Interpreting Z-Scores
By convention, a z-score of +2 or −2 represents a measurement that is 2 standard deviations above or below the mean value, usually recognized as the thresholds for normal in a pediatric population. The following classification applies to most cardiac structures:
- Z-score between −2 and +2: Normal range
- Z-score +2 to +3: Mildly dilated; warrants monitoring
- Z-score >+3: Moderately to severely dilated; requires follow-up and likely intervention
- Z-score ≥+10 (coronary arteries): Giant aneurysm (e.g., Kawasaki disease); definite intervention required
For aortic dilation specifically, the 2024 AHA scientific statement on pediatric aortopathy provides a severity grading scale: mild (z-score 2-3), moderate (z-score 3-4), severe (z-score 4-5), and very severe (z-score >5).
Body Surface Area: The Indexing Variable
BSA is the most widely accepted and predictive determinant of cardiovascular structure growth in children because of the known relationship between cardiovascular growth and cardiac output, and between cardiac output and total body size.
BSA Formulas
Multiple BSA formulas exist, and the choice matters:
Haycock formula: BSA = 0.024265 × height^0.3964 × weight^0.5378 (height in cm, weight in kg). This formula has been shown to correlate most strongly with cardiac structure sizes in both neonates and older children and is recommended by ASE guidelines.
Mosteller formula: BSA (m²) = √(height [cm] × weight [kg] / 3600). Widely used due to simplicity.
Du Bois formula: BSA = 0.007184 × height^0.725 × weight^0.425. Used in some coronary z-score systems.
The same BSA formula should be used consistently for serial measurements. Errors in height or weight measurement cascade through the z-score calculation and can lead to misclassification. Use the BSA Calculator to avoid manual math mistakes.
Available Z-Score Reference Systems
Multiple z-score models have been published, and because they use variable measurement techniques and statistical methodologies, there may be a wide range of z-scores for a particular measurement in the same patient depending on which model was used.
General Cardiac Structures
The most commonly used nomograms in the United States are from Boston Children's Hospital and the Pediatric Heart Network (PHN).
Pediatric Heart Network (Lopez, 2017): Derived from 3,215 healthy children from 19 centers across North America. This is the largest multicenter dataset with documented racial and ethnic diversity. Z-scores were found to be independent of age, sex, race, and ethnicity for each measurement.
Boston Children's Hospital: Originally based on 496 children but subsequently expanded to >2,000. Uses similar methodology to PHN and correlates highly with PHN z-scores (mean correlation 0.99).
Detroit/Pettersen (2008): Based on 782 children from a single center. Z-scores diverge from PHN at high body surface areas, possibly because there were fewer subjects in this category in the Detroit database.
Italy/Cantinotti: Based on 1,151 Caucasian children from a single center. Provides additional measurements not available in other systems.
A 2021 comparison study found that for most measurements, PHN z-score curves were similar to Boston and Italian curves but diverged from Detroit curves at high BSA. Despite excellent correlation between models, significant differences in z-scores were seen for many measurements. This is important when comparing publications using different models and for clinical care, particularly when z-score thresholds guide diagnosis and management.
The Echo Z-Score Calculator on Centilo supports both the PHN/Lopez and Pettersen/Detroit datasets, allowing side-by-side comparison for the same patient.
Coronary Artery Z-Scores
For coronary artery assessment (particularly in Kawasaki disease), the most rigorous systems are:
Dallaire (Canadian): Uses a square root function of BSA; provides normative data for the left circumflex branch. Employs Du Bois, Haycock, and Mosteller formulas.
Kobayashi (Japanese): Uses lambda-mu-sigma method for regression analysis of BSA.
Both systems perform equally well when applied across populations, though the Canadian system defines a higher proportion of abnormalities. The AHA risk stratification for Kawasaki disease is based on formulas from the National Heart, Lung, and Blood Institute Pediatric Heart Network. Use the Coronary Z-Score Calculator for coronary-specific assessment.
Neonates and Young Infants
Standard z-score systems may be less reliable at the extremes of BSA. A 2023 JACC study of >13,000 healthy newborns found that LV parameters measured in the first week of life differed so substantially from those measured in subsequent weeks that separate reference intervals were required for these two periods. For neonatal echocardiography, age-specific reference intervals should be used when available.
Emerging Systems
A 2025 Canadian study of >20,000 children developed z-score equations adjusted for body size, BMI, and age using generalized additive models (GAMLSS). These demonstrated less bias in subgroups of overweight, young, and early school-aged children compared to BSA-only models.
Limitations of Z-Scores
Obesity
BSA-based z-scores have systematic biases in overweight and obese children. A multicenter study found that existing BSA-based z-scores incompletely adjust for weight and BMI, leading to underestimation of >0.8 z-score units in subjects with higher BMI compared with lean subjects. This occurs because BSA-based models overestimate predicted dimensions with increasing weight or BMI.
Height-based normalization results in higher cardiovascular z-scores in heavier children, while BSA-based normalization results in higher z-scores in lighter children. Increasing BMI has opposite effects on height-based versus BSA-based z-scores. Models using height and weight as independent predictors may minimize residual associations with body size in children with abnormal body habitus.
Measurement Variability
A small error in measurement can translate into a significant difference in z-scores, changing classification, particularly in young patients. Interobserver variability for echocardiographic measurements is reported as ≥5% difference.
Extremes of Body Size
Z-scores from different models diverge at both extremes of BSA. The Detroit z-scores particularly diverge from PHN z-scores at high BSA. For neonates at the extremes of BSA, existing reference intervals often yield differing z-scores for the same measurement.
Clinical Applications
Kawasaki Disease
Coronary artery z-scores are central to risk stratification and management. The AHA classification system uses z-scores to categorize coronary involvement:
- No involvement: z-score always <2
- Dilation only: z-score 2 to <2.5
- Small aneurysm: z-score ≥2.5 to <5
- Medium aneurysm: z-score ≥5 to <10, or absolute dimension <8 mm
- Large/Giant aneurysm: z-score ≥10, or absolute dimension ≥8 mm
High-risk features warranting consideration of intensified initial therapy include baseline coronary artery z-score ≥2.5 and age <6 months. Contemporary studies show patients with small CAAs (z <5) have near-universal normalization and near-zero risk of adverse cardiac events, while those with z-score ≥10 remain at substantial risk.
Cardiomyopathy
Serial z-scores track disease progression. An LV z-score climbing from 2 to 4 over months signals the need to discuss transplant listing or mechanical support.
Aortopathy
The 2024 AHA scientific statement recommends z-score-based severity classification for aortic dilation, with intervention thresholds varying by underlying diagnosis (e.g., Marfan syndrome, bicuspid aortic valve).
Post-Surgical Surveillance
Trends matter more than single values. Remodeling back toward normal z-scores is reassuring; progressive increases push toward re-intervention.
Practical Recommendations
- Use the same z-score system consistently within your institution and for serial measurements in individual patients. The same nomogram should be used for serial measurements.
- Verify your echo system's reference set. Spot-check automated calculations against the Echo Z-Score Calculator occasionally; software errors occur.
- Ensure accurate anthropometrics. Inaccurate height or weight is a common source of z-score error.
- Report z-scores prominently. Include both the raw measurement and the z-score. The z-score should not be buried at the end of the report.
- Optimize measurement technique. Poor measurements yield poor z-scores. If a measurement looks borderline, obtain it in two planes.
- Consider obesity effects. In obese children, BSA-based z-scores may underestimate true abnormality. Clinical correlation is essential.
- Use age-appropriate references for neonates. Standard z-score systems may be less reliable in the first week of life.
- Interpret z-scores in clinical context. A z-score of 2.1 does not mandate intervention. Consider symptoms, chamber function, rate of change, and underlying diagnosis. A stable or normalizing z-score on serial echoes is reassuring; progressive increases push toward earlier intervention.
The z-score is a tool, not a decision. Clinical judgment remains essential.
References
- Lopez L, Saurers DL, Barker PCA, et al. Guidelines for Performing a Comprehensive Pediatric Transthoracic Echocardiogram: Recommendations From the American Society of Echocardiography. J Am Soc Echocardiogr. 2024;37(2):119-170.
- Lopez L, Colan SD, Frommelt PC, et al. Recommendations for Quantification Methods During the Performance of a Pediatric Echocardiogram. J Am Soc Echocardiogr. 2010;23(5):465-495.
- McCrindle BW, Rowley AH, Newburger JW, et al. Diagnosis, Treatment, and Long-Term Management of Kawasaki Disease: A Scientific Statement From the American Heart Association. Circulation. 2017;135(17):e927-e999.
- Morris SA, Flyer JN, Yetman AT, et al. Cardiovascular Management of Aortopathy in Children: A Scientific Statement From the American Heart Association. Circulation. 2024;150(11):e228-e254.
- Vøgg ROB, Sillesen AS, Wohlfahrt J, et al. Normative Echocardiographic Left Ventricular Parameters and Reference Intervals in Infants. J Am Coll Cardiol. 2023;81(22):2175-2185.
- Jone PN, Tremoulet A, Choueiter N, et al. Update on Diagnosis and Management of Kawasaki Disease: A Scientific Statement From the American Heart Association. Circulation. 2024;150(23):e481-e500.
- Lopez L, Frommelt PC, Colan SD, et al. Pediatric Heart Network Echocardiographic Z Scores: Comparison With Other Published Models. J Am Soc Echocardiogr. 2021;34(2):185-192.
- Lopez L, Colan S, Stylianou M, et al. Relationship of Echocardiographic Z Scores Adjusted for Body Surface Area to Age, Sex, Race, and Ethnicity: The Pediatric Heart Network Normal Echocardiogram Database. Circ Cardiovasc Imaging. 2017;10(11):e006979.
- Lauzon-Schnittka J, Plante V, Dahdah N, et al. Z Scores for Pediatric Echocardiography Dimensions Adjusted for Body Size, BMI, and Age. Circ Cardiovasc Imaging. 2025;:e017944.
- Plante V, Gobeil L, Xiong WT, et al. Alternative to Body Surface Area as a Solution to Correct Systematic Bias in Pediatric Echocardiography Z Scores. Can J Cardiol. 2021;37(11):1790-1797.
- Mahgerefteh J, Lai W, Colan S, et al. Height Versus Body Surface Area to Normalize Cardiovascular Measurements in Children Using the Pediatric Heart Network Echocardiographic Z-Score Database. Pediatr Cardiol. 2021;42(6):1284-1292.