ABSTRACT

The use of the growth curves for the assessment of pediatric growth is of great importance. In this study, different models are investigated for the prediction of head circumference of 0–2 aged Turkish children.

Measurements of head circumference for 0–2 aged children living in Ankara were analyzed. Linear-Linear, Exponential, Linear, Parabolic, Richards and Wilmink models were used for prediction of 0–2 aged children. Goodness of fit of the models was evaluated based on goodness of fit criteria.

When evaluated the predicted growth curves based on coefficient of determination (R2), mean squuare error and iteration number values, Linear-Linear, Exponential, Linear, Parabolic, Richards and Wilmink models performed well for the prediction of head circumference. The best models are Richards for girls, and Linear-Linear for boys based on goodness of fit criteria, respec-tively.

As a result, several models had different performance for the prediction of non-linear growth process. The growth curves achieved from the models are useful for monitoring of the head circumference.

Keywords: Prediction model, head circumference, growth curve, children