Nutritional status of mentally retarded children in northwest spain: II. Biochemical indicators.

Aim: To evaluate the nutritional status of mentally retarded children in the region of Galicia, northwest Spain, on the basis of serum biochemistry variables. Methods: The following serum biochemistry variables were determined in a sample of 128 mentally retarded children (81 boys and 47 girls): albumin, prealbumin, retinol-binding protein, transferrin, ferritin, ceruloplasmin, iron, calcium, phosphorus, copper, total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides, alkaline phosphatase, transaminases and carnitine. A preliminary statistical analysis indicated that most information content could be maintained taking into consideration only five of these variables. A factor analysis of the resulting 5 × 128 variables-by-subjects data matrix was then performed, identifying three factors (FB1, FB2 and FB3) that together explained 74% of total variance. Taking these factors as indicators of nutritional status, the data were then analysed for possible effects of age, gender, socioeconomic and family environment, IQ, presence/absence of cerebral palsy, quality of diet, appetite and use of antiepileptics. Results: The analysis suggests that most subjects were in the normal nutritional range, but that about 37% showed either borderline or definite malnutrition. The FB1 score showed a significant, positive correlation with age, while FB2 score was negatively correlated with age. The FB1 score showed a significant positive correlation with socioeconomic class, and was significantly higher among children from inland-rural areas than among children from coastal or urban areas. The FB1 score was also positively correlated with parents’mean age. Finally, the results suggest that antiepileptic treatment with phenobarbital or diphenylhydantoin led to reductions in serum levels of calcium and phosphorus. Conclusion: Malnutrition as revealed by biochemical variables is highly prevalent among mentally retarded children, and the prevalence is affected by socioeconomic and family environment variables.