Serum Phe and Tyr concentrations
Our results indicate that Phe concentrations in adult PKU patients are subject to only very slight diurnal fluctuations implicating that a single blood sample may reliably reflect the Phe-control in this group of patients. Variation seems to be small especially in relation to the daily mean Phe concentration. Several studies to evaluate the variations of Phe concentrations in PKU patients have been performed in the past. Studying a population of nine PKU patients (age 1 to 20 years) van Spronsen et al. also observed only small daily fluctuations of Phe concentrations especially when related to the daily mean Phe concentrations . However, amino acid measurements in this study were limited to the first half of the day (8.30 am -1.30 pm).
In contrast, McDonald et al. studied plasma Phe concentrations in 16 children (1 to 18 years) with PKU over a time period of 24 hours : The median difference between highest and lowest concentrations observed (155 μmol/l/day) did not differ much from the results found in our study cohort (132 μmol/l/day). However, as they studied mainly children under a more Phe-restricted diet, fluctuation of the Phe concentrations in relation to the patients’ daily mean values was much higher compared to our data from adult patients. Ferguson studied twelve young PKU patients (9 to 15 years) under different regimens of Phe intake and distribution of protein substitute and found marked differences of serum Phe profiles even within the same group . Both studies lead to the conclusion that single samples give an incomplete and non-representative indication of Phe-control in many children with PKU.
We could confirm earlier observations by van Spronsen et al. that Tyr concentrations show larger diurnal variations than Phe concentrations . The greater variability could be explained by the high content of Tyr in the amino acid supplements taken by the majority of patients and the relatively small plasma pool, resulting in higher proportional changes.
In accordance with earlier reports, we found maximum Phe concentrations for most patients during the morning hours (8 am - 12 am) [8, 10–13]. An overnight rise in Phe concentration is generally thought to be a result of protein catabolism in the fasting state . However, we did not find a linear overnight rise in Phe concentrations in serum as suspected by Farquhar et al. . Nevertheless, as highest Phe and lowest Tyr concentrations were found in the morning, blood sampling for the monitoring of amino acid concentrations in PKU patients should preferably be done at this time of day.
Short-term effects of meals on Phe and Tyr concentrations
In healthy individuals Phe and Tyr concentrations rise significantly after meals . Fingerhut et al. studied amino acid concentrations in dried blood spots of 92 probands (< 1 to 48 years) and found postprandial increases of Phe and Tyr concentrations of 18% and 14%, respectively .
We did not observe a consistent effect of food intake on mean Phe concentration in serum. While after breakfast and lunch a significant rise in Phe levels was observed no such effect became evident after dinner. Notably, in our study setting the patients received their regular diet and the effect of excessive Phe intake was not tested. It needs to be taken into account that mean Phe levels in our patients were much higher than in healthy controls. Therefore, even if the absolute postprandial increase of Phe concentration was the same, the percental change was much lower and thus yielded no statistical significance. In contrast, Tyr concentration increased to more than 300% in individual patients, especially after additional ingestion of amino acid supplements rich in Tyr.
In a study with nine PKU patients van Spronsen et al. also found Phe concentrations to remain rather stable (postprandial increase to 116%) when compared to postprandial Tyr concentrations that were as high as 548% in single patients .
In a different study by the same authors plasma Phe responses to different distributions of the daily Phe allowance over the day were tested in seven PKU patients (1 to 20 years) . Even after single meals containing 75% of the individual daily Phe allowance an increase of mean plasma Phe concentration of only 13% was observed. Extreme peak concentrations of Phe were not detected and the rises were only transient with Phe levels returning to baseline values within two to three hours. Similarly, MacDonald et al. who monitored Phe concentrations in 16 PKU patients (1 to 18 years) over a 24-hour period found no rise in Phe levels in response to Phe consumption .
In a recent study, van Rijn et al. investigated the effect of an additional Phe load on blood Phe concentrations in six adult patients . In this study population Phe concentrations before the study were within the target range of 120–600 μmol/l. Phe loads equivalent to 100% and 200% of each patient’s individual daily Phe intake were given once per week and Phe concentrations in dried blood spots were measured in daily intervals. Mean Phe concentration during the days before the Phe load did not differ significantly from days after the load suggesting that an extra, incidental intake of 100% - and in some cases 200%- of the individual daily Phe intake is tolerated by patients with well-controlled PKU.
Twenty- four- hour variability in blood Phe concentration could also be affected by the amount and distribution of Phe-free amino acid supplement. MacDonald et al. found strong negative correlation between the amount of protein substitute taken by the time of the evening meal and the change in plasma Phe concentrations during the day. The more protein substitute was taken early in the day, the greater was the fall in plasma Phe concentrations during the course of that day . In all of our study patients taking relevant amounts of amino acid supplements the doses were distributed evenly throughout the day. Therefore, we can draw no conclusions on the effects of an uneven distribution on the variability in blood phenylalanine concentrations.
Short-term effects of exercise on Phe and Tyr concentrations
Our short-term endurance training (ergometry) showed no significant effect on mean Phe concentration within the two hours after the exercise. In contrast, the mean Tyr concentration showed a slight, albeit significant increase within one hour and a sharp decrease during the following hour. To our knowledge, no data are available with respect to the short-term effects of exercise on amino acid concentrations in PKU patients so far. It has been shown that in healthy individuals exercise may have a profound acute effect on protein metabolism : Although protein is not normally an important energy source for exercising muscle, there is a significant increase in the rate of amino acid catabolism during exercise . Furthermore, an increase of whole body protein breakdown has been documented in several studies [27–29]. In the post-exercise state, however, whole body protein synthesis occurs , . Phe and Tyr can neither be synthesized nor degraded by skeletal muscle and thus provide a measure of the net rate of protein degradation (i.e. the rate of protein degradation minus the rate of protein synthesis) . An increase in plasma concentrations of both amino acids between 20–90% during exercise has been reported in some studies [31–33] while other authors found Phe and Tyr concentrations to remain stable .
As Phe released by protein breakdown cannot be metabolized normally in patients with PKU, more pronounced effects could be expected compared to healthy persons. However, the fact that we have observed very stable Phe concentrations during and after exercise implicates that short-term sportive activity and the concomitant protein breakdown do not put adult patients at risk for extreme peak concentrations of Phe.
Comparison of amino acid concentrations in serum and dried blood
Amino acid analysis in dried blood spots by tandem mass spectrometry yielded lower concentrations (-28%) compared to the corresponding serum levels. It was tried to reduce common errors in the application of whole blood on filter paper cards by the fact that all dried blood samples were prepared by the same person. The concentration difference between serum and dried blood can be explained by the volume displacement effect  in samples containing cellular matter. Deproteinization and centrifugation of serum eliminates the volume fraction of cellular components and distributes the remaining soluble analytes in a smaller volume, thus resulting in a higher concentration value . Because amino acid analysis in dried blood spots is commonly used in the long-term monitoring and dietetic management of PKU patients, it has to be noted that corresponding Phe concentrations in serum are significantly higher.
Notably, we found an unexpectedly poor correlation between Phe concentrations in serum and dried blood spots, whereas Tyr concentrations correlated to a higher degree. As the same analytical methods and samples were used for both amino acids, this lack of correlation is probably not due to preanalytical or analytical differences. We hypothesize that due to stability of Phe concentrations with fluctuations of only 3%, subtle changes in Phe concentration may have been masked by analytical errors resulting in the poor correlation of the two methods. More pronounced changes in blood Tyr concentration led to better statistical correlation.
Comparison of amino acid concentrations in serum and MD fluid
In our patients the mean Phe concentration measured in MD fluid reached only approximately 60% of the concentration measured in serum. Rolinski et al. studied the determination of amino acid concentrations by subcutaneous MD in nine newborn infants . In contrast to our results, they found higher concentrations of Phe and Tyr in MD fluid compared to plasma (108% and 127% of the corresponding plasma concentration, respectively). Thus, it was concluded that these amino acids are synthesized and/ or released by the subcutaneous tissue . However, it has to be taken into account that tissue composition and metabolic activity of the subcutis in newborns may differ considerably from those in adults . Furthermore, Rolinski et al. studied newborns with hypoglycemia. Since this may prompt proteolysis in order to enable gluconeogenesis, it can be hypothesized that protein catabolism may have contributed to higher tissue amino acid concentrations in the newborns. Independent of the patients’ age the position of the catheter can have a major impact on the relative recovery because vascularisation of the tissue influences the diffusion rate. The position of the catheter in the deep adipose tissue with low metabolic activity would therefore be associated with lower amino acid concentrations in the MD fluid.
We found no significant correlation between Phe concentrations in MD fluid and serum. As already discussed above, the low variability of Phe concentrations could be a complicating factor impeding good correlation results. On the other hand, Rolinski et al. have shown that amino acid values in MD fluid only partly reflect plasma values  and suspected a tissue specific amino acid pattern in subcutaneous tissue. As we have only determined Phe and Tyr concentrations instead of a more comprehensive amino acid pattern, we can neither confirm, nor disprove their hypothesis.
Assuming a possible delay in reaching an equilibrium between amino acid concentrations in serum and adipose tissue, we additionally performed correlation analysis with a delay of one and two hours. Such a time lag is likely because the MD samples were collected continuously over a period of one hour, while blood was drawn at the end of this one-hour period. However, even with these time delays no significant correlation could be found.
In summary, subcutaneous MD does not reflect blood Phe and Tyr concentrations in adults with PKU.