Structure-specific variation in pediatric echocardiographic Z scores across the nutritional spectrum: four body-surface-area nomograms and a BMI-adjusted GAMLSS model

Q: What are the key findings of this study?

Higher BMI Z-scores were inversely associated with BSA-based Z-scores in left-sided structures, especially the aorta (β -0.325 to -0.182, p<0.001), a bias largely abolished by a BMI-adjusted GAMLSS model.

Q: What is the clinical evidence from this study?

Study design: Observational. Population: Structurally normal pediatric hearts across the nutritional spectrum (including overweight and obesity) (n=1883). Intervention: Body mass index Z score (BMIz) / Nutritional status (Overweight and Obesity) vs. Eutrophic status (BMIz -1 to +1) and comparison across four BSA-based nomograms vs a BMI-adjusted GAMLSS model. Primary outcome: Association between cardiac structure Z scores (specifically at the four aortic levels) and BMI Z score (BMIz) across BSA-based nomograms (β -0.325 to -0.182, p=<0.001).

Q: What does this research mean for the field?

Higher BMI Z-scores are inversely associated with BSA-based echocardiographic Z-scores in pediatric left-sided structures, a bias that is largely abolished by a BMI-adjusted GAMLSS model. Novelty: ClaimNovelty.INCREMENTAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

Q: What question did this study set out to answer?

To determine whether pediatric echocardiographic Z scores vary with body mass index Z score across four body-surface-area nomograms, and to assess whether a BMI-adjusted model mitigates this bias.

Rodolfo Amstein; Valentina Metcalfe; Claudia Trincado G.; Álvaro Passi‐Solar; Guillermo Larios G.

doi:10.21203/rs.3.rs-10963763/v1

November 4, 2026Open Access

Structure-specific variation in pediatric echocardiographic Z scores across the nutritional spectrum: four body-surface-area nomograms and a BMI-adjusted GAMLSS model

View Full Paper

Ask AI

Bookmark

Share

Key result

Higher BMI Z-scores skew BSA-based aortic dimensions lower, a bias eliminated by BMI-adjusted GAMLSS modeling.

β -0.325 to -0.182
P<0.001
n=1,883

Why the study?

Standard pediatric echocardiographic Z score nomograms derive mostly from eutrophic populations, leaving it unclear whether cardiac structure Z scores vary with body mass index across body-surface-area-based nomograms and whether a BMI-adjusted reference mitigates this bias.

Does a BMI-adjusted GAMLSS echocardiographic Z score reference mitigate BMI-related bias in cardiac structure Z scores compared to standard BSA-based nomograms in children and adolescents aged 5–18 years across the nutritional spectrum?

Key Points

To determine whether pediatric echocardiographic Z scores vary with body mass index Z score across four body-surface-area nomograms, and to assess whether a BMI-adjusted model mitigates this bias.
Retrospective study of 1,883 children aged 5–18 years (55.1% male) with structurally normal echocardiograms evaluated between 2016 and 2021.
Assessed 11 cardiac structure diameters using four body-surface-area nomograms (Pediatric Heart Network, Detroit, Toronto, Italy) and a BMI-adjusted GAMLSS model for 10 structures.
Tested linear regression associations between cardiac Z scores and BMI Z score across three nested models with false discovery rate correction.
Absolute diameters of all structures increased with BMIz (p<0.001), but Z scores for left-sided structures showed strong inverse associations, most markedly at all four aortic levels (β −0.325 to −0.182 across all 45 comparisons) and the left ventricular end-diastolic diameter (β −0.219 to −0.154).
Ventricular walls showed variable associations across nomograms, while right-sided annuli varied non-monotonically, with mean Z scores peaking in eutrophic or overweight groups before dropping in obesity.
The GAMLSS model eliminated negative associations for the aortic valve, aortic root, ascending aorta, mitral valve, and LVEDD (|β| ≤ 0.062), but bias persisted in structures lacking BMI in their model formulation and reversed to positive in the pulmonary valve annulus (β +0.128 to +0.155).

Study Design

Type

Observational (n=1,883)

Multicenter

No

Structured PICO

Does a BMI-adjusted GAMLSS echocardiographic Z score reference mitigate BMI-related bias in cardiac structure Z scores compared to standard BSA-based nomograms in children and adolescents aged 5–18 years across the nutritional spectrum?

P

Population

1,883 children and adolescents aged 5–18 years (mean age 10.3 years, 44.9% female) with structurally normal echocardiograms spanning the nutritional spectrum (25.8% overweight, 15.8% obese) evaluated at a single center in Chile between 2016 and 2021.

E

Exposure

Echocardiographic Z score computation using a BMI-adjusted reference based on generalized additive models for location, scale and shape (GAMLSS; Lauzon-Schnittka 2025 nomogram) adjusted for height, weight, age, and body mass index Z score (BMIz) across ten cardiac structures

C

Comparator

Four standard body-surface-area (BSA)-based echocardiographic Z score nomograms (Pediatric Heart Network [PHN], Detroit, Toronto, and the Italian nomogram of Cantinotti et al.) applied across eleven cardiac structures

O

Outcome

Association between cardiac structure Z scores and body mass index Z score (BMIz) across eleven cardiac structures (left ventricular end-diastolic diameter, septal wall, posterior wall, left atrium, mitral valve annulus, aortic valve annulus, tricuspid valve annulus, pulmonary valve annulus, aortic root at the sinuses of Valsalva, sinotubular junction, and ascending aorta) assessed by standardized linear regression β coefficientssurrogate

Main Result

Effect estimate: β -0.325 to -0.182

p-value: p=<0.001

Standard BSA-based pediatric echocardiographic Z score nomograms systematically underestimate left-sided cardiac structure dimensions in children with overweight and obesity, a bias that is largely eliminated in aortic and chamber dimensions by using a BMI-adjusted GAMLSS reference.

Limitations

Retrospective, single-center design
Body composition (lean vs fat mass) was not directly measured
Reproducibility and inter-observer measurement variability were not assessed
The Toronto nomogram is from an unpublished dataset, limiting characterization of its methodological details
Pubertal stage and physical activity were not assessed, leaving potential for residual confounding
Body composition was not measured directly
Reproducibility was not assessed despite evidence of greater measurement variability in obesity
The Toronto nomogram is unpublished, preventing full characterization of methodological details
Pubertal stage and physical activity were not assessed, so residual confounding cannot be excluded

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

Standard BSA nomograms may underestimate left-sided Z-scores in obese children; supports BMI-adjusted GAMLSS models but leaves validation to prospective studies.

Abstract

Purpose Z score nomograms, the standard for interpreting pediatric echocardiographic dimensions, derive mostly from eutrophic populations. We assessed whether cardiac structure Z scores vary with body mass index Z score (BMIz) across four body-surface-area (BSA)-based nomograms, and whether a BMI-adjusted reference mitigates this bias. Methods Retrospective study of 1883 children aged 5–18 years with structurally normal echocardiograms (2016–2021). Eleven cardiac structure diameters were analyzed and Z scores computed with four nomograms (Pediatric Heart Network, Detroit, Toronto, Italy) and, for ten structures, with a fifth based on generalized additive models for location, scale and shape (GAMLSS) adjusted for body size, BMI and age. Associations with BMIz used linear regression in three nested models with false discovery rate correction. Results Overall, 25.8% were overweight and 15.8% had obesity. Absolute diameters increased with BMIz in every structure (p < 0.001), yet Z scores of most left-sided structures were inversely associated with it, most strongly for the four aortic levels (β − 0.325 to − 0.182, significant in all 45 nomogram-by-model comparisons) and for left ventricular end-diastolic diameter. Ventricular walls showed no consistent pattern; right-sided annuli varied non-monotonically. Under GAMLSS these associations were abolished in the most affected structures, persisted in the sinotubular junction and walls, and reversed in sign in the pulmonary annulus. Conclusion Z scores of most left-sided structures were lower in obesity than their absolute dimensions suggest, so interpretation must account for nutritional status and the reference. Persisting bias follows model construction: the three structures whose GAMLSS equations omit BMI are exactly those still biased.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Amstein et al. (2026) conducted an observational in Structurally normal pediatric hearts across the nutritional spectrum (including overweight and obesity) (n=1,883). Body mass index Z score (BMIz) / Nutritional status (Overweight and Obesity) vs. Eutrophic status (BMIz -1 to +1) and comparison across four BSA-based nomograms vs a BMI-adjusted GAMLSS model was evaluated on Association between cardiac structure Z scores (specifically at the four aortic levels) and BMI Z score (BMIz) across BSA-based nomograms (β -0.325 to -0.182, p=<0.001). Higher BMI Z-scores were inversely associated with BSA-based Z-scores in left-sided structures, especially the aorta (β -0.325 to -0.182, p<0.001), a bias largely abolished by a BMI-adjusted GAMLSS model.

synapsesocial.com/papers/6aed6e573dd5ffb8e184d017 https://doi.org/10.21203/rs.3.rs-10963763/v1

View Full Paper

Ask AI

Bookmark

Share

Structure-specific variation in pediatric echocardiographic Z scores across the nutritional spectrum: four body-surface-area nomograms and a BMI-adjusted GAMLSS model

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Structure-specific variation in pediatric echocardiographic Z scores across the nutritional spectrum: four body-surface-area nomograms and a BMI-adjusted GAMLSS model

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading