From anthropometrical evaluation to the prediction of nutritional facts at infancy
Mardones-restat, F.; Jones, G.; Diaz, M.
Enfoques en Atencion Primaria 5(1): 3-10
1990
ISSN/ISBN: 0716-2774 PMID: 12283399 Document Number: 371879
This study developed instruments to predict the unsatisfactory growth of children in the context of social programs, (health, education, nutrition, social work, food and employment) which measure the probability of malnutrition using anthropometry. Variables used were biological and social (family incompetency). A cohort of 822 breastfeeding infants were followed prospectively from prenatal until 12 months of age in 5 primary health clinics in Santiago, Chile. Weight by age was the dependent variable, defined as "unsatisfactory growth" if it was under the 20th percentile established by the World Health Organization. By 1 year, 32% of the study population was in this category, reflecting national statistics; at 3 months 8% of the infant population and no more than 2% of those breastfeeding were under the 20th percentile. 25 high risk variables were selected that also demonstrated a high correlation with weight and age and these were evaluated by using 3 statistical methods: 1) the relative simple risk (RRS) used by the WHO; 2) logistical regression; and 3) classification and regression trees (CART).