N. T. Ho
Background: Prepubertal body mass index (BMI) has been reported to associate with pubertal growth spurts, but whether BMI reshapes the entire distribution of pubertal growth, not just its average, is unknown, particularly in Southeast Asian children. Methods: We studied 8,018 boys, 5,495 girls from the 2018-2025 cohort of schoolchildren in 3 major cities in Vietnam with SITAR derived growth curves and prepubertal BMI z scores. Individual height SITAR size, timing, and intensity (velocity) parameters was modeled with linear quantile mixed models across quantiles 0.05 to 0.95, separately by sex, with a sensitivity model re-centered on each child's own age at peak height velocity. Results: In boys, the BMI to velocity association reversed across the distribution, from 0.045 cm/year per SD at the 5th percentile to -0.101 at the 95th (heterogeneity p<0.0001), and this reversal persisted after re-centering on each child's own pubertal peak (p<0.0001). In girls, the association stayed positive at every quantile, from 0.132 to 0.187, without significant variation across the distribution in chronological age (heterogeneity p=0.22), though a significant distributional pattern reemerged once re-centered on pubertal peak (p=0.03). BMI compressed the spread of pubertal timing in both sexes and of intensity in girls (p<0.0001 for timing in both sexes, p=0.000002 for intensity in girls). Individual timing and intensity were inversely correlated in boys (r=-0.85) and positively correlated in girls (r=0.74). Conclusion: Prepubertal BMI reshapes the distribution of pubertal growth in sex-specific ways. Clinical and epidemiological pediatrics should account for childhood BMI's complex, distribution-specific, sex-dependent impact on adolescent development.