Distinct Peri-pubertal Growth Phenotypes Among Vietnamese Urban Children: A Functional Clustering Analysis of 2018-2025 School Health Data

Background

Child growth is usually summarized by an average trajectory, an approach that can hide clinically meaningful heterogeneity between children. This study aims to identify various growth phenotypes in Vietnamese children.

Methods

This study analyzed retrospective longitudinal anthropometric data from annual school health check of children in Hanoi, Hochiminh, and Haiphong, Vietnam from 2018 to 2025. Children with at least 4 visits spanning at least 3 years were included (8,612 boys and 6,107 girls). Height for age trajectory was summarized with functional principal component scores and combined with SITAR derived random effects for growth size, pubertal timing, and tempo, then clustered with sex specific Gaussian mixture models. We evaluated phenotype associations with prepubertal BMI category, cluster stability, and whether phenotype improved prediction of late adolescent height beyond baseline prepubertal variables.

Results

Ten growth phenotypes in boys and twelve in girls were identified. Phenotype membership was associated with prepubertal BMI category in both sexes. A tall, accelerated growth phenotype in boys (4.2% of boys) had the highest prepubertal obesity prevalence (53.5%) and strongest obesity association in external validation. The girls’ phenotype solution was more sensitive to visit based eligibility criteria than the boys’ solution. Adding phenotype to a baseline model improved prediction of late adolescent height in both sexes, more strongly in boys.

Conclusions

Combining functional and SITAR based clustering identified clinically distinguishable growth phenotypes in Vietnamese children linked to prepubertal BMI and late adolescent height. Sex specific validation is needed before these phenotypes inform clinical application.

Key messages

• Combining functional and SITAR based clustering identified 10 and 12 distinct growth phenotypes in Vietnamese boys and girls, linked to prepubertal BMI and improved prediction of late adolescent height.

• A tall, accelerated growth phenotype in boys carried the highest obesity prevalence in the cohort.

• Girls’ phenotypes were less stable, so sex specific validation is needed before clinical application.