As reflected in our study cohort, the clinical heterogeneity of PA is broad, ranging from neonatal onset with severe neurologic symptoms to asymptomatic children. However, most patients seem to manifest within the neonatal period. About three quarters of the study participants became symptomatic within the first 5 days of life already. Late-onset forms were rare. Asymptomatic individuals, such as the 13-year-old girl described by Wolf et al. , were the exception even among children identified by newborn screening who received early treatment. Similar observations have been made by other authors. Sass et al.  reported a prevalence of 86% early manifestations in a study of 42 PA patients whereas in a French retrospective study with 17 PA patients 71% were classified as early onset patients . Of 25 PA patients reported by Lehnert et al.  92% presented within the neonatal period.
Poor intellectual development is still the rule in PA. About three quarters of our study patients were mentally retarded. Median IQ was only 55 and the vast majority of patients required special education. This is in accordance with earlier case series [8, 9]. However, our data allow a more positive view when compared to outcome studies of the 1990s [2, 7]. Although improved acute and long-term management have increased the survival rates within the last decades, the neurologic outcome of PA patients is still unsatisfactory. As recently shown, the IQ is negatively correlated with the number of metabolic crises, but not simply with patients’ age .
Neuroradiologic findings are rather nonspecific in patients with PA ranging from white matter abnormalities to widening of sulci and fissures, delay in myelination, brain atrophy and basal ganglia changes [22–24]. In our study, brain imaging yielded pathologic findings in 42% of patients (11/26) with basal ganglia changes being the most common abnormality.
Physical development und endocrinologic aspects
Impaired physical development is a common problem in patients with PA. Sass et al.  and van der Meer et al.  have reported a tendency towards decreased body height in PA patients with a mean height SDS of −1.19 and −2.0, respectively. In accordance with those studies, height was also decreased in our patients especially when compared to patients’ target heights calculated from parental heights. Our data are indicative of an early onset and progressive growth retardation in PA patients. In contrast, patients’ BMI were found to be higher compared to population standards after the first year of life. Failure to thrive has been postulated to be due to iatrogenic dietary protein restriction, frequent infections and metabolic decompensations [21, 25]. However, in our patients growth neither correlated with patients’ daily protein intake nor with the number of metabolic decompensations. To further elucidate the cause of growth retardation, several endocrinologic parameters were determined. Secondary deficiency of IGF-1 can be caused by malnutrition or hypothyroidism . IGF-1 concentrations were decreased in the majority of patients, however, significant correlation with the patients’ body height could not be shown. Touati et al.  also reported low IGF1 concentrations in patients with PA, but growth curves of those patients were within normal limits. Severe hypothyroidism was detected in none of our patients. Abnormal puberty which has been reported as a problem before  was also not found apart from one case of hypergonadotropic normogonadism.
The most common complications in our patients were hematological abnormalities including anemia, neutropenia, thrombocytopenia or pancytopenia, findings that have been described previously [28, 29]. Cardiac complications, especially cardiomyopathy, may contribute significantly to the mortality of PA patients with several fatal cases reported . More recently, the association between PA and long-QT syndrome has been described . Jameson and Walter  reported the first case of a child with PA and long-QT syndrome suffering a life-threatening event secondary to long QT syndrome. In our patients, long-QT syndrome was rather rarely reported (24% of patients) compared to data from an observational longitudinal study with 10 PA patients published by Baumgartner et al.  who found a prolonged QTc interval (>440 ms) in 70% of patients beyond infancy. However, ECGs were only documented for 38/55 patients of our cohort and in most cases, no exercise and 24h ECGs were performed.
Osteoporosis or osteopenia have only been documented in one patient. However, as bone density assessment is usually not done on a regular basis in most patients, osteoporosis might be underdiagnosed.
The rather high incidence of impaired hearing ability in our PA population suggests a connection between both disorders. This is in contrast to an earlier assumption by Brosch et al. , who reported four PA patients with sensoneurinal hearing loss.
Pancreatitis seems to be a rare complication [35, 36] and was documented in two of our patients only. Optic atrophy, observed in one of the patients studied here, has been reported in four patients with PA [37, 38] before.
There is still controversy about the optimum natural protein intake of patients with PA and different therapeutic regimens are used by different metabolic centers.
The median natural protein intake in our patients ranged from 0.8 to 1.1 g/kg body weight per day within the first 5 years, well in accordance with the recommendations given by Sass et al. . Most patients’ diet was supplemented with a precursor-free amino acid mixture. L-Carnitine was supplemented in 93% of the patients. The high frequency of feeding disorders in this series of patients must be stressed and is in agreement with a report by Touati et al. . About half of all patients received tube-feeding to promote growth and ensure metabolic stability.
Quality of life
While the neurologic outcome of PA has been addressed by several case reports and small case series, very little is known about the quality of life of PA patients. In our study, there might be an ascertainment bias because the questionnaire could only be filled in by those patients whose cognitive impairment was only moderate. Among all categories, family life was the top-ranked aspect with regard to the patients’ quality of life. This may well reflect the significant role of the parents and the family ties that can arise by caring for a child that suffers from a chronic, potentially life-threatening disease. Notably, evaluation of the questionnaire data suggests, that most patients consider their quality of life high. Most patients appeared even to perceive themselves as healthy and felt not affected in their daily life. The results of this self-evaluation resemble very much the data of a healthy control group .
From the parents’ perspective, difficulties in the interaction of patients with peers are the most common problem. Notably, according to their parents’ estimate, abnormal findings are more common in PA patients in all categories addressed by the questionnaire when compared to healthy children. This includes emotional issues and conduct problems. One half of all patients were considered to have psychological problems or to show at least borderline abnormalities. The percentage of PA patients who – based on their parents’ view- have psychological problems was about four times higher compared to healthy controls .
Chronic diseases may severely affect family life and pose psychosocial stress to parents and siblings. The results of the FaBel-questionnaire revealed higher values for five of the six subscales for PA patients when compared to the healthy controls reported by Ravens-Sieberer . Those scales include daily and social strains, personal strains and worries about the future, financial burdens and strains on siblings. These findings underline that apart from the medical care affected families also need psychological and social support. Therefore, they strongly advocate for the inclusion of psychologists and social workers into the interdisciplinary metabolic team.