Executive functions in preschool children with moderate hyperphenylalaninemia and phenylketonuria: a prospective study

Phenylketonuria (PKU) is an autosomal recessive inherited metabolic disorder with a prevalence of 1 in 10,000 births in Europe caused by a deficiency of phenylalanine hydroxylase (PAH). Management recommendations have recently been drafted at the American [1] and European [2] levels. A distinction is now made between patients with PKU (phenylalanine (Phe) levels ≥ 360 μmol/L) and with moderate hyperphenylalaninemia (MHP) (120 μmol/L > Phe < 360 μmol/L). Neonatal screening for PKU makes it possible to rapidly implement specific management to reduce plasma Phe levels. For MHP patients, a residual amount of enzymatic activity would be present thereby avoiding severe neurodevelopmental disorders which can occur if PKU is left untreated [3].

Neonatal treatment can prevent the development of encephalopathy related to hyperphenylalaninemia (HPA). However, recent studies have shown a «suboptimal» cognitive outcome. The intelligence quotient (IQ) in the PKU is average but significantly lower than in the control population e.g., [4]. Additionally, processing speed, fine motor skills, and episodic memory impairment are observed in patients with PKU [5]. In subjects with MHP, these cognitive functions are likely to be relatively preserved e.g., [6, 7].

Impairment in executive functions (EFs) may explain the neuropsychological phenotype observed in PKU e.g., [8] and is suspected in MHP [9, 10].

EFs refer to a variety of high-level functions required to perform goal-oriented behaviours, primarily involving the prefrontal cortex and its networks. They allow the individual to adapt to new situations, particularly when automatisms are insufficient to carry out the task at hand [11].

However, very little research has examined the incidence of executive dysfunction during preschool years, with mixed results (see Table 1). In all studies involving preschool-age PKU patients, performance was below that of the control group or the test norm [12, 13]. There are also more complaints in the questionnaires about flexibility and control [14]. In MHP patients, the results are contradictory: one study found no disorders [15], while the other reported a continuum of impairment between PKU and MHP patients compared to controls [16]. Limitations inherent to the studies include the fact that the classification of PKU/MHP was variable with Phe levels ranging between 240 and 600 µmol/l for MHP [16]. Furthermore, the assessment of EFs is often limited [17] and only one study has combined performance tests with parental daily life measures [14].

Several studies have investigated the relationship between biochemical markers and cognitive impairment in early-treated PKU. Currently, it appears that the best predictor of neuropsychological disorders, and especially executive disorders, is the mean / median levels of Phe during lifetime and the variability of these levels over time. Recent studies have also shown a possible link between EFs disorders and an elevated Phe/Tyrosine (Tyr) ratio over time or at the assessment time [see 5].

In this study, we propose to explore the hypothesis of early executive dysfunction in HPA patients, while distinguishing PKU and MHP patients according to the new classifications, and to examine the links between EFs and certain metabolic variables (mean Phe levels and Phe/Tyrosine ratio over the course of a lifetime, variability of Phe levels over the course of a lifetime, Phe levels and Phe/Tyrosine ratio at inclusion). We used a protocol that combined several executive performance-based tasks with daily life questionnaire (parents and teachers). It is crucial to identify EFs disturbances from preschool age to promote their management given their profound impact on environmental adaptation, academic success and, learning [18].

Based on previous studies, we hypothesized that children in the HPA group would manifest impaired EFs (and increased reports in daily life) compared to the control group (1). We also expected to observe a continuum of executive impairment with increasing severity of HPA, resulting in poorer test performance and increased reports on daily life questionnaires in PKU patients compared to MHP patients (2). Lastly, associations are expected between metabolic indicators and test/questionnaire scores: the higher the levels, the worse the executive performance/the higher the complaints (3).

The objective of this study was to investigate EFs in children with HPA children and the possible impact of certain metabolic variables on their development in accordance with the new classification recommendations between PKU and MHP [1, 2]. Furthermore, the approach used to study executive functions included combining performance-based tests with daily life questionnaires, including perspectives from parents and the school environment.

For the HPA group, the pattern of results in our study does not support our initial hypothesis suggesting early executive difficulties compared to the general population. Rather, the results show some variability in the HPA group depending on the tasks considered. In performance tests, patients manifested difficulties in the Hand-Game test, suggesting possible deficiencies in the inhibition of a motor response: they make more errors (trend) and self-correct significantly less, which may suggest a relative lack of control of the action. In contrast, HPA children perform better in three tests assessing verbal inhibition (Sun-Moon Stroop, error score), cognitive flexibility (DCCS) and affective decision-making (CGT, trend only). Finally, they performed similarly to controls on all other measures (i.e., the majority) of visual and auditory verbal working memory (forward and backward digit span, forward and backward visuospatial span), verbal inhibition (Sun-Moon Stroop interference score), and flexibility (preschool Brixton test). On the daily life questionnaires (BRIEF-P), there was an equivalent level of reports from both parents and teachers, apart from increased difficulties identified by teachers on the inhibition scale. This significant heterogeneity in the results could be explained by the fact that preschool children are likely to show significant variability in their motor and cognitive development, making it impossible to identify a specific profile. Our results differ from the only study that compared preschool MHP and PKU patients [16], which showed group-wide executive impairment in inhibition and working memory. However, the Phe levels considered were higher and did not yet follow the new classification recommendations. In fact, in our sample of children with HPA, mean Phe levels were quite low and according to current management recommendations (< 360 µmol/L: 299.831 + -93.526), a threshold considered likely to prevent cognitive disorders and especially executive disorders [9, 12, 30].

When comparing the results between the two subgroups of patients, our data confirm that there are possible early executive difficulties, but that they vary significantly according to the type of hyperphenylalaninemia (PKU or MHP), which tends to confirm our second hypothesis. In fact, three executive indicators referring to cognitive inhibition, verbal working memory, and visuospatial working memory were significantly worse for PKU children than for MHP’s, on three different tests: the Sun-Moon Stroop error score, the forward digit span, and the backward spatial span. These data support the idea that executive difficulties appear early in PKU children, in contrast to MHP children. Our results appear to be consistent with previous research that has investigated the executive functions of preschool PKU children using performance tests, both in comparison to control subjects and to test norms [12‐14, 17]. For MHP children, our results are partly consistent with previous data in preschoolers. Indeed, the study by Sharman et al. [15] found no evidence of executive impairment in their longitudinal study that included 4 patients with MHP (according to the new recommendations). In contrast, Diamond et al. [16] showed a continuum of impairment between PKU and MHP compared to control subjects. However, these outcomes should be interpreted while recognizing that patients with MHP had Phe levels up to 600 µmol/L (new recommendations 120 µmol/L > Phe < 360 µmol/L), and executive impairment mainly observed in children with MHP with the highest levels.

The executive impairment that preferentially affects patients with PKU could be explained by the nature of the enzyme deficiency, which is more severe in these patients. Despite strict diet control, the "toxic metabolites" pass the blood–brain barrier and could affect EFs earlier than in the case of children with MHP (in case of intercurrent diseases such as fever, virus, etc.). In fact, this phenomenon is likely to lead to an alteration in intracerebral protein synthesis and the synthesis of certain neurotransmitters, particularly dopaminergic ones. Dopamine has an important role in the frontal-subcortical circuits underlying the emergence of executive functions [16, 31]. Similarly, excess Phe has a negative impact on myelin production and can lead to white matter lesions, especially in the frontal lobe, in the form of T2 hyperintensity [2, 32]. On the contrary, MHP patients have residual enzymatic activity that would allow them to reduce the metabolic consequences of the deficit and, to maintain lower concentrations of Phe [3]. This would limit the cognitive impact and delay the potential onset of executive difficulties. Analysis of the intellectual profile of our sample provides additional answers to this issue. Indeed, while PKU and MHP children are comparable in all IQ indicators with scores in the normal range, the WMI tends to be significantly lower in patients with PKU (p = 0.051, moderate effect size), which could confirm a particular or even specific fragility or EFs, at an early age, in the case of PKU. This working memory impairment is also found in the study by de la Parra et al. [33] when they analysed intellectual efficiency scores in school aged PKU and MHP children. Specifically, the intellectual profile was similar between the two groups of patients despite more pronounced difficulties in PKU children. It is therefore quite possible that executive difficulties are not yet visible in younger children with MHP because of partial enzymatic impairment but that they may appear over time. Two recent reviews of the literature point in this direction and report indicators of executive dysfunction in older MHP patients in terms of working memory and inhibition [34, 35].

However, our results show that poorer executive performance in preschool PKU patients compared to MHP patients only affects 3 of the 7 performance-based tasks and that there are no significant differences in the daily life questionnaires of parents or teachers. Therefore, the executive difficulties identified in some tests are not found in daily life. These data do not compare well with previous work in the literature, as daily life questionnaires have only been used once in preschool children and only those with PKU [14]. In that study, parents' reports were higher compared to controls, but only on the flexibility and control scales, which is not the case in our results. Alternatively, the only study that used a daily life questionnaire in school-aged MHP patients (BRIEF) did not present executive reports from parents [15], in contrast to what has been shown at this age in PKU [14, 36, 37]. In view of these data, it is possible that the differential in executive impairment between PKU and MHP may not yet be sufficiently noticeable in young children by family members and in the school environment, since EFs are still relatively unused in daily life during this period.

Finally, regarding the links between metabolic variables and EFs, the hypothesis is only partially tested. Indeed, significant correlations (p ≤ 0.01) sporadically emerged between the inhibition index at the BRIEF-P teacher version and the variability of rates over the lifespan: the greater the variability of rates, the more inhibition complaints there are among teachers of HPA children. More specifically, in the PKU subgroup, a negative and significant correlation was found between Brixton preschool and Phe level on day of inclusion. Therefore, the higher the Phe level, the more difficult PKU subjects will be to engage in flexibility skills on this test. Conversely, in children with MHP, a positive and significant correlation was found between the global score of the DCCS and the average rate of Phe during life. It seems that the higher the average rates, the more efficient the flexibility capacities required by this test. Moreover, no correlation was found between the metabolic variables and the 3 executive tests where PKU patients performed worse than MHP patients. This is partly consistent with data from the literature, since several studies have also shown negative correlations between Phe levels at inclusion and EFs in preschoolers with HPA [13, 16, 17]. Similarly, variability in rates throughout life can impact the extent of executive impairment in PKU [12, 16, 30]. However, there is also evidence that average Phe [13] or average Phe/Tyr ratio appears to correlate with some executive impairment in children with PKU [16, 38]. The sporadic presence of correlations does not rule out their import as a predictor of executive difficulties in HPA. However, it is possible that due to the young age of the patients, these metabolic indicators lack sensitivity. Adherence to treatment is likely to be most optimal in the early years of life before greater variability in levels emerges during school age and adolescence [39].

This study has several limitations that must be considered in the scope of the outcomes. Firstly, the sample size remains small (HPA preschoolers = 23) and limits the statistical power of the analyses carried out. It should be noted, however, that the sample remains large for a rare disease and for the age group investigated (3–5 years). Moreover, our population does not include patients lost to follow-up, with potential diet failures. These children have a metabolic control in the management recommendations, which may be a potential bias in the interpretation of our results in view of our hypothesis: the worse the metabolic control the worse the outcome scores should be. Furthermore, an unaffected sibling’s control group would have been ideal for a better matching of educational and socioeconomic level. However, the socioeconomic level was considered in the constitution of the control group, which limits the potential bias. In addition, the comparison of results between the 3 subgroups (PKU, MHP and controls) would have been interesting but it was not privileged because of the variability of the performances on such small numbers and by the random character of the effects that this type of comparison generates. Finally, the tests administered are mostly experimental tests, due to the lack of clinically validated tools for preschool children. Therefore, these tasks may lack sensitivity. The ideal would be to create a standardized battery for the evaluation of executive functions in preschoolers based on a unitary approach to executive functions at this period [18]. To consider the fatigability of the youngest children, we should propose 3 or 4 tests that have already shown their sensitivity in various clinical situations [40], such as the DCCS test [41], the Sun-Moon Stroop [42] or the Hand-Game test [20].. It should also be borne in mind that executive difficulties generally become more pronounced with age and that the impact of metabolic variables also modulates with age, and more difficult compliance with metabolic control [39]. Therefore, it is crucial to investigate EFs in school-aged patients with MHP and PKU, using performance tests and daily life measures, while measuring potential links with different metabolic indicators. Ideally, longitudinal follow-up is the best option, to gain a developmental perspective of the executive profile of PKU and MHP patients.

This study confirms the existence of early executive difficulties in PKU patients, which do not seem to affect MHP children. These difficulties are identified through some performance-based tests but are not perceived by parents or teachers at this stage of development (using BRIEF-P questionnaires). The comparable intellectual profile of PKU and MHP children suggests a possible specific early fragility of EFs in PKU and suggests the importance of their clinical identification in early childhood. Finally, although there is no overall link between the executive profile of the patients and the various metabolic variables considered in this study, the sporadic correlations identified for the Phe level at inclusion, the mean Phe level, and the variability of Phe levels throughout life suggest possible links, which remain poorly predictive for this period of life, but which could constitute useful benchmarks for subsequent neurodevelopmental follow-up. For this, longitudinal studies rather than cross-sectional studies of the same cohort of patients at different ages would be necessary. This would make it possible to refine the dynamics of the onset of executive disorders and their possible causes to propose appropriate treatment.