Appendix B: Final Report of September 15, 2000-A Fifteen Year Follow-Up on Participants in the Collaborative Study of Children Treated For Phenylketonuria (PKUCS 1967-1984)

The content in this publication was accurate at the time it was published, but it is not being updated. The item is provided for historical purposes only.​

By Robert Koch, (1) B. Burton, (2) J. Coldwell, (3) Richard Erbe, (4) M. Geraghty, (5) G. Hoganson, (6) R. Peterson, (7) W. Rhead, (8) R. Rouse, (9) R. Scott, (10) C. Shear, (11) Janet Thomas, (12) J. Wolff, (13) A. Stern, (1) C. Azen (1)

(1)Childrens Hospital Los Angeles, Division of Medical Genetics, Department of Pediatrics and University of Southern California California, School of Medicine, Los Angeles, California; (2)Children's Memorial Medical Center, Chicago, Illinois; (3)Children's Medical Center, Tulsa, Oklahoma; (4)Childrens Hospital, Genetics Division, Buffalo, New York; (5)Johns Hopkins Hospital, Institute of Genetics, Baltimore, Maryland; (6)University of Illinois at Chicago, Department of Pediatrics, Chicago, Illinois; (7)San Diego Regional Center, San Diego, California; (8)University of Iowa, Division of Medical Genetics, Iowa City, Iowa; (9)Children's Hospital, University of Texas Medical Branch, Galveston, Texas; (10)University of Washington School of Medicine, Department of Pediatrics, Seattle, Washington; (11)Developmental Pediatrics, Boca Raton, Florida; (12)Childrens Hospital, Denver, Colorado; (13)Waisman Center on Mental Retardation and Human Development, University of Wisconsin, Madison, Wisconsin

Funded by NIH Contract No. N01-HD-2-3148 from the National Institute of Child Health and Human Development (NICHD), Bethesda, Maryland; Childrens Hospital Los Angeles Research Institute, USC School of Medicine, Los Angeles, California; Computational Assistance was provided by the NIH NCRR GCRC M01 RR-43 CDMAS Project and was performed at the GCRC at Childrens Hospital Los Angeles, Los Angeles, California.


Two prior preliminary reports of the PKU Follow-Up Project have been submitted to the National Institute of Child Health and Human Development (NICHD). The initial report included data on 42 enrollees (27-32 years of age) as of January 1, 2000; the second report on 4-15-00 included 53 enrollees, and this report includes 67 enrollees.

Table 1 illustrates the response that the project has received from the Participating Clinics in the original Collaborative Study of Children Treated for Phenylketonuria (PKUCS 1967-1984). The Participating Clinics varied greatly in their ability to locate participants for evaluation, varying from 20 to 87% for a mean of 52%. Thus, nearly 50% have been lost to follow up for a variety of reasons. The clinics that were able to locate greater than 50% of their enrollees offered continuing services for adults with PKU.

Despite a national trend to treat older persons with PKU, many were lost to follow up services. It must be assumed that the persons who could not be located are probably not on any restriction of phenylalanine in their diet.


After a 15-year interval between the two studies, 67 of the 129 enrollees who were still active at the end of the original study, were located and evaluated. There were no significant differences between those who were and those who were not evaluated as adults on diagnostic, treatment, demographic or childhood psychometric variables, except that those who were re-evaluated had higher Parental SES (p=0.045). Table 2 reveals that seven of the 67 enrollees never discontinued their phenylalanine restricted diet, while 36 discontinued before 6.5 years of age. 13 between 6.5 and 12 years of age and 11 between 12.5 - 20 years. Twenty-six of the 60 discontinuers resumed dietary treatment (44%), and remained on diet for periods ranging from a few months to more than 10 years. Those who never discontinued had a tendency toward lower diagnostic phe, earlier initiation of diet and lower phe levels during childhood than those who discontinued at various ages.


Medical findings varied considerably (see Table 3). The findings were based partially on self-reporting by the enrollees and therefore, it is difficult to be certain of their accuracy, except for the neurological signs which were based on a physical examination performed by the physician in the Participating Clinics. These findings are more readily documented, as is the reported instance of seizures. Although the number of participants does not lend itself to statistical comparison among groups, the fact that neurological findings were reported in 22% of those who discontinued diet before 6.5 years of age, in 27% of those who discontinued diet between 6.5 - 12.5 years, and in only 11% of those treated longer than 12.5 years and none of those who never discontinued the diet suggests a beneficial effect of longer treatment. The one person with seizures discontinued diet prior to 6.5 years of age. Mental disorders occurred in a pattern similar to neurological findings: 19%, 17%, and 10% of the respective discontinuation groups, and none of the continuers. The hypertension reported in 2 diet continuers were associated with increased BMI values.

It is of interest that 3 of the 7 diet continuers are obese, with body mass index over 30. This may be associated with the use of low protein products but needs further evaluation. No cancer was reported in any of the participants but eczema and asthma were more frequent in diet discontinuers.

Educational and Social Status

Educational status was obtained by self-reporting and social class was categorized by the Hollingshead Two Factor Index of Social Position, based on the education and occupation of the head of household. As shown in Table 4, an increasing percentage of participants graduated from college, as dietary continuation increased (26% to 71%). For those graduating from high school and some college, the figures are 63%, 77%, 36% and 29%. For those with less than a high school education, percentages were 11, 0, 9 and 0 respectively. Thus, the data support the prolongation of treatment well beyond childhood and adolescence. Note that social data also support this. Those who stayed on diet longer had higher percentages in the two highest classes, while those who discontinued by 6.5 years had the highest percentage (53%) in the two lowest classes. In paired comparisons, 27 PKU Collaborative Study adults who had non-PKU siblings had comparable percentages who were college graduates (41% each) and in the two highest Hollingshead class (31% each). These findings were consistent across diet discontinuation groups. All groups had percentages of college graduation at or above the reported national average of 24%.

MRI/MRS Findings

Data remain unchanged since the previous report, since staff has concentrated its efforts on the Filipek study which is still ongoing. Table 5 provides data gathered on 15 subjects who were included in the MRI/MRS pilot study. In data already published, normal brain content of phenylalanine is 0.1 mM and carriers exhibit 0.2 mM. Six of the subjects (I-17, L-22, L-10, S-01, A-08 and Q-04) demonstrate low brain phe content. In two continuers (A-08 and Q-04) the blood phe levels are in concert with very low brain levels. In the other 4 subjects, the brain phe content is significantly below what one would expect in view of their elevated blood phe levels. The IQ data, however, are inconclusive. The 7 continuers exhibited a mean IQ of 112 and the 7 discontinuers, a mean of 102. The latter group contains 2 individuals with IQ's of 75 and 83. The individual with the IQ of 83, discontinued diet treatment at age 6 years, but resumed ingesting a medical food prior to having her brain phenylalanine content determined, and that may have lowered her phenylalanine brain level due to the large neutral amino acid content in the restricted phenylalanine dietary product (brain level of phe 0.22 mM). The 10 subjects with low brain phenylalanine content (<0.40mM) exhibited a mean IQ of 108 and the 4 with high brain levels 104. The numbers are too small, at this time, to draw any conclusions from these data.


The protocol for this part of the PKU Follow-Up Study includes 20 subjects who will have neuropsychological and MRI segmental evaluations, and simultaneous blood and brain phenylalanine levels performed by the new scanner. Extensive neuropsychological testing has been administered to 13 patients to date.

This portion of the Follow-Up Study has been hampered by two unexpected problems. The installation of the new MRI/MRS equipment at CHLA was delayed. Completion and state licensing finally occurred on July 17, 2000. The new equipment is a state of the art General Electric instrument. Unfortunately, the equipment at University of California at Irvine (UCI) utilizes Siemens brand equipment from Germany. Dr. Rex Moats has been working closely with the UCI group, but it has only been resolved as of 2 weeks ago. Therefore, this part of the study has been inexorably delayed and it is unlikely to be completed by October, 2000.

During the remaining time we are hopeful that 10 of the 20 subjects will be available to Dr. Filipek to conduct this part of the study. Five subjects were examined for this part of the examination, but these were deemed incompatible with the segmentation software. Dr. Moats is hoping to solve this problem so the data are not lost. Two are compatible.

Mutation Study of the Phenylalanine Hydroxylase Enzyme

Data from 66 of the 67 enrollees have been sent to Dr. Guttler in Denmark and are tabulated in Table 6. The mutations have been rated by Dr. Guttler as severe (zero conversion of phenylalanine (phe) to Tyrosine (Tyr), moderate (some conversion of phe to tyr), mild (more conversion of phe to tyr) and mild hyperphenylalaninemia (MPH) (significant conversion of phe to tyr). The latter were excluded from the study in 1967-1984. Thus, none have been identified in the present sample. Mutations for eleven of the 67 enrollees are still in process of identification. Those with two severe mutations had a median WAIS-R IQ of 95 in comparison to 100 in those with one severe and one moderate and 101 in the severe/mild groups. Table 6 illustrates these results.

The preliminary results are compatible with those reported by Guttler et al from the Maternal PKU Collaborative Study, but are as yet incomplete because 11 are still in the process of identification. If these preliminary results are further validated, it suggests that the severity of the mutation reduces the quality of treatment and therefore, their higher blood phenylalanine levels decrease adult intelligence as measured by the WAIS-R.

Psychological and Achievement Data

Figure 1 illustrates the mean Wechsler test scores from age 8 years to adulthood. The drop in IQ between 10 and 12 years of age may be due partly to the change from the WISC to the WISC-R, which the literature suggests may account for a difference of 7 points. No decline in mean IQ scores was observed between age 12 and adulthood. Figure 2 shows the WRAT scores from age 8 to adulthood. The adult scores show that the decline from age 10 to 12 in arithmetic scores is reversed by adulthood, and that mean spelling scores improve as well.

Figure 3 reveals that the arithmetic subtest scores on the WAIS-R are below the population mean, as are information, digit span, and object assembly. Figure 3 indicates that the adult PKU mean subtest scores are at the population mean in vocabulary, similarities, picture completion, picture arrangement, block design and digit symbol. It should be noted that the arithmetic subtest of the WAIS-R versus the arithmetic achievement score of the WRAT-3 measure vastly different cognitive and psycholinguistic functions. Direct comparison of the differences between these two domains are misleading.

Figure 4 illustrates the importance of diet restriction continuation into adulthood. Thirty-three enrollees who discontinued dietary treatment at 6 years of age, scored a mean IQ of 91 on the WAIS-R but 14 points below their parents' mean score. Continuation of dietary treatment in 13 enrollees to 12 years of age improved enrollee mean scores to 95 but were still 15 points below the parent mean IQ score of 110. Continuation of dietary therapy beyond 12 to 20 years in eleven and indefinitely in 7 nearly equalized enrollee mean scores to parental mean IQ WAIS-R scores. Again, the numbers are small but the trend is important, and diet discontinuation before age 12 resulted in WAIS-R IQ scores significantly below parental mean IQ.

Figure 5 is significant in that it shows that the blood phe level is very important. Enrollees with a mean blood phenylalanine level below 15.9 mg% scored a mean WAIS-R IQ 6 points below their parents, while those between 16 to 19.0 mg%, scored 8 points lower than their parents. When the levels were consistently above 20 mg%, the PKU adult scores were below the mean of their parents by as much as 16 IQ points. Even though the numbers are small, the trend is important to consider.

This observation is fortified by Figure 6 which shows the subjects with 2 severe mutations of the phenylalanine hydroxylase gene tend to have lower mean intellectual quotients than those that do not.

Tables 7 and Table 8 present adult WAIS-R and WRAT-3 mean scores and change from age 12 for PKU Collaborative Study adults grouped by age at discontinuation and by current phe level. The importance of parental IQ as an indicator for judging the significance of offspring achievement on the WAIS-R and the WRAT-3 test scores is seen here.

Table 9 shows correlation coefficients between adult test scores and parent IQ, years on diet and phe levels at various ages. All test scores are positively correlated with parent IQ and negatively correlated with phe levels at age 6, age 10 and the usual adult phe level. After controlling for parent IQ, all IQ scores had significant partial correlations with phe at age 6 and 10. Verbal and Full Scale IQ scores also had significant partial correlations with years on diet and adult phe levels.

WRAT-3 Reading and Spelling scores had significant partial correlations with phe at age 6, after controlling for parent IQ.

In another approach to evaluating the relative important of a number of demographic and treatment factors to adult PKU test scores, we performed stepwise regression for each of the 6 scores, with these possible predictors: maximum diagnostic phe level, age at first treatment, parent IQ, father's education, years on diet and phe level at age 6 years, 10 years and currently. Variables were added to the model until the p-value for including the next variable was 0.15 or greater. Table 10 summarizes these analyses for WAIS-R scores and Table 11 for WRAT-3. Parent IQ or education, age first treated and childhood phe were the best predictors. Adult phe level improved the models for the IQ scores, but not the WRAT-3 scores.

Neuropsychological Assesment of Adults

Neuropsychological assessment to date has consisted of 14 subjects between ages 27-32. All subjects had an IQ in the low average to high average range, with the exception of one subject with Borderline IQ. Preliminary findings suggest that the sample appears to show greater than expected difficulty on a number of tasks that measure executive skills including mental flexibility and initiation. The group showed a tendency toward perseveration on several tasks. Findings also suggest that verbal learning and memory skills are below expectation. Focused attention, visuo-perceptual skills, expressive and receptive language as well as fine motor skills appeared intact. Standardized self-report measures of depression and anxiety were in the normal to mild range on 11 subjects and in the moderate/severe range for 3 subjects. Findings also suggest that the group with a concurrent phe level >15 mg% scored lower than the group with phe level <15 mg% on measures of reaction time, sustained attention, mental flexibility, verbal fluency and expressive naming.

In conclusion, this sample appears to show greater than expected difficulty on a number of tasks which measure executive skills and verbal memory. Although this pilot study data appears to show a trend in the total group toward deficits in executive skills that may be more pronounced with higher phe levels, no statistical testing has been completed since the study is incomplete. A sample size of 20 is anticipated. This will allow for further investigation of these preliminary results and to assess the impact of treatment upon cognitive outcome.

Listings of raw data for the 67 patients, ordered by adult phe level, and by age at diet discontinuation appear in the Key Variables for Cases Evaluated for PKUCS Followup Study.


In summary, the PKU Follow-Up Study data strongly support continued dietary therapy into adulthood. The justifications are as follows:

  1. Funding agencies will not provide dietary supplies for the restricted phenylalanine diet unless there is a strong recommendation that it is beneficial. Without a national consensus, the fractured resources associated with insurance companies, Health Maintenance Organizations, Federal and State funding agencies, suggest that organized resources will not be available. Eligibility for Medicare coverage should be explored to provide the care needed for PKU adults.
    1. The PKU Follow-Up Study data has shown that early diagnosis and early and long-term treatment are very important to a successful outcome. Persons with phenylketonuria associated with two severe mutations of the phenylalanine hydroxylase gene are at a greater risk for poor control of blood phenylalanine levels. The data suggest that blood levels below 16 mg% (960 micromoles/Liter) foster improved intellectual outcome.
    2. Health resources must be better organized to provide necessary interdisciplinary treatment, assessment, and guidance.
    3. It is especially critical for female adults who must be treated more rigorously to prevent congenital defects associated with the Maternal PKU Syndrome in their offspring.
  2. Newer technology needs to be explored systematically. MRI/MRS, Phenylalanine
    Hydroxylase Mutations, and the use of Tetrahydrobiopterin therapy, which are needed to improve therapy and long-term follow-up, offer new avenues to learn more about this devastating disorder. Resources for mutation studies are unavailable in the USA. NICHD should fund such a national resource; preferably one in the East, one in the Midwest and one in the West to facilitate services.
  3. A demonstration project is needed to show that the older untreated adult PKU population languishing in state institutions, group homes, and those still living in their own home would have improved behavior, decreased medical problems such as seizures, decreased need for skilled nursing care, repeated hospitalizations and psychotropic drugs. It is not unusual to find untreated PKU individuals on several psychotropic drugs that actually cost more than dietary restriction of phenylalanine.

    A trial such as this would take 4 years to demonstrate significant improvement due to the severity of the disorder in this older untreated PKU population.

    We know behavioral improvement occurs rapidly in treated persons with dietary control of phenylalanine, but motor, seizure control and some improvement in intelligence requires more time for habilitation to occur.

Recommendations for the Consensus Panel to Consider

Based on the follow-up of 67 enrollees in the 1967-1984 PKU Collaborative Study, the Committee may wish to extend this study for 6-12 months to obtain more data on former enrollees who are willing to participate, but were unable to do so due to various factors such as job obligations, pregnancy, personal reasons, etc. In addition, IRB regulations delayed recruitment in 3 Participating clinics.

With the publicity generated by the Consensus Conference, I am confident that 15-20 additional enrollees could be identified. The original funding was for an estimated 90-100 enrollees.

The neuropsychological evaluations are completed in 13 subjects, but interpretations are not available as yet. Every effort will be made to send you a report on these subjects soon.

first | previous | next | last

top of pageBACK TO TOP