The Use of Principal Components in the Analysis of Cross-sectional Growth Data

Cross-sectional data, consisting of 33 anthropometric measures of the body and head, were collected on a sample of 256 Irish male children between the ages of 8 and 12 years. Principal components extraction followed by Varimax rotation produced five orthogonal components accounting for nearly 79% of the total variation. These components represent general body size, body mass, head breadth, upper craniofacial height, and head length. Individual component scores were computed and analyzed to determine their relationship with age. The first component was found to have a high positive correlation with age and showed a constant increase across the age range studied. The remaining components were age-independent and represent non-growth-related vectors of variation