Holtzman 1975.

Study characteristics
Methods	Randomised, controlled parallel study. Duration 2 years.
Participants	10 participants diagnosed with PKU in the neonatal period and commenced on a low‐phenylalanine diet by 1 month of age, were randomised to either continue on the low‐phenylalanine diet or to terminate the diet at 4 years of age. 5 participants were randomised to each group and the number of males and females in each group was the same (3 males, 2 females).
Interventions	The diet continuation group was continued on a low‐phenylalanine diet. Those in the termination group gradually relaxed the diet following a standard protocol until the diet was unrestricted.
Outcomes	Blood phenylalanine levels were measured at the start of the trial, then at 2, 4, 6 and 8 weeks after the initial time point, and again 6 months, 1 year and 2 years later. Phenylalanine intake was assessed from diet diaries carried out at the same time points as the blood phenylalanine levels. The following physical examinations were carried out when the children were 4, 4.5, 5, 5.5 and 6 years of age: weight gain; height; and head circumference. Neuropsychological examinations were also carried out at these time points and included the following: Stanford Binet intelligence test; the vocabulary; similarities; animal house and the geometric designs subtests of the Wechsler preschool and primary scale of intelligence, auditory and visual sequential memory and auditory reception subtests of the Illinois test of psycholinguistic abilities. A behaviour rating of each child's performance during the test was also obtained.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Participants randomly allocated, method not given.
Allocation concealment (selection bias)	Unclear risk	Not discussed.
Blinding (performance bias and detection bias) Participants	High risk	Parents blinded as to which group child was in until time that restricted diet ceased, at which point they knew treatment group.
Blinding (performance bias and detection bias) Clinicians	High risk	Not blinded.
Blinding (performance bias and detection bias) Outcome assessors	Low risk	Psychological examiner was blinded to diagnosis of PKU.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Data on some of the participants were missing at several of the intermediate time points and at the end of the study, reasons for drop outs given (Holtzman 1975).