Predicting factors of neurodevelopmental performance in children with phenylketonuria

In phenylketonuria (PKU), high phenylalanine (Phe) levels hamper neurodevelopment impairing executive function later in life. While the second has been more studied, fewer data exist on predictors of PKU patients' development in specific populations. To contribute to the field, we performed a retrospective analysis of predictors of neurodevelopment in PKU patients in a Portuguese cohort. We analyzed the retrospective data on the metabolic control of 89 patients, as their health and familial features. Griffith's Mental Development Scale performance at age 6 (GMDS6) was used to assess neurodevelopment. Our cohort included 14 GMDS6_low and 75 GMDS6_high patients. In a multivariate analysis, the better predictors of neurodevelopment were the metabolic control at age 3 and year of birth (n = 87, β₀  = -121, β₁  = -1.77, β₂  = 0.06, LRchi2(2) = 13.61, Prob > chi2 = 0.001, Pseudo R2 = 0.1773). With this model, it was possible to define a safety cut-off of 7.8 mg/dL for the Phe level at age 3 (sensitivity = 72.6%, specificity = 78.6%), confirming the safety of the cut-off of 6 mg/dL already used in the clinical practice. Our study supports the relevance of metabolic control to predict the neurodevelopment of PKU patients, in the historical context of the disease management.