Neuropsychological outcome of experimental manipulation of phenylalanine intake in treated phenylketonuria

Abstract

Blood phenylalanine concentrations were experimentally increased for 3 months by means of a phenylalanine‐complemented amino acid supplement in a group of 16 children aged 10‐16 years with classical phenylketonuria who had been treated early and who had remained on the restricted diet. Average concentrations achieved during challenge were between 1000 and 1300 µmol/L. Psychological outcome was measured by a neuropsychological battery consisting of tests of verbal and spatial memory, attention and fine motor coordination. A triple‐blind, repeated measures, randomized, crossover design was adopted to control for practice and expectancy effects. Subjects were assessed at baseline and at the end of the first and second phenylalanine manipulation periods. Significant interactions (ANOVA) emerged as predicted for phenylalanine concentrations, but similar crossover effects were not found for any of the neuropsychological tests. The results suggested that medium‐term hyperphenylalaninaemia in treated PKU is not harmful to psychological functioning in older children and adolescents who have been continuously treated up to and beyond age 10 years, though the susceptibility of executive functions needs to be further researched. The findings add some weight to the idea that by late childhood the vulnerability of the nervous system to the neurotoxic influence of phenylalanine may be much reduced.

Dietary interventions for phenylketonuria

“…Several of the included studies did not provide details on allocation concealment or generation of randomisation sequence. Only two of the four studies employed an intention-to-treat analysis (Clarke 1987;Gri iths 1998). As would be expected, blood phenylalanine concentration, which was assessed in all the studies, was found to be significantly higher in people with PKU on a relaxed phenylalanine diet than in those on a low-phenylalanine diet.…”

“…We included four studies (19 references) in this review, with a total of 251 participants (Clarke 1987;Gri iths 1998;Holtzman 1975; US/PKU Collaborative). Two of these studies employed a crossover design (Clarke 1987;Gri iths 1998).…”

“…We included four studies (19 references) in this review, with a total of 251 participants (Clarke 1987;Gri iths 1998;Holtzman 1975; US/PKU Collaborative). Two of these studies employed a crossover design (Clarke 1987;Gri iths 1998). Combined group data only were available from these two studies, which meant that the unsatisfactory approach of ignoring the cross-over design had to be employed to allow these data to be included in the analysis.…”

“…In this study treatment groups were "randomly formed" using table of random numbers. Generation of randomisation sequence was unclear in the remaining three included studies leading to an unclear risk of bias for these studies (Clarke 1987;Gri iths 1998;Holtzman 1975).…”

“…Intention-to-treat analysis was employed in two of the studies, so we judged these to have a low risk of bias (Clarke 1987;Gri iths 1998). We judge the other two studies to have an unclear risk of bias, since there were several drop outs from the studies, but reasons for these were briefly described in the published papers (Holtzman 1975; US/PKU Collaborative).…”