Shape-invariant modelling of human growth

Area of Science:

• Biomathematics and Human Growth Modelling
• Longitudinal Data Analysis

Background:

• Accurate modelling of human height growth is essential for understanding developmental trajectories.
• Existing models may have limitations in fitting longitudinal data, especially during puberty.

Purpose of the Study:

• To introduce a new mathematical algorithm for modelling human height growth, applicable to other variables.
• To develop a shape-invariant model (SIM) for bias-free fitting of longitudinal growth data.
• To compare different SIM variations, including a 'switch-off model', for biomathematical insights.

Main Methods:

• Developed a mathematical algorithm that iteratively refines functional form guesses using data.
• Implemented a shape-invariant model (SIM) for fitting longitudinal data from 1 to 20 years.
• Compared an additive two-component SIM with a 'switch-off model' where puberty inhibits non-pubertal growth.

Main Results:

• The shape-invariant model (SIM) provides approximately bias-free fitting for longitudinal height data.
• The 'switch-off model' variant demonstrated superiority in several aspects compared to the additive model.
• Identified qualitative growth features: midspurt, pre-pubertal dip, and asymmetric pubertal peak.

Conclusions:

• The shape-invariant model (SIM) is a robust approach for biomathematical modelling of human growth.
• The 'switch-off model' offers a more accurate representation of pubertal growth dynamics.
• Preliminary analysis of individual parameters supports existing findings on sex differences and adult height correlations.