

ORIGINAL RESEARCH 

Year : 2018  Volume
: 2
 Issue : 3  Page : 147154 

Estimation of Zscores of cardiac structures in healthy Indian pediatric population
Bhadra Trivedi^{1}, Manish Chokhandre^{1}, Poornima Dhobe^{2}, Swati Garekar^{1}
^{1} Department of Pediatric Cardiology, Fortis Child Heart Mission, Fortis Hospital, Mumbai, Maharashtra, India ^{2} Statistician, Independent Consultant, Mumbai, Maharashtra, India
Date of Web Publication  10Dec2018 
Correspondence Address: Dr. Bhadra Trivedi Department of Pediatric Cardiology, Fortis Child Heart Mission, Fortis Hospital, Mulund (w), Mumbai  400 078, Maharashtra India
Source of Support: None, Conflict of Interest: None
DOI: 10.4103/jiae.jiae_25_18
Introduction: Nomograms of pediatric cardiac structures are an effective tool to differentiate between normal and abnormal changes in dimensions of the heart. There is impending need for nomograms of ZScores of echocardiographic data derived from Indian children. Objective: The main objective of this study is to gather echocardiographic data from the healthy Indian pediatric population visiting the pediatric cardiology outpatient clinic and to derive the ZScores for various cardiac structures. Materials and Methods: All the echocardiographic studies from an eligible normal Indian population at a single centre were assessed. All the studies were performed on a single vendor echocardiography machine using weight appropriate neonatal, pediatric, and adult probes. Statistical Analysis: Body surface area (BSA) was used as an independent variable in a nonlinear regression analysis for the predicted mean value of each of the 19 echocardiographically measured structures. Results: The total number of children evaluated during the study period was 596, with age ranging from newborn to 16 years. The total parameters collected in the study were 8102. The correlation with Haycock's BSA and an individual parameter was found to be the most sensitive predictor of normal progression with age. Relationship of individual parameters with BSA was represented in the form of graphs. Conclusion: This study of normal Indian pediatric population is the largest Indian study to date. The regression formulae along with the graphs can be used to acquire the Z score of 19 individual echocardiographic parameters.
Keywords: Congenital heart disease, echocardiogram, nomograms, pediatric cardiology, ZScores
How to cite this article: Trivedi B, Chokhandre M, Dhobe P, Garekar S. Estimation of Zscores of cardiac structures in healthy Indian pediatric population. J Indian Acad Echocardiogr Cardiovasc Imaging 2018;2:14754 
How to cite this URL: Trivedi B, Chokhandre M, Dhobe P, Garekar S. Estimation of Zscores of cardiac structures in healthy Indian pediatric population. J Indian Acad Echocardiogr Cardiovasc Imaging [serial online] 2018 [cited 2020 Aug 3];2:14754. Available from: http://www.jiaecho.org/text.asp?2018/2/3/147/247027 
Introduction   
The echocardiogram forms the cornerstone for pediatric cardiac evaluation, an important aspect of which is a quantification of cardiac structure in terms of dimensions. The dimensions of a child's heart structure are affected by his or her hemodynamics and somatic growth.^{[1],[2]} Nomograms are an effective tool to differentiate between normal and abnormal changes in dimensions of the heart. Most published nomograms are derived from the Caucasian population.^{[3],[4],[5]} These have been included in the recent American Society of Echocardiography and the European Association of Echocardiography guidelines.^{[6]}
The Indian population has so far been evaluated by ZScores derived from such studies and guidelines. It has been proven that the Caucasian anthropometric data are not representative of Indian children.^{[7]} In this context, growth charts developed by the Indian Academy of Pediatrics^{[8]} are widely used in the Indian subcontinent. The prevalence of the congenital heart disease (CHD) in India is not well studied but has estimated to be between 4.2/1000^{[9]} and 19.1/1000 newborns.^{[10]} Thus, there is an impending need for nomograms of ZScores of echocardiographic data derived from Indian children.
Objective
The objective of this study is to gather the echocardiographic data from the healthy pediatric population visiting the pediatric cardiology outpatient clinic situated in a tertiary care center in the metropolis of Mumbai and to derive the ZScores for various cardiac structures inclusive of the aortic arch.
Materials and Methods   
All the stored echocardiographic studies of a normal echocardiogram from the eligible Indian population at the Fortis Hospital Mulund, Mumbai, between January 2014 and August 2016 were studied. The study was approved by the Institute Review Board.
Echocardiographic examination
All the echocardiographic studies were performed on a Philips iE33 Machine (Philips Medical Systems, Bothell, WA) using weight appropriate neonatal, pediatric, and adult probes, which include S12–S4, S8–S3, or S5–S1 MHz sector array transducers. All the echocardiograms were performed by single echocardiographer. All echocardiograms in the laboratory followed a rigorous quantitative protocol.^{[2]} All echocardiograms were electrocardiography gated, with a minimum of 2 heartbeats recorded per loop. All measurements were made live (online) as they are considered essential elements of our report. Infants were sedated with oral trichlophos if required. The echocardiography planes were standard and the twodimensional and Mmode measurements were made following prescribed methods. Measurements of aortic arch were taken as described by Kaine et al.^{[11]} Echocardiography data were stored in digital imaging and communications in medicine format, on DVDs. Measurements were stored electronically after stripping off personal identifiers.
Inclusion and exclusion criteria
The details of all the children visited the pediatric cardiology outpatient department (OPD) were reviewed. Only those children aged 0 days to 16 years with a normal echocardiogram report, normal physical examination and clinical history and absent systemic illness were included in the study. Normal echocardiograms included those with a tiny patent ductus arteriosus and/or a patent foramen ovale. Exclusion criteria were presence or suspicion of a syndrome, systemic illness, any structural heart defect (other than that mentioned above), and any hemodynamic abnormality like systemic or pulmonary hypertension. Incomplete studies in terms of nonstandard acquisitions were excluded. The protocol for obtaining measurements is detailed in [Table 1] and [Figure 1].  Table 1: Details about measurement of an individual parameter employed in the study
Click here to view 
 Figure 1: Location of echocardiographic measurements of the parameters. LV: Left ventricle, RV: Right ventricle, RPA: Right pulmonary artery, Ao Ann: Aortic annulus, STJ: ST junction, A Ao: Ascending aorta, TV: Tricuspid valve, MV: Mitral valve
Click here to view 
Statistical analysis
The parameters were indexed with the body surface area (BSA). There have been various studies which have established the relationship between growing cardiac structures and height, weight, BSA of an individual.^{[12],[13],[14],[15],[16]}
Body surface area was used as an independent variable in a nonlinear regression analysis for the predicted mean value of each of the 19 echocardiographically measured structures. Because of the problem of heterogeneous variances with these measurements across the range of BSA, a logarithmic transformation using the natural logarithm was performed on 19 structures before the fitting of the regression models for the purpose of stabilizing the variances.
A polynomial model of the third power was selected as the final regression model. Studentized error residuals were used to detect the outliers to be excluded from analysis.
The first step was to transform the measurement of each structure by computing its natural logarithm (ln), which is represented as:
y = ln(measurement)(1)
The transformed echocardiographic measurements were entered into a nonlinear regression model as the dependent variable and BSA, BSA2, BSA3 as predictors (independent variables).
Expectedy = β_{0}+β_{1}* BSA+β_{2}* BSA^{2}+ β_{3}* BSA^{3}(2)
Once the regression coefficients (β0, β1, β2, and β3) were obtained, the equation (2) was then transformed back to measurement's original scale by exponentiating y as explained in the following equation (3).
The Mean Y obtained from the equation (3) is used to standardize the observed value of the parameter using the equation number (4) as shown below, which in turn will produce Z score as a final product.
To plot a graph between echocardiographic measurements and BSA, the regression equations were considered in original units and 6 curves corresponding to the z score = ±1, ±2, ±3 were plotted for each echocardiographic measurement.
Data analysis were performed using SAS version 9.1 (SAS Institute, Cary, NC, USA).
Results   
The total number of children evaluated during the study period was 596, with age ranging from newborn to 16 years. A maximum number of children were within their 1^{st} year (n = 349), followed by 1–5 years of age (n = 119). The total parameters collected in the study are 8102. The distribution is shown in [Table 2]. [Table 3] and [Table 4] show the distribution of the study population in relation to weight and height respectively.
Out of the total 596 study participants, there was a slight preponderance of males (59%). The most common reason for referral among all age groups was the detection of a murmur, which was almost 50%. The other common referral reasons were apparent cyanosis and tachycardia.
The relation between each parameter and the BSA, height, age, and weight was studied using the regression equations. Among the various possibilities, the correlation with Haycock's BSA and an individual parameter was found to be the most sensitive predictor of normal progression with age. The coefficients for each equation are given in [Table 5].
R^{2} is the standard statistic used to measure how well data fits the selected regression model (goodness of fit) and has a value ranging from 0 to 1, with 1 representing a perfect fit and 0 representing a total lack of fit. In [Table 5] R^{2} values for most of the parameters are above 0.60 which indicates that there is a strong relationship between BSA and the measurement. The interventricular septum thickness varies throughout the cardiac cycle. The errors in accurate measurement of interventricular septal thickness (in systole and diastole) may be a contributing factor for lower R^{2} value. The lower R^{2} value of the aortic isthmus can be attributed to the less number of the observations.
[Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10], [Figure 11], [Figure 12], [Figure 13], [Figure 14], [Figure 15], [Figure 16], [Figure 17], [Figure 18], [Figure 19], [Figure 20] show the graphical presentation of the relationship of individual parameters with BSA. The graphs also show estimated regression lines corresponding to the ZScores of 0 (black line), while the red and green lines correspond to a Z score of ± 1 and 2, respectively. The graphs representing the cardiac valves, the branch pulmonary arteries, left ventricle internal dimensions, intersinus distance, sinotubular junction, ascending aorta, and the distal transverse arch is more “compact” as compared to other parameters. This is due to higher regression coefficient of these parameters.  Figure 2: Scatter plot of the mitral valve annulus (mm) versus body surface area (m^{2})
Click here to view 
 Figure 3: Scatter plot of the tricuspid valve annulus (mm) versus body surface area (m^{2})
Click here to view 
 Figure 4: Scatter plot of the aortic valve annulus (mm) versus body surface area (m^{2})
Click here to view 
 Figure 5: Scatter plot of the pulmonary valve annulus (mm) versus body surface area (m^{2})
Click here to view 
 Figure 6: Scatter plot of the right pulmonary artery (mm) versus body surface area (m^{2})
Click here to view 
 Figure 7: Scatter plot of the left pulmonary artery (mm) versus body surface area (m^{2})
Click here to view 
 Figure 8: Scatter plot of the interventricular septal thickness in diastole (mm) versus body surface area (m^{2})
Click here to view 
 Figure 9: Scatter plot of the interventricular septal thickness – systole (mm) versus body surface area (m^{2})
Click here to view 
 Figure 10: Scatter plot of the left ventricle internal diameter – diastole (mm) versus body surface area (m^{2})
Click here to view 
 Figure 11: Scatter plot of the left ventricle internal dimension – systole (mm) versus body surface area (m^{2})
Click here to view 
 Figure 12: Scatter plot of the posterior thickness in diastole (mm) versus body surface area (m^{2})
Click here to view 
 Figure 13: Scatter plot of the posterior wall thickness in systole (mm) versus body surface area (m^{2})
Click here to view 
 Figure 14: Scatter plot of the intersinus distance (mm) versus body surface area (m^{2})
Click here to view 
 Figure 15: Scatter plot of the sinotubular junction (mm) versus body surface area (m^{2})
Click here to view 
 Figure 16: Scatter plot of the diameter of ascending aorta (mm) versus body surface area (m^{2})
Click here to view 
 Figure 17: Scatter plot of the diameter of proximal transverse arch (mm) versus body surface area (m^{2})
Click here to view 
 Figure 18: Scatter plot of the diameter of the distal transverse arch (mm) versus body surface area (m^{2})
Click here to view 
 Figure 19: Scatter plot of the diameter of the isthmus (mm) versus body surface area (m^{2})
Click here to view 
 Figure 20: Scatter plot of the diameter of the descending aorta versus body surface area (m^{2})
Click here to view 
How to calculate ZScores
The Z score for an individual parameter can be calculated in two different ways.
First is using the statistical equations attached to the study.
Example with our data:
Weight: 8 Kg; Height: 110 cm; BSA: 0.5, Measured mitral valve annulus – 15 mm.
For mitral valve: β0 = 1.875, β1 = 2.865, β2 = 2.29, and β3 = 0.684 and insert these values in equation (2) to obtain the mean for mitral valve annulus size for BSA = 0.5.
Mean y = 1.875 + 2.865 × 0.5–2.29 × 0.52 + 0.684 × 0.53
= 2.821
Next, take the natural log of the observed mitral valve annulus size (15 mm) of a patient using equation (1) and standardize it using the following formula:
= − 0.68
The formulae for calculation of ZScores (of all 19 parameters) using this study are given in Appendix 1 (attached as an addon file).
The second method employs the graphs. Plotting the individual parameter on the graph against the BSA will give an estimated Z score. [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10], [Figure 11], [Figure 12], [Figure 13], [Figure 14], [Figure 15], [Figure 16], [Figure 17], [Figure 18], [Figure 19], [Figure 20] show the scatter plot of the measurement of an individual parameter versus BSA.
Discussion   
The measurements of cardiac structures and their comparison with nomograms are essential for preoperative planning for most congenital heart defects^{[17],[18],[19]}ZScores are essential to monitor the disease progression for the management of various acquired heart diseases such as Kawasaki disease or rheumatic heart disease.^{[2],[20]} Major pediatric cardiac centers across the world have developed their own nomograms.^{[21],[22],[23]}
The Z scores of cardiac structures of the Indian pediatric population remains a challenge. The only Indian study is a recent one that excluded the population below 4 years.^{[24]} Our study, on the other hand, focuses on the younger population with the majority (74%) <4 years and including 132 neonates. With the majority of CHD being detecting at the preschool age group, our study is likely to be of more relevance.
The demographic profile of our study in terms of the number of neonates we studied, matches a study by Tacy et al. However, this study included only neonates (n = 129) and reported on atrioventricular valve annuli sizes exclusively.^{[25]}
The mean BSA of Indian versus Caucasian children in various age groups differed significantly. This agrees with the reported demographic differences between normal Indian and Caucasian children and has been attributed to racial differences.^{[26]} The nomograms, therefore, should be derived from regional population and ideally, one cannot be substituted for another.
It is imperative to note that nomograms for cardiac structures in the pediatric population are derived from regression formulae which are unique to each study's demographic profile. Because of this, the Z scores derived from the different studies have variations between them.^{[27]} Our study has 177 children under the age of 3 months (BSA of 0.25 m^{2}); the number being significantly higher than in the Detroit study (n = 82).^{[22]} The data spread provides a more accurate estimation of the ZScores for the age and the BSA group in discussion.
Rogé et al. showed that the either of height, weight, BSA, or cube root of weight can be used for regression as they are all strongly associated with the growth of cardiac structure.^{[28]} BSA was chosen here to make the results comparable to various published studies on the same subject.
There are no representative data of Indian children for aortic root and arch measurements. It has been observed that the Western nomograms tend to underestimate the Z score of Indian children and overestimate the severity of the arch hypoplasia, which does not concur with intraoperative findings. This study is a first attempt to derive root and arch nomograms for Indian children in an age <4 years
Conclusion   
This study of normal Indian pediatric population is the largest to date. The regression formulae along with the graphs can be used to acquire the ZScores of 19 individual echocardiographic parameters. The ready to use calculator is provided in the online version of this article.
Limitations
As with other studies reporting ZScores in pediatric echocardiography, the z scores are derived from extrapolations using regression formulae rather than an actual measurement. Only those children with a normal echocardiogram, history, and physical examination who visited pediatric cardiology OPD were included, this could be potential for selection bias. Any children with preexisting illness or syndromes were excluded as mentioned earlier.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References   
1.  Gutgesell HP, Rembold CM. Growth of the human heart relative to body surface area. Am J Cardiol 1990;65:6628. 
2.  Lopez L, Colan SD, Frommelt PC, Ensing GJ, Kendall K, Younoszai AK, et al. Recommendations for quantification methods during the performance of a pediatric echocardiogram: A report from the pediatric measurements writing group of the American Society of Echocardiography Pediatric and Congenital Heart Disease Council. J Am Soc Echocardiogr 2010;23:46595. 
3.  Roman MJ, Devereux RB, KramerFox R, O'Loughlin J. Twodimensional echocardiographic aortic root dimensions in normal children and adults. Am J Cardiol 1989;64:50712. 
4.  Daubeney PE, Blackstone EH, Weintraub RG, Slavik Z, Scanlon J, Webber SA, et al. Relationship of the dimension of cardiac structures to body size: An echocardiographic study in normal infants and children. Cardiol Young 1999;9:40210. 
5.  First T, Skovránek J, Marek J. Normal values of 2dimensional echocardiographic parameters in children. Cesk Pediatr 1992;47:2604. 
6.  Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al. Recommendations for chamber quantification: A report from the american society of echocardiography's guidelines and standards committee and the chamber quantification writing group, developed in conjunction with the European association of echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18:144063. 
7.  Tarozzi A. Growth reference charts and the nutritional status of Indian children. Econ Hum Biol 2008;6:45568. 
8.  Khadilkar VV, Khadilkar AV. Revised Indian academy of pediatrics 2015 growth charts for height, weight and body mass index for 518yearold Indian children. Indian J Endocrinol Metab 2015;19:4706. 
9.  Chadha SL, Singh N, Shukla DK. Epidemiological study of congenital heart disease. Indian J Pediatr 2001;68:50710. 
10.  Bhardwaj R, Rai SK, Yadav AK, Lakhotia S, Agrawal D, Kumar A, et al. Epidemiology of congenital heart disease in India. Congenit Heart Dis 2015;10:43746. 
11.  Kaine SF, Smith EO, Mott AR, Mullins CE, Geva T. Quantitative echocardiographic analysis of the aortic arch predicts outcome of balloon angioplasty of native coarctation of the aorta. Circulation 1996;94:105662. 
12.  Epstein ML, Goldberg SJ, Allen HD, Konecke L, Wood J. Great vessel, cardiac chamber, and wall growth patterns in normal children. Circulation 1975;51:11249. 
13.  Sheil ML, Jenkins O, Sholler GF. Echocardiographic assessment of aortic root dimensions in normal children based on measurement of a new ratio of aortic size independent of growth. Am J Cardiol 1995;75:7115. 
14.  King DH, Smith EO, Huhta JC, Gutgesell HP. Mitral and tricuspid valve anular diameter in normal children determined by twodimensional echocardiography. Am J Cardiol 1985;55:7879. 
15.  Snider AR, Enderlein MA, Teitel DF, Juster RP. Twodimensional echocardiographic determination of aortic and pulmonary artery sizes from infancy to adulthood in normal subjects. Am J Cardiol 1984;53:21824. 
16.  McComb BL, Munden RF, Duan F, Jain AA, Tuite C, Chiles C, et al. Normative reference values of thoracic aortic diameter in American College of Radiology imaging network (ACRIN 6654) arm of national lung screening trial. Clin Imaging 2016;40:93643. 
17.  Awori MN, Leong W, Artrip JH, O'Donnell C. Tetralogy of fallot repair: Optimal z score use for transannular patch insertion. Eur J Cardiothorac Surg 2013;43:4836. 
18.  Stewart RD, Backer CL, Young L, Mavroudis C. Tetralogy of fallot: Results of a pulmonary valvesparing strategy. Ann Thorac Surg 2005;80:14318. 
19.  Schwartz ML, Gauvreau K, Geva T. Predictors of outcome of biventricular repair in infants with multiple left heart obstructive lesions. Circulation 2001;104:6827. 
20.  Chubb H, Simpson JM. The use of ZScores in paediatric cardiology. Ann Pediatr Cardiol 2012;5:17984. 
21.  Roberson DA, Cui W, Chen Z, Madronero LF, Cuneo BF. Annular and septal Doppler tissue imaging in children: Normal z score tables and effects of age, heart rate, and body surface area. J Am Soc Echocardiogr 2007;20:127684. 
22.  Pettersen MD, Du W, Skeens ME, Humes RA. Regression equations for calculation of z scores of cardiac structures in a large cohort of healthy infants, children, and adolescents: An echocardiographic study. J Am Soc Echocardiogr 2008;21:92234. 
23.  Cantinotti M, Scalese M, Murzi B, Assanta N, Spadoni I, De Lucia V, et al. Echocardiographic nomograms for chamber diameters and areas in caucasian children. J Am Soc Echocardiogr 2014;27:12799200.e2. 
24.  Gokhroo RK, Anantharaj A, Bisht D, Kishor K, Plakkal N, Aghoram R, et al. A pediatric echocardiographic Z score nomogram for a developing country: Indian pediatric echocardiography study – The Z score. Ann Pediatr Cardiol 2017;10:318. 
25.  Tacy TA, Vermilion RP, Ludomirsky A. Range of normal valve annulus size in neonates. Am J Cardiol 1995;75:5413. 
26.  Agarwal DK, Agarwal KN. Physical growth in Indian affluent children (birth6 years). Indian Pediatr 1994;31:377413. 
27.  Daubeney PEF, Delany DJ, Anderson RH, Sandor GGS, Slavik Z, Keeton BR, et al. Pulmonary atresia with intact ventricular septum: range of morphology in a populationbased study. J Am Coll Cardiol 2002 May 15;39 (10):1670–9. 
28.  Rogé CL, Silverman NH, Hart PA, Ray RM. Cardiac structure growth pattern determined by echocardiography. Circulation 1978;57:28590. 
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10], [Figure 11], [Figure 12], [Figure 13], [Figure 14], [Figure 15], [Figure 16], [Figure 17], [Figure 18], [Figure 19], [Figure 20]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]
