Phase variation in child and adolescent growth.

Area of Science:

• Pediatrics
• Biostatistics
• Human Growth and Development

Background:

• Child growth involves increases in height and maturational status.
• Functional data analysis can separate height (amplitude) and maturation tempo (phase) variations.
• Existing methods may not fully capture the complex interplay of these growth components.

Purpose of the Study:

• To develop and validate a novel, automated approach for child growth modeling.
• To differentiate and quantify amplitude and phase variations in longitudinal height data.
• To explore the relationship between growth patterns and pubertal development.

Main Methods:

• Combined Principal Component Analysis (PCA) with the Maximum Likelihood Principle.
• Analyzed longitudinal height data from 361 children in Polish and Swiss growth studies.
• Applied functional data analysis techniques to separate amplitude and phase variations.

Main Results:

• Phase variation peaks during puberty, explaining significant height variance (up to 40% in girls, 50% in boys).
• Identified 8 significant amplitude and 4 significant phase components.
• Phase components showed significant correlations with clinical signs of puberty.

Conclusions:

• The combined PCA and Maximum Likelihood Principle offers a powerful, automated tool for growth modeling.
• This approach provides estimates for future growth, adult stature, and developmental tempo.
• Preliminary findings suggest utility for automated screening in child growth assessment.