The parents of 12 children participated, of which seven with an English-speaking background and five with other backgrounds. 10 children had mutations in GLDC and 2 in AMT, and all carried at least one missense variant, the majority of which is known to confer residual activity (Table 4). There were equal numbers by sex, presence of epilepsy or not, and taking DMP or not (Table 5). All children used benzoate, but only five children achieved target plasma glycine levels [2]. Four children received psychopharmaceuticals of several groups.

The salient results of the questionnaires are shown in Table 2 and summarized in Table 6.

According to ICD-10 [24], based on formal assessments provided to us, three children had a total IQ score (TIQ) consistent with moderate ID, and one child with mild ID. Furthermore, for one child parents just indicated mild to moderate ID, whereas four children did not have an intelligence test available. The other parents were not aware of an intelligence test for their child (Table 5).

Adaptive behavior based on the ABAS GAC score revealed very low scores, including very low scores in at least two of the three domains, establishing a quantitative criterion for impairment in adaptive behavior, in keeping with a population with ID. All children achieved very low scores in the Practical and Conceptual domains, whereas the Social domain was a strength, with six of 12 children achieving a low to even a (below) average score. The highest score was in the Social domain for eight children, and the Practical domain for four children.

Of the Dutch speaking children, one child fell below the norm score range for the GAC and for the Conceptual score. A second child fell below the norm for only the Conceptual score. For the remaining children, norm scores for the Conceptual domain had a mean of 59, median 57, range 49 – 73; for the Practical domain a mean score of 61, median 60, range 48 - 82; for the Social domain a mean score of 69, median 73, range 54 – 86; and for the GAC total score the mean was 60, median 61, range 47 - 76. Nine children had a GAC score and 10 a Social domain subscore above the 0.1the percentile, which is discrepantly better than expected for their very low IQ.

Three children did not speak more than two-word sentences, and one child failed the validity criterion and was not included for analysis. The eight remaining children had a high GCC (>90th percentile) indicative of a general communication problem (six >99th percentile), affecting most subscales. In contrast to these subscales, only two children scored weak for the Non-verbal Communication Scale and two for the Social Relationships Scale. Seven of eight children had a high pragmatic score (≥90th percentile), indicative of pragmatic language problems. The eighth child with the lower pragmatic score also had the highest ABAS score. Two children used alternative communication tools like signs or communication devices.

All questionnaires identified major motor problems.

Scores on the MSC questionnaire indicated that the motor development level of four children was lower than that of children aged 4 y, the lower limit of this questionnaire, and eight children were consistent with children between 4 y 0 m and 5 y 6 m, median 4 y 3 m.

On the DCDDaily-Q, all children took longer to learn ADL (>95th percentile), and the frequency of participation in ADL and quality of performance of ADL was generally low. Children achieved respective scores corresponding to those of children aged 6 y (n=2 and n=1), children aged 5 y (n=2 and n=1), but most of them scored even lower than children aged 5 y (n=5 and n=8). Only two children (#17, #44) achieved participation scores corresponding to children aged 7-8 y (the upper limit of this questionnaire); for one of them, this was also the case for the quality score. One parent mentioned a decrease in motor performance, another parent mentioned good progress in fine motor skills, although not reflected in the measurements. The M-ABC Checklist also identified motor problems, with one child scoring equivalent to 6 y old children, five to 5 y old children, and five scored below that of children 5 y of age, and one child failed scoring for too many "not observed" responses. All parents indicated multiple non-motor factors that affect movement (Part C of M-ABC Checklist), half of them recognize seven or more factors (range 4-11): Distractible (n=12); Hesitant/forgetful, Impulsive and Overactive (n=9); Underestimates own ability (n=8); Disorganized and Lacks persistence (n=7); Timid and anxious (n=6); Overestimates own ability (n=5); Upset by failure (n=4); Passive (n=2); Unable to get pleasure from success (n=1). Parents additionally mentioned chorea (n=2), balance problems (n=3), dystonia and ataxia (n=1 each). Two parents noted variability in motor functioning during the day or over days, depending on emotional state, environment, and how well NKH was controlled.

On the DBC, seven of the 12 children had a TPS in the highest quartile. On the five subscales, four children scored in the highest quartile for the disruptive scale, seven children for the self-absorbed scale, five children on the communication disturbance scale, four children on the anxiety scale (none of which took fluoxetine), whereas no child on the social relating scale. Two children had no maladaptive behavior within the highest quartile across all scales, interestingly both with a mutation in AMT.

This problem behavior did not relate to overall functioning. Two of three children with overall better functioning (DBC TPS < P75, SCQ < 15, ABAS GAC > P0.1, Motor > 4y and DBC ADHD symptoms ≤ 6/9), had an abnormal score on the DB Anxiety scale. Further, three children with overall score of very weak performance (ABAS GAC ≤ P0.2, insufficient speech to administer the CCC-2, motor functioning <3y4m), had generally adequate DBC scores.

All parents provided examples of characteristics of the restricted, repetitive behaviors symptom domain for ASD, with nine of 12 parents’ examples fitting at least two characteristics of this symptom domain, which is one of the conditions that supports a DSM-5 diagnosis of ASD [39]. Parents mention several specific fears. For one child, according to the parent, the many hospital interventions in her early childhood may have caused post-traumatic stress disorder.

On the SCQ, the mean score was 17, median 16, range 7-27. Seven children exceeded a score of 15, considered indicative of possible ASD [29, 30]. Of these seven children exceeding the cut-off, when considering their current functioning, only two still meet this cut-off, both with a score of 15. Thus, this data suggests that the children exhibited more autistic behaviors around the age of 4 to 5 y, than they did at the current age of the questionnaire.

On the ABAS, the social domain appeared to be a strength. Six children had a norm score in the low to (low) average range above 70. The highest norm score was for the social domain in two-thirds of children. On the DBC social relations scale, no child scored in the highest quartile. On the CCC2, no child had a high SIDI (>90th percentile), which together with a high GCC would have indicated ASD. Rather, on the contrary, eight subjects had a low SIDI (<10th percentile), while they scored >90th percentile on both the GCC and the pragmatic score. Despite this, only one scored >95th percentile for both Social Relationships and Interests. Three children scored >95th percentile for the Interests Scale, but without an increase in the score for the Social Relationships Scale, and one child scored >95th percentile for Social Relationships Scale while the child had a normal score for the Interests Scale. Three children had normal scores on both scales [31, 32]. Taken together, these results do not indicate a systematic indication for ASD.

In the DBC questionnaire, 10 parents indicated at least two-thirds of the nine ADHD characteristics present. In the M-ABC Checklist Part C non-motor factors, of the five ADHD related symptoms, five symptoms were present in three children, four in six children, and three in two children. Of all non-motor factors affecting movement, the seven factors present in >50% of children all related to ADHD: Distractible (n=12), Hesitant/forgetful, Impulsive and Overactive (n=9) and Disorganized (n=7). There was no significant correlation between both ADHD related scores (DBC-related ADHD questions and M-ABC Part C ADHD related questions) (Spearman ρ=0.24 p=0.46), reflecting that they evaluated different aspects of this domain. A single child (#17) had less than the 50% of ADHD related symptoms present on M-ABC part C, and had the fewest ADHD symptoms on the DBC, but was not different from other children on the other questionnaires.

In an exploratory ANCOVA analysis for factors influencing the TPS on the DBC, which was normally distributed, on a linear model using as variables sex and use of DMP, and as covariates age and last glycine level, only sex and DMP use were significant factors whereas glycine level and age were not. In an optimized model of TPS (F=5.79, p=0.021), predictors sex (F=8.78, p=0.018) and DMP use (F=13.17, p=0.007) were significant. Girls had more problems than boys and children taking DMP had a higher score than those not taking DMP. Similarly, the similar ANCOVA analysis of the ADHD 9 questions score of the DBC had significant predictors of sex (F=9.33, p=0.016, worse in girls) and use of DMP (F=6.41, p=0.035, worse when taking DMP) with the overall model (F=4.17, p=0.047) without a significant interacting term. Similar correlations existed with sex and DMP use on the DBC disrupt score, although the overall model was less significant. Indeed, on the DBC, the TPS significantly correlated with ADHD 9 subscore (R=0.886, p<0.001) and the Disrupt subscore (r=0.899 p<0.001), but not with the DBC Social Relations score. Glycine levels did not correlate with any outcome factor, and there was no difference in any outcome class by glycine level whether in control or not. The single child with both alleles having a variant conferring residual activity (#23) did not fare better than other children. No statistical relationship was found with the ABAS scores.