TGet al. [29] have found that a higher BP is associatedwith a higher cIMT, even after adjusting for cardiovascular risk factors, but they did not have a clear diagnostic criteria of BP for the observed efect. It shouldbe noted that cIMT measurements in children mayvary greatly between imaging protocols and machines,but absolute values are much lower than those associated with severe cardiovascular events in adults(usually > 1.0 mm).

Carotid distensibility is probably the most standardizedand used carotid arterial stifness index. Measuring CDmeans quantify changes in arterial diameter in responseto BP changes from diastole to systole. Like cIMT distensibility has also been related to all cause mortality, cardiovascular morbidity and mortality and the presence andseverity of cardiovascular diseases in adults [30]. It represents an evaluation of functional abnormalities beforestructural modifcation began. Elaine M et al. [31] foundthat prehypertensive youth had increasing arterial stifness and a graded increase in cIMT. But our study foundno diferences in CD and cIMT among the three groups.Te possible reasons were that the sample size was small,the inner diameter changes of carotid artery were measured manually not by automatic instrument, and thepulse pressure measured was non-central method et al.

Tere are some limitations of this study that should benoted. First, this was a retrospective study, thus selectionbias was inevitable. Second, the sample size was small,and the regional distribution was uneven, and becauseof the small sample size of children with obesity, we didnot evaluate the efects of obesity on cardiovascularstructure and function alone. Finally, a coefcient of variation of the mean value of the intra-session within- andbetween-operator variability respectively was missingand there was no agreement data between the Dopplertechnique and tonometry or piezoelectric mechanotransducer technique, as well as intra- / inter- reproducibility for our center when assessing cfPWV. In view ofthese limitations, follow-up studies should expand thesample size and increase the analysis of cardiovascularefects in individuals with obesity. Our future researchdirections include longitudinal studies to assess changes in cfPWV over time in healthy children and children withhypertension and obesity. We can do further consistencystudy between ultrasonographic and automatic tonometey methods.

In conclusion, although essential hypertension has noobvious clinical symptoms, Doppler ultrasonographyshowed that the heart and blood vessels of children withhypertension undergo structural and functional changes.Our data provide fundamental support for the argumentthat hypertension is having an efect on target organdamage of cardiovascular system in adolescents andyoung adults.

Conclusions

Doppler ultrasonography shows that the heart and bloodvessels of children with hypertension undergo structuraland functional target organ damages.

Acknowledgements

Not applicable.

Authors’ contributions

YYD, as the corresponding author, was the chief investigator. WL, CH, MH,QQX, HW, PPG, LS, and HTL were co-investigators. WL, CH, QQX, HW, and PPGcollected and managed ultrasonography data. MH, LS, and HTL collectedand analyzed clinical data. WL drafted the manuscript. HTL participated indata interpretation and manuscript drafting. All authors read and approvedthe fnal manuscript. WL and CH equally contribute to this study as frstco-authors.

Funding

This work was supported by grants from the National Natural Science Founda‑tion of China (Nos. 81870365 and 81970436), Young Scientists Fund of theNational Science Foundation of China (Nos. 81800437 and 81900450), andJiangsu Provincial Medical Young Talents (QNRC2016764).

Availability of data and materials

The data and material in the current study are available from the correspond‑ing author on reasonable request.

Declarations

Ethics approval and consent to participate

This retrospective study was approved by the Ethics Committee of theChildren’s Hospital of Soochow University with a waiver for written informedconsent obtained from each participant’s legal guardian/next of kin.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.Received: 15 May 2021 Accepted: 7 July 2021

Published online: 21 July 2021