*Article* **Protein Labelling Accuracy for UK Patients with PKU Following a Low Protein Diet**

#### **Dilyana Kraleva 1, Sharon Evans 2, Alex Pinto 2, Anne Daly 2, Catherine Ashmore 2, Kiri Pointon-Bell 1, Júlio César Rocha 3,4 and Anita MacDonald 2,\***


Received: 21 September 2020; Accepted: 27 October 2020; Published: 10 November 2020

**Abstract:** A phenylalanine (protein)-restricted diet is the primary treatment for phenylketonuria (PKU). Patients are dependent on food protein labelling to successfully manage their condition. We evaluated the accuracy of protein labelling on packaged manufactured foods from supermarket websites for foods that may be eaten as part of a phenylalanine-restricted diet. Protein labelling information was evaluated for 462 food items ("free from", *n* = 159, regular, *n* = 303), divided into 16 food groups using supermarket website data. Data collection included protein content per portion/100 g when food was "as sold", "cooked" or "prepared"; cooking methods, and preparation instructions. Labelling errors affecting protein content were observed in every food group, with overall protein labelling unclear in 55% (*n* = 255/462) of foods. There was misleading, omitted, or erroneous (MOE) information in 43% (*n* = 68/159) of "free from" foods compared with 62% (*n* = 187/303) of regular foods, with fewer inaccuracies in "free from" food labelling (*p* = 0.007). Protein analysis was available for uncooked weight only but not cooked weight for 58% (*n* = 85/146) of foods; 4% (*n* = 17/462) had misleading protein content. There was a high rate of incomplete, misleading, or inaccurate data affecting the interpretation of the protein content of food items on supermarket websites. This could adversely affect metabolic control of patients with PKU and warrants serious consideration.

**Keywords:** phenylketonuria; food labelling; protein content; free from; gluten free

#### **1. Introduction**

Phenylketonuria (PKU) is a rare, autosomal recessive inborn error of metabolism due to low or absent activity of the enzyme phenylalanine hydroxylase (PAH), required for degradation of phenylalanine to tyrosine. It causes elevated levels of phenylalanine in the blood and brain and if untreated, leads to severe, irreversible, intellectual disability [1]. Maintaining low blood phenylalanine levels within defined target ranges prevents phenylalanine toxicity [1]. Although it can be managed with a combined approach of dietary and pharmaceutical treatment, the only treatment option in the UK is a lifelong, phenylalanine-restricted diet [2]. Dietary management is stringent, requiring discipline and tenacity, and it is well established that many patients with PKU of all ages are unable to sustain satisfactory blood phenylalanine control [3,4]. Although there are multiple causes for unsatisfactory metabolic control, relatively small deviations from dietary prescription can adversely affect blood phenylalanine levels in patients with classical PKU [5].

Dietary treatment involves the avoidance of high-protein foods such as meat, fish, eggs, cheese, seeds, soya, and nuts and a limited intake of natural protein from foods such as cereals, potatoes, milk, and some vegetables. Any natural protein intake should be calculated, measured, and controlled and up to 80% of patients tolerate <10 g/day [3]. Fruit and vegetables with a phenylalanine content ≤75 mg/100 g, butter, oils, and sugars are given without restriction [2]. Dietary protein is supplemented with synthetic protein, either phenylalanine-free amino acids or low-phenylalanine glycomacropeptide, with added vitamins and minerals to meet nutritional protein requirements. Caregivers and people with PKU are trained in reading and interpreting the protein amounts on manufactured food labels. They are reliant on supermarkets and manufacturers to provide accurate and easily interpreted information about protein content on food labels. Almost every UK supermarket offers an online food delivery service and survey data suggest that 29% of people purchase food via online shopping [6], with online grocery shopping available in at least 60 countries worldwide [7]. Website supermarket shopping is popular for those with special dietary requirements, giving the opportunity to examine food labels prior to food purchase [8]. Patients with PKU and their families can browse the protein nutritional analysis of foods and examine information about food preparation, cooking, and reconstitution of foods. Some online supermarket websites are intuitive to dietary needs and can even create a specific dietary profile that will highlight products that should be avoided for food allergies, although the needs of patients with PKU are not considered [9].

On 25 October 2011, the European Parliament and Council adopted Regulation (EU) No 1169/2011 that issued legal standards for the labelling and information given to consumers by food manufacturers (called the "Food Information to Consumers (FIC) Regulation") [10]. This regulation has been applied since 2016. When pre-packaged foods are sold "online", it is regulated that the responsibility for providing mandatory food information (except the date of minimum durability or the "use by" date) sits with the owner of the online website (the responsible food business operator). Online pre-packaged mandatory food information should include information on the weight and volume of food (net quantity information), a list of ingredients, protein content per 100 g/100 mL, and instructions for use or cooking, if applicable. For non-packaged food, there are fewer rigid stipulations, but the food business operator is required to provide allergen information [11].

In PKU, if foods are eaten because of inaccurate or ambiguous website or food labelling information it may cause unexplained, poor blood phenylalanine control. In countries like the UK, there is high reliance on manufactured foods, so reliability of food labelling information is particularly important. Inadequate, misleading, or unclear information about protein content may deter caregivers/patients with PKU from purchasing specific foods. Therefore, it is in the best interests of manufacturers to supply suitable and trustworthy information that is easy to understand and accurate. The clarity of food labelling and food information on online supermarket websites remains unstudied for people with special dietary requirements such as PKU.

The aim of this study was to evaluate the accuracy of protein labelling on packaged manufactured foods from supermarket websites for foods that may be eaten as part of a phenylalanine-restricted diet. Patients with PKU may use some "free from foods", particularly gluten free, which may have a lower protein content than foods containing wheat or milk, so emphasis was placed on this group of foods.

#### **2. Materials and Methods**

From January 2019 to April 2020, 462 packaged food items were examined using descriptive information given by major UK supermarkets (Asda, Morrisons, Sainsbury's, Tesco, and Waitrose) available on their website. For each food, factors that may alter or affect the protein analysis were tabulated. The selection of foods was not random as this was conducted based on food popularity and common usage. Foods were chosen based on their potential suitability in a protein-restricted diet and mainly had a protein content <10 g/100 g. A selection of "free from" (all gluten-free) and regular foods

were examined. Both commercial branded products and the supermarkets' own brands were included. Meat and fish products were avoided, although two types of regular cheese were included to compare information with "free from" varieties. Items were divided and analysed by the following food groups: bread and bread products, breakfast cereals, vegan cheese, cakes, sweet biscuits, pastries/tarts, crackers, chocolate, crisps, desserts, flours, gravies/sauces, pasta, vegetable foods, dried pot noodles and yoghurts. The supermarket websites accessed were required to give a product description and nutritional analysis for each food.

The following data were collected for each food item: product description, ingredients, preparation, cooking instructions and usage, net and portion size. Specific information collected about protein content included: protein content per portion as cooked/prepared, protein content per portion as sold; protein content per 100 g expressed as cooked/prepared, protein content per 100 g as sold; any misleading information about protein content per 100 g or per food portion (e.g., when a food protein content states 0.0 g per portion, but was >0.1 g/100 g or if the protein content for each portion was only described as <0.5 g with no other relevant information, or if there was a discrepancy between the ingredients listed and the protein content); reconstitution instructions for dry powders, including protein content per portion/100 g supplied when protein analysis was given after dry products had been reconstituted.

All information was transferred onto a database and coded according to the accuracy of information. All products were checked twice by two different dietitians for accuracy and to minimise any risk of bias. Statistical analysis was performed using Mann–Whitney unpaired *t*-tests to compare numbers of misleading, omitted, or erroneous (MOE) foods in "free from" and "regular" food groups and to compare types of MOE information between the two groups. Percentage error in "free from" and "regular" foods were also compared using Wilcoxon signed-rank tests.

#### **3. Results**

#### *3.1. Accuracy of Product Description*

The product description, ingredients, net weight, portion size, protein content per 100 g cooked and uncooked, and preparation and cooking instructions were checked for 462 foods from five supermarket websites (Asda, Morrisons, Sainsbury's, Tesco, and Waitrose). There were 159 "free from" foods and 303 "regular" foods (Table 1). All the "free from" foods were gluten free. Overall, 255 of 462 (55%) foods had information that was MOE from the website product information given by supermarkets, thereby affecting the interpretation of protein content for food items. There were fewer inaccuracies in "free from foods" (MOE, 68 of *n* = 159 foods, 43%), compared with regular foods (MOE, 187 of *n* = 303 foods, 62%) (*p* = 0.007, Wilcoxon signed-rank test) most notably for breads, bread products, and flours.

#### *3.2. Types of Misleading Omitted, or Erroneous (MOE) Information in Food Product Information that A*ff*ected Protein Content*

All types of MOE information are categorised by issue in Table 2. Some food items had more than one descriptive issue that affected interpretation of their protein content.

#### 3.2.1. Food Label did not Distinguish if Protein Content was for Food Item when Cooked/Prepared or as Sold

Thirty two percent (*n* =146/462) of foods required further cooking or preparation. Of these 7% (*n* = 10/146) did not specify whether the protein analysis per 100 g/weight of food was for the cooked/prepared or 'as sold' product. When expressed per portion size this figure was 14% (*n* = 20/146).


**Table 1.** Number of individual foods with misleading, omitted, or erroneous (MOE) information that would affect the protein content calculation from information given on the supermarket websites.

Abbreviations: MOE, foods with misleading, omitted, or erroneous information. \* Mann–Whitney unpaired *t*-test; \$ Pot noodles: a mix of dehydrated noodles, assorted dried vegetables and flavouring powder in a pot. They are prepared by adding boiling water. ID = insufficient data.



\* Mann–Whitney unpaired *t*-test. \*\* total number of foods requiring cooking/preparation *n* = 146; free from *n* = 41, regular *n* = 105.

#### 3.2.2. Missing Information about Protein Content per 100 g (either Omitted Protein Content for Cooked/Prepared Weight or for Uncooked/Unprepared Weight)

Protein content was given for uncooked weight only but not cooked weight for 58% (*n* = 85/146) of foods requiring preparation. In contrast, 33% (*n* = 48/146) of foods gave protein value for cooked but not uncooked weight. These issues were more commonly "regular" foods than "free from" foods (*p* = 0.03 and 0.02 respectively).

#### 3.2.3. Missing Information about Protein Content per Portion Size

Twenty-seven percent (*n* = 125/462) of foods did not give a weight for an estimated portion/serving size; these were more likely to be "regular" foods than "free from" foods (*p* = 0.001). Ten percent (*n* = 47/462) of foods did not give the protein analysis of the portion size. Protein content was omitted per cooked portion size in 35% (*n* = 51/146) of foods or omitted per portion as sold in 26% (*n* = 38/146).

#### 3.2.4. Omitted Cooking Instructions

This information was omitted in 8% (*n* = 12/146) of food items.

#### 3.2.5. Missing Net Size

This information was omitted in 5% of foods (*n* = 21/462).

#### 3.2.6. Misleading/Incorrect Protein Content

Four percent (*n* = 17/462) of foods had misleading protein analysis. Either the protein content per portion size stated 0 g protein when the analysis per 100 g stated a protein content >0.0 g (more commonly in "regular" foods than "free from" foods; *p* = 0.004), or the protein content per 100 g or per portion stated <0.5 g but did not give a specific protein amount. One food had an incorrect protein analysis; this product contained 70% peas, and it was stated that it contained a protein content of only 0.5 g/100 g when it should have contained around 4 g/100 g.

#### 3.2.7. Missing Protein Analysis

Two foods contained no protein analysis (a jelly and frozen potato product). These products were produced by the same manufacturer.

#### 3.2.8. Preparation/Reconstitution Information

Twelve percent (*n* = 18/146) of foods requiring preparation gave the protein analysis only after a product had been reconstituted/prepared with milk even though milk was not part of the ingredients list. Consequently, this "theoretical" protein analysis portrayed these foods to be unsuitable in a low-protein diet. The protein content of the dry ingredients was not given. This was more likely to occur in "regular" foods than in "free from" foods although it did not reach statistical significance.

#### 3.2.9. Omitted Ingredients List

Two percent (*n* = 9/462) of foods did not give an ingredients list.

#### *3.3. Frequency of Misleading, Omitted, or Erroneous (MOE) Food Information for Food Groups and for Individual Foods*

Thirty eight percent (*n* = 96/255) of foods with MOE information had one inaccuracy, 37% (*n* = 95/255) had two inaccuracies, and 16% (*n* = 41/255) had three inaccuracies regarding their information, which affected the interpretation of the food protein content. Bread and bread products, cake, and biscuits commonly had missing information about portion sizes. Pasta and vegetable products regularly had omitted information about the protein content for cooked or uncooked product (either per 100 g or per portion size). Pot noodles were particularly misleading; their protein content was commonly given per 100 g reconstituted weight rather than dry weight, but this was unclear. The protein content of "regular" custards, instant desserts, and some "regular" and "free from" cereals were only given after reconstitution with milk, and commonly had unclear portion sizes.

The frequency of MOE information and the number of problems for the same food items ("free from" and "regular" food items) that would affect their protein content given on the supermarket websites are presented in Figures 1 and 2.

**Frequency of misleading, omitted or erroneous (MOE) for "Free from" food items**

**Figure 1.** Frequency of misleading, omitted, or erroneous (MOE) information for "free from" food items that would affect their protein content given on the supermarket websites.

**Frequency of misleading, omitted or erroneous (MOE) for Regular food items**

**Figure 2.** Frequency of misleading, omitted, or erroneous (MOE) information for regular food items that would affect their protein content given on the supermarket websites.

#### **4. Discussion**

This research indicates that interpreting the protein content for some common supermarket foods available via online websites is inadequate, unclear and even misleading for people with PKU. Information about the protein content per portion size was sometimes omitted or indeterminate,

particularly for "regular" foods compared with "free from" foods. For "regular" dried products requiring reconstitution, the protein content was commonly given only after the product has been prepared with "added" cow's milk, which then increased the protein content of the food, rendering it unsuitable for most people with PKU. For other products consisting of dry ingredients, it was sometimes uncertain if protein labelling was for the dry product or after preparation. For products such as gluten-free biscuits, the protein content was stated as <0.5 g per portion only, even though the protein per 100 g was much higher, and the food item included protein-containing ingredients. Not all products identified net weight.

Food regulations, manufacturers, and online food business operators have not considered the impact of any inaccurate product information for people on very low-protein diets. Fortunately, mandatory FIC nutrition labelling for pre-packaged foods does include protein content, but it is listed only after energy, fat (including saturates), and carbohydrates (including sugar). For non–prepacked foods, there is no requirement in the EU FIC regulations for any nutrition information to be provided, but many manufacturers voluntarily declare the protein content. The FIC regulations states that food manufacturers are not required to do their own laboratory analysis for protein content and it is possible for a food business operator to calculate the values themselves (1) from the known or actual average values of the ingredients used or (2) from generally established and accepted data [12]. The accuracy of protein measurement by these methods is unknown and the definition of what is meant by "generally established and accepted" data is not given, so manufacturers could interpret this in different ways. It is also unknown how many food businesses estimate the protein content by using published protein values of similar foods rather than estimating individual foods by chemical analysis.

Some patients with PKU tolerate a minimal amount of protein (3 to 4 g/day) so accurate protein information is crucial [13–15]. In conflict, the FIC regulations apply protein tolerances to food labels on the basis that protein analysis is not precise due to natural variations in ingredient composition and changes in production. They appear unaware of the needs of patients on very low-protein diets. For foods containing protein <10 g/100 g, they state that the protein content may be within ±2 g; for foods containing protein 10–40 g/100 g, the protein content is ±20%; and for foods containing protein >40 g per 100 g, protein content is within ±8g[16,17]. Additionally, rounding guidance suggested by the EU states that for food containing protein ≥10 g/100 g or 100 mL, the protein should be declared to the nearest 1 g (no decimals); protein between <10 g and >0.5 g/100 g or mL to the nearest 0.1 g; and protein at ≤0.5 g/100 g or mL as "0 g" or "<0.5 g." We identified eight foods, particularly "free from" items, that stated that the food portion contained <0.5 g protein, even though each portion could have contained 0.4 g protein (<0.5 g); this amount would need to be calculated in a very low-protein diet as it may impact on metabolic control. Some patients with PKU have unexplained fluctuating daily blood phenylalanine levels and some of this may be due to the approximate nature of food protein labelling [18].

The FIC regulations state that the nutrition declaration is required for the food as sold, but, instead and where appropriate, it can relate to the food as prepared, provided sufficiently detailed preparation instructions are given. It is therefore possible to include only the nutrition information "as prepared" for foods such as dehydrated powdered soup or desserts. This is deceptive and unsafe for people with PKU. Commonly, we found that nutrition labels for "regular" dessert mixes were calculated based on their preparation with cow's milk, and this should be avoided in PKU. The addition of milk substantially increased the protein content, even though many of the raw ingredients of dessert mixes were low in protein. By declaring protein content after preparation with other added ingredients, products appear unsuitable for patients on a low-protein diet, even though it may have been possible to consume the food product if it had been made up with a low-protein milk alternative. Additionally, giving the protein content of pot noodles after preparation is confusing and this has led to several incidents when caregivers/patients have miscalculated and underestimated their protein content [19].

It was common for supermarket websites to omit information about whether the protein analysis was associated with cooked or uncooked food. The protein content of a food product will vary depending on if it is dry cooked, fried, microwaved, or uncooked [20]. Commonly, foods such as potatoes have a high water content, and dry cooking results in moisture loss and a more concentrated protein amount [20]. These protein differences must be considered in a low-protein diet. The FIC regulations state that instructions on how to prepare and cook the food, including heating in a microwave oven, must be given on the label if they are needed [21–23]. If the food must be heated, the temperature of the oven and the cooking time should usually be stated, so it is sensible to give food analysis both for the "as sold" state and for "cooked", as recommended.

The protein content of a portion size was either omitted or the portion size was not quantified by weight for 172 of 462 foods (37%), particularly in "regular" foods, contributing to the difficulty in calculating the protein content of foods consumed in a phenylalanine-restricted diet. The FIC regulations state that the portion or consumption unit should be easily recognisable by the consumer, quantified on the label in close proximity to the nutrition declaration, and the number of portions or units contained in the package must be stated on the label. The "consumption unit" information requires improvement.

Legislation on food labelling gives instruction to producers and retailers; it also gives the consumers rights to basic information. We have shown that the information on the protein content of foods via supermarket websites is inaccurate and potentially harmful to those with PKU. Unlike allergies, there is little understanding of the essential role of a very low-protein diet and the harmful impact of poor control on patients' neuropsychological health [9]. This may also apply to other patients with inherited disorders of protein metabolism such as Maple Syrup Urine Disease or Tyrosinaemia type I or II. They also rely on accurate food labelling to manage their dietary treatment safely.

There appears to be no audit or regular assessment of supermarket websites to check accuracy of information that is provided to the consumer. Manufacturers should indicate on food labels how they have estimated protein content. We identified a packaged food product containing 70% peas and 30% carrots (with no other added ingredients), but it stated that it only contained protein 0.5 g/100 g, when it should have contained a protein amount of around 4 g/100 g (based on the established protein content of peas). For children to be given a food they enjoy in error, leads to additional psychological stress and guilt for the parents. We identified two other packaged products made by the same company without any protein analysis. These products were targeted at young children, so likely to be mistakenly eaten by a population very vulnerable to the impact of high blood phenylalanine concentrations.

This study did have some limitations. Although almost 500 foods were examined, matched numbers and types of "free from" foods were not compared with regular foods. However, as an overall group, "free from" foods website supermarket information gave more comprehensive data that would enable the consumer to assess the protein content of the product consumed. This was commonly due to the low availability of some "free from" foods, such as pot noodles or dessert pots. There was not an equal number of foods examined in all the different food groups, with small numbers of the following products examined: cheese, yoghurts, ice cream, and vegetable products. Food products in this study were not chosen by random, but commonly selected in order of popularity and usage, so it is accepted that there are limitations in product selection, especially with gluten-free or "free from" products that are usually purchased by patients with coeliac disease or food intolerance. Website information was not compared with product labelling on packages as purchased from the supermarket shelves, which may have identified further discrepancies.

#### **5. Conclusions**

Obtaining accurate information about the protein content of some foods from online supermarket website information is challenging. A high proportion of incomplete, misleading, or inaccurate data was identified that directly affected the interpretation of the protein content of food items. Inadequate protein food labelling is likely to contribute to the difficulties in maintaining good metabolic control in PKU. It is important that all dietitians, patients, and families of patients with PKU are aware of the food label limitations and potential problems.

Although food producers and business operators are expected to provide information to consumers that is clear and accurate, little attention is paid to the exactness of protein food labelling. The FIC regulations should be reconsidered, with more attention given to monitoring the accuracy of information provided by supermarket websites. Poor awareness of the impact and inattention to the factors that affect food protein content and carelessness about the accuracy of protein labelling can adversely affect the neurological health of people with PKU and deserves urgent consideration.

**Author Contributions:** Conceptualisation, D.K., A.M., and S.E.; methodology, D.K., A.M., and S.E.; formal analysis, D.K., S.E., A.P., and A.M.; investigation, D.K.; data curation, D.K.; writing—original draft preparation, D.K.; writing—review and editing, D.K., S.E., A.P., A.D., C.A., K.P.B., J.C.R., and A.M.; visualisation, D.K., S.E., A.P., A.D., C.A., K.P.-B., J.C.R., and A.M.; supervision, A.M. and K.P.-B.; project administration, A.M. and K.P.-B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors have no conflicts of interest to declare related to the paper's content.

#### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Review* **Protein Substitutes in PKU; Their Historical Evolution**

**Anne Daly \*, Sharon Evans , Alex Pinto, Catherine Ashmore and Anita MacDonald**

Birmingham Women's and Children's Hospital, NHS Foundation Trust, Birmingham B4 6NH, UK; evanss21@me.com (S.E.); alex.pinto@nhs.net (A.P.); catherine.ashmore@nhs.net (C.A.); anita.macdonald@nhs.net (A.M.) **\*** Correspondence: a.daly3@nhs.net

**Abstract:** Protein substitutes developed for phenylketonuria (PKU) are a synthetic source of protein commonly based on L-amino acids. They are essential in the treatment of phenylketonuria (PKU) and other amino acid disorders, allowing the antagonistic amino acid to be removed but with the safe provision of all other amino acids necessary for maintaining normal physiological function. They were first formulated by a chemist and used experimentally on a 2-year-old girl with PKU and their nutritional formulations and design have improved over time. Since 2008, a bioactive macropeptide has been used as a base for protein substitutes in PKU, with potential benefits of improved bone and gut health, nitrogen retention, and blood phenylalanine control. In 2018, animal studies showed that physiomimic technology coating the amino acids with a polymer allows a slow release of amino acids with an improved physiological profile. History has shown that in PKU, the protein substitute's efficacy is determined by its nutritional profile, amino acid composition, dose, timing, distribution, and an adequate energy intake. Protein substitutes are often given little importance, yet their pharmacological actions and clinical benefit are pivotal when managing PKU.

**Keywords:** phenylketonuria; protein substitute; amino acid; glycomacropeptide

A.; Ashmore, C.; MacDonald, A. Protein Substitutes in PKU; Their Historical Evolution. *Nutrients* **2021**, *13*, 484. https://doi.org/

**Citation:** Daly, A.; Evans, S.; Pinto,

Academic Editor: Jamie I. Baum Received: 29 December 2020 Accepted: 28 January 2021 Published: 2 February 2021

10.3390/nu13020484

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Amino acids are unique substrates providing nitrogen, hydrocarbon skeletons and sulphur [1]. They are essential precursors for the synthesis of proteins, peptides, and low molecular weight substances such as glutathione, dopamine, nitric oxide, and serotonin [1]. In phenylketonuria (PKU), dietary treatment was made feasible with the introduction of low/free phenylalanine synthetic proteins (protein substitutes), that have gradually advanced with time. In the 1950s, these were originally derived from protein hydrolysates, but in the 1970s, phenylalanine-free amino acids were introduced. Protein substitutes provide the building blocks of tissue proteins and their amino acids are essential for the synthesis of hormones, enzymes, and other cellular processes. Therefore, their composition and nutritional profile is fundamental, helping prevent neurological devastation, allowing normal growth and biosynthetic functions. The original technology for making protein substitutes was crude and limited but now precision manufacturing has improved their quality.

Although Følling [2] first identified phenylpyruvic acid in the urine of untreated children with PKU, it was Penrose who recognised that it was a genetic recessive disorder and named it phenylketonuria (PKU) [3]. He was also the first to try a dietary treatment based on fruit, sugar, olive oil, and vitamins, but this protein-free diet lacked essential phenylalanine and all other amino acids, resulting in malnutrition and so the treatment was abandoned [4]. Twenty years later, protein substitutes were introduced, but their central role in the management of PKU remains undervalued.

#### **2. Early Studies**

Følling and Penrose [5] both demonstrated that giving phenylalanine to a PKU subject increased the excretion of phenylpyruvic acid. The type of phenylalanine ingested as

D or L isomers had different effects on phenylpyruvic excretion, with L phenylalanine leading to a greater production of phenylpyruvic acid. Similarly, in non-PKU subjects, L phenylalanine was the preferred metabolised substrate, with D and DL isomers leading to small amounts of phenylpyruvic acid but an absence when the L form was given due to its complete metabolism. From these studies, they concluded that phenylpyruvic acid excreted in PKU patients was due to an incomplete breakdown of phenylalanine. Tyrosine when administered had no effect on urine phenylpyruvic acid excretion, concluding this was metabolised normally. It was not until 1944 that Bernheim [6] demonstrated that the main metabolic pathway for phenylalanine was by parahydroxylation of phenylalanine to tyrosine. In 1953, Jarvis showed that it was the inability to perform this hydroxylation that resulted in phenylketonuria [7].

Penrose and Quastel [5] conducted a series of feeding studies where they found that by lowering the natural protein intake by >50% resulted in an immediate reduction in urinary phenylpyruvic acid in a patient with PKU. However, after the second day of treatment, urine phenylpyruvic acid re-appeared and increased over subsequent days. The authors noted a weight loss over the same time and hypothesised that catabolism led to the production of phenylpyruvic acid.

In 1951, a positive ferric chloride screening test in a symptomatic 2-year-old girl from Birmingham, UK, preceded the first successful dietary treatment in PKU. She was only the third child to be tested with the ferric chloride test at Birmingham Children's Hospital [8]. In PKU, phenylpyruvic acid present in urine causes the characteristic greenish-blue colour reaction when a few drops of ferric chloride are added [9]. On presentation, she was unable to talk, walk, or engage with her surroundings; her mother waited for the doctor every morning outside the hospital laboratory as she refused to accept that there was no treatment for her daughters' condition. Louis Woolf designed the first successful protein substitute formulation used in PKU. He was a chemist with a commercial background and had used hydrolysed casein to produce amino acids as a treatment for malnutrition after the Second World War. Cost was a priority in the post war years and protein hydrolysates were readily available and cheaper than pure amino acids. In 1949, he suggested that supplementation of carbon treated casein hydrolysate with appropriate amounts of missing amino acids (including a source of phenylalanine to prevent deficiency) could treat PKU. He was unable to convince his medical colleagues at Great Ormond Street (GOS) Children's Hospital, London to try his proposed treatment. He recalls: "*At GOS, the suggestion floated like a lead balloon, I was told not unkindly that I should be devising new diagnostic tests, not dreaming up crazy treatments for conditions that everybody knew were untreatable*" [10]. In collaboration with Drs Bickel, Hickman, and Gerrard from Birmingham, a modification of Louis Woolf's protein substitute was given to the 2-year-old child with PKU [11].

#### **3. The First Protein Substitute**

Casein was hydrolysed by boiling it for several hours with concentrated hydrochloric or sulphuric acid to produce a thick black amino acid liquid. This solution was neutralised with sodium hydroxide and then purified by the addition of activated carbon and finally filtered to produce a clear solution of amino acids. This solution contained phenylalanine, which was removed by a second filtration method using activated charcoal. This removed the aromatic amino acids: phenylalanine, tryptophan, and tyrosine (although a small residual amount of phenylalanine was detectable). To nutritionally improve the protein substitute carbohydrate, fat, vitamins and minerals were added, together with tryptophan and tyrosine [11]. This unpalatable solution was then mixed with sugar, wheat starch, double cream and water and given as a formula to infants. In older children, it was either flavoured with tomatoes and given as a soup [12], made into a blancmange with sugar, margarine and wheat starch [13] or mixed with vegetable oil, and sugar and flavoured with artificial flavourings [14].

The production of the original formula was difficult and time consuming and had to be done in a cold room or it would deteriorate. The black charcoal covered everything and as the first formula was prepared, the sight of Dr. Bickel wrapped in layers of jumpers topped by a charcoal smudged lab coat became a common sight [8]. Woolf identified that a small amount of phenylalanine should be added to the formula as it was an essential amino acid [15]. He stressed the need for careful monitoring, and he was also the first to propose treatment for life in PKU [12,16,17].

Phenylalanine-free amino acids as a protein substitute for PKU were first tried in the USA in the 1950s [18], but had to be abandoned most likely due to the pure amino acid mixtures causing vomiting.

#### **4. Amino Acid Requirements**

In the early stages of making the protein substitute, the exact amino acid composition of the casein hydrolysate was unknown. Casein was low in sulphur containing amino acids and cysteine was also partly removed by hydrolysis and charcoal filtration. Bickel [19] suggested adding L cystine to the hydrolysate and Woolf proposed the addition of DL methionine [20].

The amounts of tryptophan and tyrosine added to the first protein substitutes were determined by amino acid requirements established in the early 1950s; this knowledge was pioneered by Rose [21–25], Holt and Snyderman [26,27]. It was estimated that an adult man required 1000 mg/day of L phenylalanine to maintain nitrogen equilibrium or 300 mg/day if tyrosine was provided [24]. Synderman [28] suggested that 90 mg per kg/day of L phenylalanine was needed by an infant, but this was reduced to 25 mg/kg per day if sufficient tyrosine was supplied. Other essential amino acid requirements were estimated from the work of Rose, Leverton [23,29] and Swendseid [30]. In Woolf's [12] original formula, 25 mg/kg/day of L tryptophan and 25 mg/kg/day of L methionine were given in addition to 50 mg/kg/day of L tyrosine, a surrogate essential amino acid in PKU (Table 1).


**Table 1.** The original composition of the first protein substitute designed by Louis Woolf (1958).


There were challenges when administering the artificial diet in PKU [11]. The first child to start treatment was admitted to hospital for 6 weeks. The musty smell associated with PKU disappeared, plasma and urinary phenylalanine concentrations returned to normal, and there was a negative ferric chloride test when the diet was commenced. However, the child lost weight and within 5 days of treatment, plasma tyrosine concentrations were un-recordable (with a change in hair pigmentation), and plasma phenylalanine was raised. Tyrosine (1.5 g/daily) was added, correcting the low plasma tyrosine concentrations and temporarily arrested weight loss. However, after a further 3 weeks, aminoaciduria was noted, and in the fifth week, blood phenylalanine increased and phenylpyruvic acid reappeared in the urine; this was associated with weight loss, vomiting, and the child was described as unwell. These observations were important, highlighting that tyrosine became an essential amino acid in PKU as a consequence of the biochemical block in converting phenylalanine to tyrosine. Aminoaciduria, a result of weight loss and catabolism due to phenylalanine deficiency, led to an increase in catabolism and a subsequent increase in blood phenylalanine. Adding a measured amount of phenylalanine back into the diet (typically 250–500 mg or equivalent to around 5–10 g protein/day) increased plasma and urine concentrations, but to levels significantly below pre-treatment concentrations. Laboratory analysis was laborious, each blood test was analysed in duplicate and the production of a chromatogram took 3 days of intensive labour [8].

After 6 months of treatment, this child made remarkable progress; followed by a cascade of successful case studies. Woolf [17] reported 3 cases, with a further publication of 10 cases in 1958 [12], in which children were treated from the age of three weeks to 5 years of age. Armstrong and Tyler [31] reported the treatment of five children, and Armstrong and Binkley in 1956 [32] followed the progress of an infant starting treatment at 40 days of age. All reported that a low phenylalanine diet, supplemented with a low phenylalanine protein hydrolysate corrected the major biochemical abnormalities.

It was also established that sufficient carbohydrate and fat (including a source of linoleic acid) was necessary to prevent protein catabolism [20,33–35]. Woolf reported that the daily intake of hydrolysate should be high correcting for the inefficient utilisation of the amino acids [12].

#### **5. Commercial Protein Hydrolysate Preparations**

Production of the hydrolysate moved from hospital laboratories to commercial production in late 1953/early 1954. In Europe, Cymogran 1954/5 (Allen and Hanbury, London, UK), XP Albumaid (1960) (Scientific Hospital Supplies, Liverpool, UK) and Minafen (designed for infants in 1955), (Trufood Ltd., Guildford, UK) were developed and the US produced Lofenalac (Mead Johnson, Chicago, IL, USA) in 1958. In the spirit of commercial interest, Trufood and Allen and Hanbury agreed to share production with one company making an infant substitute Minafen (Trufood)and the other (Allen and Hanbury) a preparation for older children Cymogran. Limited practical instructions were provided on how to reconstitute these formulas and families had to weigh the prescribed powder. The main difference between hospital and factory production was the use of ion exchange resins to separate phenylalanine, dispensing with the sodium hydroxide and carbon filtration. These synthetic filters consisted of microbeads from resin or polymers, allowing the separation and purification of the hydrolysed casein. These products were supplemented with variable amounts of vitamins, minerals, carbohydrate and fat.

#### **6. The First UK PKU Guidelines**

In 1960, the UK Ministry of Health [9] provided guidelines on screening and early detection of PKU, together with recommendations on optimal blood phenylalanine concentrations and provision of protein substitutes. They proposed screening by the ferric chloride test at 4–6 weeks of age (which was later replaced by the Guthrie method in 1969 [36]). To prevent phenylalanine deficiency, a target blood phenylalanine concentration slightly above normal was recommended (90–120 mmol/L), with blood phenylalanine monitoring done twice weekly until stability was achieved, and then weekly or monthly monitoring was required.

In infancy, a protein substitute, formulated and reconstituted similar to regular milkbased infant formula was recommended. A second protein substitute with a lower energy content was advocated for older children.

#### **7. Nutritional Deficiencies with Early Protein Substitutes**

In the early history of treating PKU by diet, there were concerns about 'over- treating' patients and maintaining very low phenylalanine blood concentrations. Nutritional deficiencies, malnutrition, and even death were linked to dietary treatment [37]. In the 1960s, severe skin rashes in babies on Minafen (Allen and Hanbury Ltd., London, UK) were reported [38–40]. Woolf [12] described a child with faltering growth and hair loss when acetyl DL tryptophan was accidentally given instead of DL tryptophan; stopping the acetyl derivative immediately reversed the symptoms. Studies in animals fed synthetic low phenylalanine diets [41] led to the addition of choline, riboflavin, folic acid, and vitamin E to the hydrolysate preparations. Two reports of folic acid deficiency were described [42,43], one child had megaloblastic anaemia due to folic acid deficiency exacerbated by vomiting and poor feeding and subsequently died. Hypoglycaemia was also reported in two cases [44].

#### **8. Amino Acid Preparations**

In the late 1960s, commercial amino acid formulas were made from pure crystalline amino acids by fermentation of bacteria. They were manufactured by the Japanese at an affordable cost [45]. The first product for PKU was Aminogram Food Supplement [46,47], which had several advantages, compared to hydrolysed formulas including improved taste and a lower daily volume, with an amino acid composition that could be easily adapted for the treatment of other aminoacidopathies such as maple syrup urine disease, homocystinuria, and tyrosinaemia type 1.

Manz [48] reported anorexia and vomiting in some infants given amino acid preparations. Metabolic acidosis was observed when the preparations contained amino acids in the form of hydrochloride salts or when the ratio of sulphur containing amino acids was too high, leading to higher urinary pH and increased renal net acid excretion. Modifications in the amino acid preparations normalised the renal net acid excretion and acidosis.

The early commercial preparations of L amino acid substitutes required separate supplementation with vitamins and minerals and careful monitoring of nutritional status was essential. These vitamin and mineral supplements were commonly deficient in molybdenum, chromium, selenium, and pantothenic acid [49,50].

**L-amino acid substitutes nutritionally complete:** In 1980, the first UK amino acid preparation supplemented with carbohydrate, vitamins, mineral, and trace elements and designed for children over 1 year of age with PKU was manufactured. It was flavoured for improved taste and palatability. In 1988, a similar product (but also with added taurine and carnitine), but formulated for children over the age of 8 years and suitable for maternal PKU, was introduced [51]. From the 1990s, further advances were made in the nutritional formulations, taste, and presentation of protein substitutes (Table 2). Although selenium supplementation was added to protein substitutes from the late 1980s, many countries were wary about adding selenium to protein substitutes due to concerns about its toxicity which had been responsible for deaths in man and animals and was referred to as the 'essential poison' [52]. Consequently, this led to reports of many cases of biochemical selenium deficiency [53,54].


**Table 2.** Introduction of protein substitutes.


It was also established that the fat intake of children with PKU was low [55] and n-3 long chain polyunsaturated fatty acid status was sub-optimal [56]. This led to the addition of essential fatty acids to protein substitute powders designed for children [57]. Around the same time, long chain polyunsaturated fatty acids were added to infant protein substitutes in 2000; in a double blind randomised study, infants received either a formula with or without a supplemented fat blend of long chain polyunsaturated fatty acids (LC-PUFA). The results clearly showed the benefit of supplementation [58] and led to the addition of LC-PUFA to other products designed for older children.

Over the years, there has been much endeavor to ensure that protein substitutes meet changing nutritional trends and accommodate nutritional requirements according to life stage. In 2011, the first phenylalanine-free infant formula containing a specific mixture of prebiotic oligosaccharides was introduced [59]. This helped maintain levels of bifidobacteria and lower stool pH in infants with PKU. There is concern about increasing obesity rates in the PKU and non PKU population, so many recent protein substitutes introduced for children, teenage, and adults with PKU have a lower carbohydrate and energy composition [60]. Impact on lowering obesity rates has not yet been proven.

The nutritional adaptation of protein substitutes in order to gain clinical benefit is an area likely to grow in the future. Recently, specific nutrient combinations (containing uridine monophosphate, docosahexaenoic acid, eicosapentaenoic acid, choline, phospholipids, folic acid, vitamins B12, B6, C, and E, and selenium) have been studied in PKU mice to examine the impact on synaptic deficits in PKU [61]. The specific nutrients are precursors and cofactors for the synthesis of phospholipids thought to be beneficial in improving the neurotransmitter/synaptic changes in PKU. This combination of nutrients has been shown to have a benefit on synapse formation, morphology, and function in mouse models of Alzheimer's disease so it may be an important nutritional adaptation of protein substitutes for older patients [62].

#### **9. Choice of Protein Substitutes**

The choice, composition, and presentation of protein substitutes have expanded at fast moving rates since the turn of the century. This time was associated with the evolution of pre-packaged and premeasured products, which has not only improved convenience but also accuracy, adherence, and ease of protein substitute prescription for clinicians.

**Spoonable low volume protein substitutes:** An innovative substitute designed for young children with PKU was produced in 2000, based on phenylalanine free amino acids and starch to which a small amount of water was added, forming a gel/paste that was similar in consistency to a weaning food [60]. This low volume, fat free, lower calorie, more concentrated amino acid substitute had the advantage of allowing transition onto a second stage product from the age of 6 months, in line with complementary feeding. It was presented in premeasured sachets (dispensing with the need for large tins of formula), was easy to prepare, with a good consistency and acceptable taste. Normal infant feeding behaviour, teething, and intercurrent infection can lead to its rejection in late infancy so perseverance and a consistent approach is needed by parents [63,64].

**Ready to drink liquid protein substitutes:** In 2005, 'ready to drink' flavoured phenylalanine free amino acid pouches were introduced. These small volume, lower energy protein substitutes were convenient and compact, allowing greater independence for children and teenagers. Patients were less self-conscious taking a liquid drink compared to a powdered preparation [65]. The pharmacological efficacy of these lower volume substitutes did not compromise nutritional biochemistry or phenylalanine concentrations, which remained the same or improved [60]. One potential problem was abdominal discomfort (constipation/diarrhoea), attributed to the hyperosmolar concentration of the lower volume protein substitutes [66,67], and like all concentrated amino acid products, they should be administered with additional water.

**Protein substitute tablets:** Amino acid tablets and modular systems were also introduced around 2000. Modular systems are when a combination of amino acid tablets, capsules, liquids, powder, or bars of amino acids are used to provide daily protein substitute requirements, allowing flexibility of choice. In a randomised crossover study, it was shown that subjects with PKU successfully took at least 40% of their protein substitute as tablets, with an improvement in adherence and significantly lower blood phenylalanine concentration [68]. The quantity of tablets to meet protein requirements was around 70/day, and they were not nutritionally complete, requiring extra supplementation with vitamins and minerals. They provide an alternative for older children and adults who struggle taking conventional protein substitute. Micro-tablets of amino acids have since been introduced.

**Caseinglycomacropeptide with amino acids (CGMP-AA)**: CGMP-AA was introduced in the UK in 2017, although first used in the USA as a protein substitute for PKU in 2008 [69]. CGMP is purified from whey by anion exchange chromatography, but the final product does contain residual amounts of aromatic amino acids including phenylalanine [70]. CGMP-AA is different from amino acid substitutes; approximately 40% of the product is composed of amino acids, with the rest as a bioactive peptide; based on a macropeptide, they are associated with improved taste and palatability [71].

**Slow-release protein substitute:** A prolonged release product was first developed in 2014 [72] but there was little supporting published data demonstrating its effectiveness. In 2018, a slow release preparation containing amino acids coated with ethyl cellulose and alginate was introduced. Based on physiomimic technology, the bitter taste and smell of amino acids was improved, and as this product is not mixed with fluid, it does not have an osmolality. Most importantly, the technology prolongs the release of amino acids into the systemic circulation. Animal and human kinetic studies demonstrate a reduced peak concentration of amino acids. This new technology suggests a physiological absorption of amino acids similar to natural protein [73,74]. In a short-term observational study using prolonged amino acids in subjects with PKU, it was well tolerated, with fewer gastrointestinal symptoms and no change in blood phenylalanine concentrations [75].

#### **10. Pharmacological Importance of Protein Substitutes**

The amount of protein equivalent (g/kg) from protein substitutes affects blood phenylalanine control [76–79]. As early as 1961, an observational study performed by O'Daly [80] showed protein substitutes significantly lowered blood phenylalanine concentrations. Further studies have shown that phenylalanine tolerance is increased when total protein intake from a protein substitute is increased [79,81].

Protein substitutes have an important function at the blood brain barrier. Large neutral amino acids including phenylalanine compete for LAT1, a large neutral amino acid transporter allowing entry of amino acids into the brain [82]. Phenylalanine has a particularly high affinity for LAT1 and protein substitutes are the only source of competitive large neutral amino acids necessary to prevent excess phenylalanine entering the brain. These pharmacological effects of ingesting an amino acid rich formula are frequently neglected and given little scientific credence, and yet they have a significant impact on phenylalanine metabolism and long-term physical and neurological outcome. The gut also controls the absorption of amino acids across the epithelial membrane. Phenylalanine is transported as a carrier mediated sodium dependent process which requires energy. Similar to the blood brain barrier, large neural amino acids are transported in the gut by LAT1, also known as SLC5A7, for which phenylalanine has a high affinity [83,84].

#### **11. Protein Substitute Requirements**

Human requirements for each amino acid are specific to age, metabolic demands (immune/neuromuscular), and growth rate (protein deposition) [76,85,86]. For protein synthesis to occur, all the amino acids should be available; absence of one leads to the cessation of synthesis [1]. Snyderman [87] reported that the complete withdrawal of phenylalanine from the diet in a normal infant led to the depression in several other amino acids, the most prominent being tyrosine, hence the importance of tyrosine supplementation. Woolf showed that the nitrogen content of the artificial substitute was not an exchange for natural protein; the hydrolysate contained less nitrogen, was rapidly absorbed from the gut with greater oxidation and urinary amino acid losses [88,89], therefore sufficient product was needed to meet nitrogen requirements.

A protein substitute intake that just meets minimum WHO requirements (WHO/FAO/ UNU 2007) [90] may result in 'latent' catabolism, leading to body tissue breakdown, increasing phenylalanine concentrations. Protein utilisation is enhanced by a supply of carbohydrate and fat [91,92] further illustrated in a randomised controlled study in PKU subjects by MacDonald [93] and supported by Illsinger [94].

The European PKU Guidelines recommend that the total protein intake should supply 40% more than the FAO/WHO/UNU safe levels of protein intake [95]. However, this amount is arbitrary and unconfirmed by research [67]. A collaborative study [96] involving 63 European and Turkish IMD centres concluded that the amount of total protein prescribed by different European countries was not uniform. All centres gave higher protein equivalents than the recommended 2007 WHO/FAO/UNU [90] safe levels of protein intake with Western European centres prescribing less total protein then other European regions.

To maximise the utilisation of amino acids and minimise the variation in phenylalanine concentrations, protein substitutes should be taken frequently, a minimum of three times a day. MacDonald [97] demonstrated that the greater the amount of protein substitute consumed between waking and 4 p.m., the greater the decrease in phenylalanine concentrations. Likewise, when protein substitute was given 4 hourly for 24 h, there was a marked stablisation in phenylalanine concentrations, reducing phenylalanine variability [98].

Tyrosine, a precursor of catecholamine neurotransmitters (dopamine, norepinephrine, and epinephrine), thyroxine, and melanin, is an essential amino acid in PKU due to the limited or absent hydroxylation of phenylalanine. It is hydrophobic and the absolute quantities added to protein substitutes are not defined. Indicator amino acid oxidation studies [99] suggest tyrosine should provide 19 mg/kg/day, although current protein substitutes provide approximately 5 times above current recommendations. The importance of tyrosine was recognised by the Report of the Medical Research Council working party on PKU [100], which recommended that protein substitutes should be nutritionally complete and contain 100–120 mg/kg/day of tyrosine.

#### **12. Protein Substitute Administration**

In the early history, the practicalities of administering an acid based hydrolysed unflavoured product were particularly challenging. Bentovim [46] described the struggles families faced trying to persuade children to take the acid tasting formula: the large daily volume that needed to be consumed, regular vomiting, refusal to eat permitted food due to negative associations with the substitute, the bad smell, and lack of palatability. Furthermore, children experienced isolation and psychological difficulties particularly

in the school years. This was one of the factors leading to diet cessation as early as 6 years [101–103]. An extract from the *Cork Examiner* describes the struggles faced by one family adapting to the news that their two children had been diagnosed with PKU and the dietary changes and challenges made to improve their neurological outcome [104].

Despite the advances in technology, almost all protein substitutes have a strong taste and odour and are associated with poor palatability and breath odour. They are a burden to patients as they must be consumed a minimum of three times daily and spread evenly throughout the day. Ford [66] reported 293 of 631 participants with PKU (39% of adults, 11% of children) either did not take protein substitute or took less than their prescribed amounts.

Verbatim extract from study: *Our greatest struggle is getting our son taking his protein substitute. He refuses to take it and it can take up to 45 min for him to finish one with a lot of upset*.

Evans [105], in a case control study in PKU children of weaning age, highlighted the stress, anxiety, and struggles associated with protein substitute administration. Maternal anxiety regarding child rejection of protein substitute increased with time peaking at 12– 24 months. Similarly, in 2016, MacDonald [106] reported in 114 children with PKU, dietary management was associated with a considerable time and financial burden for caregivers, with much time spent supervising protein substitute intake.

#### **13. Conclusions**

In PKU, the early pioneers understood the physiological importance of protein substitutes. They stressed the need for a balanced amino acid profile, for even administration throughout the day, together with an adequate energy intake and dietary treatment for life. Although these principals remain unchanged 70 years later, each decade has witnessed improvements in the delivery and nutritional composition of protein substitutes, which remain of fundamental importance in the treatment of PKU. Further changes are needed, to deliver improved taste and odour-free products, with the properties of natural protein delivering a stable chemical environment associated with optimal physiological function and patient tolerance.

An extract from the *Irish Cork Examiner* describes the struggles of a family diagnosed with PKU in 1959, the son aged 4 and the daughter 2<sup>1</sup> <sup>2</sup> years old. This extract describes the determination, sacrifice, hardship, and success against the better judgement of expert advice. 25 October 1962. Phenylketonuria: A story of heartbreak and hope.

*"Treatment might help your daughter" he said "but for your son detection has come too late." He would deteriorate so much that at a later stage institutional care was inevitable. No one had attempted treatment on a child over 2 years. But the specialist was willing to give my little girl a trial. I pleaded for both of them not knowing the terrible struggle this entailed.*

*The boy was difficult, backward and had no speech while his sister could neither walk or talk and was unable to sit up alone and was extremely difficult to manage. Those first months of 1959 were a nightmare from which there was no awakening. The introduction of the unpalatable diet and the cessation of stews, broths and chocolate sundaes brought tears and tantrums. How I dreaded the ice cream vendors that first summer and the laughing lolly licking youngsters who stood on our corner. The synthetic protein (Minafen) was unpleasant to take, but I have found it can be disguised reasonably well in savouries and cookies.*

*Each child is allowed approximately 270 mg of phenylalanine per day according to body weight. If the child were to have 1oz of porridge oats this would cover 241.5 mg of their daily allowance, whereas 1oz of tomatoes would only represent 5 mg, so planning meals for them was at one time a highly complicated business. Now I possess a simplified chart of foods with a low phenylalanine content and by drawing on almost every cook book in print for ideas I have complied my own Cook Book for Phenylketonurias. Special gluten free flour must be used for bread making and Kosher margarine replaces butter, bread* *making with wheat starch was a different matter "Neolite or just plain leather "was my husband's query at my first attempt. Meals for ourselves present a real problem. It is so difficult to take a hearty T bone steak or peach meringue while two pair of eyes watch with longing. Meals out are impossible as is home entertainment, but it has been so worthwhile, the little girl unable to talk or walk takes some chasing and her speech is coming slowly. Responsiveness and alertness have taken place in slow but sure degrees. My son now 7 has made remarkable progress benefiting from a normal education.'*

*Although detection of PKU and treatment soon after birth is essential for complete recovery we have proved beyond all doubt that much can still be done.*

**Author Contributions:** Conceptualization, A.M., A.D.; writing original draft preparation, A.D., A.M. Writing review and editing, A.D., S.E., A.P., C.A. and A.M.; supervision, A.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** We would like to thank R. Cowen, P. Cowen, J. Cowen, P. Portnoi and C. Clothier for their time and contribution to the historical facts in the manuscript.

**Conflicts of Interest:** A.D. received research funding from Vitaflo International, financial support from Nutricia, Mevalia and Vitaflo International to attend study days and conferences. SE received research funding from Nutricia, financial support from Nutricia and Vitaflo International to attend study days and conferences. A.P.received an educational grant from Cambrooke Therapeutics and grants from Vitaflo International, Nutricia, Merck Serono, Biomarin and Mevalia to attend scientific meetings. C.A. received honoraria from Nutricia and Vitaflo International to attend study days and conferences. A.M. received research funding and honoraria from Nutricia, Vitaflo International and Merck Serono, and is a Member of European Nutrition Expert Panel (Merck Serono international), a member of Sapropterin Advisory Board (MerckSerono international), and a member of the Advisory Board Element (Danone–Nutricia).

#### **References**


## *Article* **Accidental Consumption of Aspartame in Phenylketonuria: Patient Experiences**

**Ella Newbould 1, Alex Pinto 2, Sharon Evans 2, Suzanne Ford 3, Mike O'Driscoll 4, Catherine Ashmore 2, Anne Daly <sup>2</sup> and Anita MacDonald 2,\***


**Abstract:** Aspartame is a phenylalanine containing sweetener, added to foods and drinks, which is avoided in phenylketonuria (PKU). However, the amount of phenylalanine provided by aspartame is unidentifiable from food and drinks labels. We performed a cross-sectional online survey aiming to examine the accidental aspartame consumption in PKU. 206 questionnaires (58% female) were completed. 55% of respondents (*n* = 114) were adults with PKU or their parent/carers and 45% (*n* = 92) were parents/carers of children with PKU. 74% (*n* = 152/206) had consumed food/drinks containing aspartame. Repeated accidental aspartame consumption was common and more frequent in children (*p* < 0.0001). The aspartame containing food/drinks accidentally consumed were fizzy drinks (68%, *n* = 103/152), fruit squash (40%, *n* = 61/152), chewing gum (30%, *n* = 46/152), flavoured water (25%, *n* = 38/152), ready to drink fruit squash cartons (23%, *n* = 35/152) and sports drinks (21%, *n* = 32/152). The main reasons described for accidental consumption, were manufacturers' changing recipes (81%, *n* = 123/152), inability to check the ingredients in pubs/restaurants/vending machines (59%, *n* = 89/152) or forgetting to check the label (32%, *n* = 49/152). 23% (*n*= 48/206) had been prescribed medicines containing aspartame and 75% (*n* = 36/48) said that medicines were not checked by medics when prescribed. 85% (*n* = 164/192) considered the sugar tax made accidental aspartame consumption more likely. Some of the difficulties for patients were aspartame identification in drinks consumed in restaurants, pubs, vending machines (77%, *n* = 158/206); similarities in appearance of aspartame and non-aspartame products (62%, *n* = 127/206); time consuming shopping/checking labels (56%, *n* = 115/206); and unclear labelling (55%, *n* = 114/206). These issues caused anxiety for the person with PKU (52%, *n* = 106/206), anxiety for parent/caregivers (46%, *n* = 95/206), guilt for parent/carers (42%, *n* = 87/206) and social isolation (42%, *n* = 87/206). It is important to understand the impact of aspartame and legislation such as the sugar tax on people with PKU. Policy makers and industry should ensure that the quality of life of people with rare conditions such as PKU is not compromised through their action.

**Keywords:** phenylketonuria; phenylalanine; aspartame; sugar tax

#### **1. Introduction**

Aspartame, a non-nutritive sweetener, is one of the most widely used artificial sweeteners and accounts for 62% of the artificial sweetener market [1]. It is a synthetic dipeptide known as N-L-alpha-aspartyl-L-phenylalanine methyl ester (C14H18N2O5) and was accidentally discovered in 1965 [2,3]. Aspartame is completely hydrolysed to phenylalanine (50%), aspartic acid (40%) and methanol (10%) in the intestinal lumen and is rapidly

**Citation:** Newbould, E.; Pinto, A.; Evans, S.; Ford, S.; O'Driscoll, M.; Ashmore, C.; Daly, A.; MacDonald, A. Accidental Consumption of Aspartame in Phenylketonuria: Patient Experiences. *Nutrients* **2021**, *13*, 707. https://doi.org/10.3390/ nu13020707

Academic Editor: Shanon L. Casperson

Received: 18 January 2021 Accepted: 19 February 2021 Published: 23 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

metabolised by esterases and peptidases [4,5]. It is around 200 times sweeter than sucrose and it is estimated that it is added to >6000 foods and drinks [6,7]. Aspartame is approved in more than 90 countries and its safety has been evaluated by the Joint FAO/WHO Expert Committee on Food Additives (JECFA), as well as by numerous national food safety authorities, including the US Food and Drug Administration (FDA) and the European Food Safety Authority (EFSA) [8–10]. Aspartame can be safely consumed by healthy individuals, but it has long been recognised as a hazard to individuals with phenylketonuria (PKU) and therefore, it should be avoided [11]. The amount of phenylalanine in aspartame containing foods and drinks is not declared on ingredient labels and its impact on metabolic control in patients with PKU is not well established [12–14].

PKU, an autosomal recessive inherited condition, is caused by mutations in the gene encoding phenylalanine hydroxylase. It is estimated to affect 0.45 million individuals worldwide, with a global prevalence of 1:23,930 live births [15]. A rigorous lifelong low-phenylalanine diet is the principal treatment option. It requires the avoidance of high protein foods such as meat, fish, eggs, lentils, nuts, soya, bread, pasta and cheese. Daily dietary phenylalanine intake is calculated, measured and continually controlled according to individual tolerance. Eighty per cent of patients tolerate <500 mg/day (10 g natural protein/day) in order to avoid elevated blood phenylalanine levels. Phenylalanine tolerance does vary between patients depending upon the severity of their disorder and the use of pharmaceutical treatment options such as sapropterin (synthetic tetrahydrobiopterin (BH4), or pegvaliase (phenylalanine ammonium lyase). Sapropterin, an oral drug, is effective in a subset of BH4 responsive patients with PKU and is usually given as an adjunct to dietary treatment [16]. Pegvaliase, delivered by subcutaneous injection, is only licensed for adults with blood phenylalanine levels above the European PKU guidelines target range [17,18]. Neither pharmaceutical treatment option is available via the National Health Service in England.

The additional scrutiny of checking all food ingredient labels for aspartame in food, drinks and drugs intensifies the complexity of management [19]. Aspartame is added to a wide variety of foods: low calorie sweeteners, soft drinks (including fizzy drinks, fruit squashes/cordials), iced tea, flavoured mineral water, energy drinks, dessert mixes, frozen desserts, syrups/dessert sauces, mints, jelly, chewing gum, fruit yogurt, ice lollies, and ice creams. It is also added to around 600 pharmaceutical products (both medically prescribed and over the counter) including chewable multivitamins and cough medications. According to European law, foods containing aspartame must declare it is added either by name or E number (E951) [20]. However, it is not mandatory for manufacturers to state the amount of aspartame added to foods, rendering it impossible for people with PKU to estimate the phenylalanine intake from this source.

A further concern in the UK is the Soft Drinks Industry Levy (SDIL) which was introduced by Her Majesty's Revenue and Customs (HMRC) in 2018 [21]. It is commonly referred to as the "sugar tax". This was devised in response to national concerns about rising childhood and teenage obesity and was designed to encourage manufacturers to reduce the added sugar content of their drinks. It is a two-tier levy system: including a standard tax rate applied to drinks with a sugar content between 5 g and <8 g per 100 mL and a higher tax rate applied to drinks with a sugar content ≥8 g per 100 mL. This "sugar tax" has been highly effective with at least 50% of manufacturers reducing the sugar content of their products [22] but it has also led to many manufacturers replacing sugar with artificial sweeteners such as aspartame, potentially marginalising the dietary choices of patients with PKU. A recent equality risk assessment conducted by the HMRC examining the SDIL, stated that they were unaware of any evidence to suggest that the existing warning on food labels about the presence of aspartame in soft drinks was inadequate for people with PKU [23].

It is important to understand the impact of added aspartame to foods, drinks and medications on people with PKU. This paper aims to examine the frequency of accidental aspartame consumption, the reasons for this, and the challenges associated with avoiding aspartame in PKU.

#### **2. Materials and Methods**

#### *2.1. Study Design*

We performed a cross sectional online survey. Patients with PKU and/or parents/ caregivers of a person with PKU were invited to take part in this study. Respondents were excluded if they did not reside in the UK. The questionnaire was built in the Online Surveys platform (https://www.onlinesurveys.ac.uk (accessed on 1 April 2020)) and placed on the UK National Society for Phenylketonuria (NSPKU) website, with additional promotion on the NSPKU Twitter and Facebook accounts between April and July 2020.

This non-validated questionnaire contained 23 questions; 10 multiple choice (6 of which invited additional comments), 8 multiple response, 3 Likert scale and 2 open-ended questions. A group of experienced research dietitians from Birmingham Women's and Children's Hospital (A.P., S.E., A.M.), a colleague at the NSPKU (S.F.) and an expert in survey methodology (M.O.) helped develop the survey with a student dietitian from Birmingham City University (E.N.). The questionnaire was also reviewed by lay people to ensure its readability.

#### *2.2. Data Collected*

Demographic information was collected about the type of respondent (patient or parent/caregiver of patients aged ≥18 y or <18 y), gender of the person with PKU and confirmation of residency in the UK. Respondents answered questions about any known consumption of foods, drinks and medications containing aspartame, the frequency this had occurred, the reason behind this accidental ingestion and any symptoms this had caused. They were also asked about their knowledge and impact of the sugar tax with respect to the aspartame content of foods and drinks, and the ease of identifying aspartame on food, drinks and medication labels in addition to other challenges of identifying aspartame in products.

Overall themes explored in the survey were: accidental consumption of aspartame in food and drinks, accidental consumption of aspartame in medications, the sugar tax, drinks choice in different venues, label checking, and the effect of aspartame addition on the person with PKU and their family.

#### *2.3. Statistics*

Quantitative data analysis (inferential and descriptive statistics) was carried out with Statistical Package for the Social Sciences (SPSS) version 25 (SPSS Inc., Chicago, IL, USA). Multiple response questions were analysed with descriptive statistics only. Statistical significance was set at *p* < 0.05.

Qualitative data analyses of open-ended responses were carried out in NVIVO v.12 PRO. The whole survey dataset was imported into NVIVO so that the coding of open-ended responses could be broken down by survey questions including demographic questions. All open-ended responses were analysed thematically.

#### *2.4. Ethics*

Ethical approval to perform this study (approval number 6085, project title "The accidental consumption of aspartame in PKU: The experiences of patients and their caregivers") was given by Birmingham City University ethics committee. Adults with PKU and parents/carers of children and adults with PKU gave their consent at the beginning of the online questionnaire. Potential respondents were also advised that data from the survey may be published in an anonymized form. If names or hospitals were mentioned in verbatim abstracts, these were removed from results presented in this manuscript.

#### **3. Results**

There were 206 wholly or partially completed questionnaires. Fifty-five per cent (*n* = 114) of respondents were adults (18 or over) with PKU or parent/carers of adults with PKU and 45% (*n* = 92) were the parent or carers of children with PKU.

All respondents were normally residents in the UK. The PKU population described by the respondents were: 58% (*n* = 119) female; 41% (*n* = 85) male, 1 respondent was 'non-binary' and 1 preferred not to say.

#### *3.1. Accidental Consumption of Aspartame in Food and Drink*

Seventy-four per cent of participants (*n* = 152/206) said that people with PKU had consumed aspartame in a food or drink; 20% (*n* = 42/206) said they had not and 6% (*n* = 12/206) said they did not know.

Of those who had consumed aspartame by accident/error, just under half (47%, *n* = 72/152) said this occurred one to three times; 17% (*n* = 26/152) said 4 to 6 times and 6% (*n* = 9/152) said that it had occurred 7 to 9 times in the last 3 years. One in ten respondents (11%, *n* = 16/152) said that accidental consumption had occurred 10 times or more. Just under one fifth (19%, *n* = 29/152) of respondents could not recount how often accidental consumption had happened. Repeated accidental consumption of aspartame was more frequent in adults with PKU than for children (*p* < 0.0001). In the last 3 years, aspartame had been consumed accidentally 1 to 3 times in 79% (*n* = 42/53) of children and 43% (*n* = 30/70) of adults. In contrast, accidental consumption of 4 to 6 times occurred in 31% (*n* = 22/70) of adults compared to only 8% (*n* = 4/53) in children. Females (79%) with PKU were more likely to report having consumed aspartame than males (67%) (*p* = 0.008, Fisher's exact test). Eleven per cent (*n* = 8/74) of females had 7 to 9 incidents, compared to 0% (0/48) of males; 18% (*n* = 13/74) of females had 10 or more incidents, 3 times the proportion of males at 6% (*n* = 3/48). Patients that answered "don't know" were excluded.

The main reasons for accidental consumption of aspartame were manufacturers' changing product recipes (81%, *n* = 123/152), inability to check the ingredients e.g., drinks purchased in a pub or restaurant or from a vending machine (59%, *n* = 89/152), forgetting to check the label (32%, *n* = 49/152), and picking the wrong product from a shelf when shopping (29%, *n* = 44/152). Other reasons described by the respondents included: served the wrong drink in a bar or restaurant, (*n* = 22), unclear labelling (*n* = 16), not realising a product contained aspartame (*n* = 11), child unsupervised (*n* = 6), or other undefined reason (*n* = 4).

Examples of the verbatim quotes for the 5 most common themes for accidental aspartame consumption.


#### *3.2. Foods/Drinks Involved in Accidental Aspartame Consumption*

The food or drinks containing aspartame most reported to be accidentally consumed were fizzy drinks e.g., Coca Cola/lemonade/Irn Bru (68%, *n* = 103/152), fruit squash/cordials e.g., Robinsons Summerfruit squash (40%, *n* = 61/152), chewing gum (30%, *n* = 46/152), flavoured water (25%, *n* = 38/152), ready to drink cartons or bottles of juice/squash e.g., Strawberry Ribena (23%, *n* = 35/152), sports drinks e.g., Lucozade/Powerade (21%, *n* = 32/152), alcoholic drinks (19%, *n* = 29/152), sweets (14%, *n* = 21/152), jelly (9%, *n* = 14/152), tonic water (7%, *n* = 11/152), mints (7%, *n* = 11/152), iced slush drinks (7%, *n* = 10/152), energy drinks e.g., Red Bull (5%, *n* = 7/152), and table top sweetener e.g., Half-Spoon (3%, *n* = 4/152).

#### *3.3. Aspartame Consumption of Medically Prescribed and over the Counter Medications*

Twenty-three per cent (*n* = 48/206) of responders said that people with PKU had been prescribed medicines by their doctors that contained aspartame. This was more likely to occur in children (30%, *n* = 28/92) than adults (18%, *n* = 20/114).

Seventy-five per cent (*n* = 36/48) said that medicines were not checked by doctors/pharmacists for aspartame, but it was identified by the person with PKU or their carer. Twenty-five per cent (*n* = 12/48) said they had been advised that it was better to take the medicine and not worry about the aspartame content. Four per cent (*n* = 2/48) of respondents said the amount of phenylalanine from aspartame was checked and the number of phenylalanine exchanges adjusted accordingly. Thirteen per cent (*n* = 6/48) gave an "other" response including: 'was given a replacement medication only after they requested for this to happen', 'they accepted the medicine even though they knew it contained aspartame', were 'refused an alternative medication', and 'health professionals (dispensing the medication) were unaware of aspartame or PKU'. Although most respondents managed to access an alternative suitable medication, it depended on the patient or carer first identifying that aspartame was on the list of ingredients on the original medication.

Most respondents (88%, *n* = 182/206) were aware that some over-the-counter medicines contained aspartame, but 20% (*n* = 37/182) had consumed aspartame from this source.

Some verbatim extracts about the experiences associated with aspartame in medications are given below.


#### *3.4. "Sugar Tax"*

Most respondents (93%, *n* =192/206) were aware of the sugar tax. Many respondents (85%, *n* = 164/192) considered the sugar tax made accidental aspartame consumption more likely (either much more likely, 59% (*n* = 114/192) or slightly more likely, 26% (*n* = 50/192)). Eleven per cent (*n* = 21/192) thought that the sugar tax made no difference to the likelihood of accidental consumption of aspartame and just over 3% (*n* = 6/192) thought that the sugar tax had make it less likely.

Eighty-nine per cent (*n* = 170/192) thought the sugar tax led to fewer choices of drinks and more than two-thirds (68%, *n* = 130/192) considered that drink costs increased. More than four in 10 respondents said that the sugar tax had caused increased stress for the person with PKU and 27%, (*n* = 52/192) reported greater social isolation. Fifteen per cent (*n* = 29/192) of respondents thought that the tax had led to worse blood phenylalanine control for people with PKU. Only 5% (*n* = 10/192) thought the tax had no effect. 'Other responses' were commonly expressions of anger, being disheartened or depressed about the situation as the sugar tax increased the burden of dietary treatment even more.

Some examples of verbatim quotes given to the open question responses about the impact of the sugar tax:


#### *3.5. Choice of Drinks in Different Venues*

Respondents stated their dissatisfaction with the supply of drinks in different venues (Table 1). This was highest in relation to leisure/sports centres (67%); followed by fast food chains (62%) and restaurants (60%). Forty-nine per cent (*n* = 81/167) were dissatisfied (fairly or extremely) with the choice of drinks in hospitals, when people with PKU attended their clinics. Museums, airports, petrol stations and other people's homes had some of the lowest dissatisfaction scores but even for these venues, dissatisfaction is high in absolute terms (i.e., there is low satisfaction across all venues and high proportions are neutral on most venues).


**Table 1.** Satisfaction with the range of drinks across various venues.

Abbreviations: n: number of respondents. This varies considerably and is low for some venues such as 'university' and 'nurseries' because these are used predominantly by particular demographic groups and those that did not use them chose 'not applicable' and did not rate.

#### *3.6. Label Checking*

Respondents checked labels for food, drinks and medicines most of the time (Table 2). Drink labels (96%, *n* = 196/205) were checked either most of the time or always which is higher when compared with food labels (81%, *n* = 165/203).

**Table 2.** Proportion of respondents who check food, drinks and medicine labels.


Food, drinks and medication labels were always checked more often by parents/ caregivers for children. For food, 44% (*n* = 49/111) of adults or carers of adults always checked labels compared with 71% (*n* = 65/92) of parents/carers of children; for drinks, 56% (*n* = 63/113) of adults or carers of adults always checked labels compared with 78% (*n* = 72/92) of parents/carers of children; for medicine, 46% (*n* = 52/112) of adults or carers of adults always checked labels compared with 85% (*n* = 77/91) of parents/carers of children. On average, both adults or carers of adults and carers of children with PKU all checked food, drinks and medicines labels for aspartame 'most of the time' but carers of children significantly more so (Table 3).

**Table 3.** Mean levels of checking food, drink and medicine labels by age group.


Abbreviations: PKU, Phenylketonuria; n: number of respondents. The mean values relate to a scale of 1 to 5 (1 = Not at all; 2 = Rarely; 3 = Sometimes; 4 = Most of the time; 5 = Always).

#### *3.7. Ease of Identifying Aspartame on the Ingredient Label*

A high proportion of respondents reported it was very easy or fairly easy to identify aspartame on ingredient labels, 63% (*n* = 130/205) for food and 65% (*n* = 133/205) for drinks compared to those who had difficulty, 22% (*n* = 45/205) for food and 23% (*n* = 48/205) for drinks (Table 4). Ease of identification of aspartame on medicines was lower with 46% (*n* = 86/189) reporting it was very easy/fairly easy and 40% (*n* = 76/189) finding it difficult. The number remaining neutral was similar for food, drinks and medication.


#### *3.8. Challenges in Identifying Products which Contain Aspartame*

The biggest challenges identified by respondents are presented in Table 5 in detail.



Perhaps unsurprisingly, the highest single response category in open-ended responses about the challenges in identifying aspartame is related to product labelling. This was mentioned by nearly half of those who responded to this question.

Verbatim quotes about the challenges relating to identifying aspartame from labels on foods and drinks:


Respondents also mentioned that there was no prominent warning about the presence of aspartame or that this information was not consistently in the same place on packaging.


• "I can easily identify aspartame in products with labels, but it is time consuming and annoying. I worry that other caregivers, e.g., grandparents, would not be able to. There is no labelling in restaurants, so we err on the side of caution and only order what we know does not contain aspartame."

Some respondents suggested that the warning on packaging should be at least as prominent as allergen warnings.


Overall, 74% (*n* = 152/206) of respondents thought that it would be helpful (fairly or extremely) if manufacturers listed the phenylalanine content of food, drink or medicines on the label. Only 6% (*n* = 13/206) were neutral and 20% (*n* = 41/206) thought it would be fairly or extremely unhelpful.

#### *3.9. Effect of Aspartame on People with PKU and Parents/Carers*

Table 6 gives the percentage of patients that reported each of the stated effects of aspartame on patients and parents/caregivers managing PKU.


**Table 6.** Reported effects of aspartame on the person with PKU and parent/carer.

Abbreviations: PKU: Phenylketonuria.

Coding of the open-ended responses about the effect of aspartame showed that the top four themes were: feelings of being different, lack of choice, stress or concern and the additional time required to check all labels. These issues are illustrated in the following verbatim quotes.


• "Checking for aspartame increases stress and anxiety especially when eating out which is supposed to be a nice/happy experience."

#### **4. Discussion**

This is the first UK survey to examine the impact of aspartame in food, drinks and medications on people with PKU and their caregivers. We found that repeated accidental aspartame consumption is common, particularly in adults with PKU. Many respondents acknowledged there may be occasions in which aspartame has been inadvertently ingested and there were many concerns about the inability to identify its presence in pre-mixed alcoholic drinks and draft soft drinks in restaurants and bars.

The most unintentionally consumed aspartame containing items included fizzy drinks, fruit squash, cordials, flavoured water, sports drinks and chewing gums. Changes to product recipes, selecting the wrong product when shopping, packaging similarities between aspartame and non-aspartame containing products, unclear labelling, and difficulties identifying aspartame in drinks purchased from restaurants and pubs were commonly identified challenges. This suggests the need for: mandatory ingredient lists for all drinks and foods in restaurants, cafes, bars, and vending machines; distinct front of package labelling when a product recipe has changed; and clear labelling when there are several products within a brand range with some containing aspartame and others not (e.g., Ribena, Fanta, Tango, Robinsons). There should also be mandatory visible "first glance" disclosure of aspartame on packaging. Recently Dutch researchers demonstrated there was wide variability in the aspartame content of soft drinks, particularly the same brand of soft drinks bought in different countries. They have urged European legislators to enforce manufacturers to declare the amount of phenylalanine obtained from aspartame on food and drink labels, so that individuals with PKU are aware of the phenylalanine content of foods and drinks [24]. This 'call for action' is supported by NSPKU Medical Advisory Panel of dietitians [25].

Accidental aspartame consumption due to medications occurred in almost a quarter of respondents. Respondents felt there was little awareness or concern about the presence of aspartame in medications amongst medical professionals when they prescribed medication for PKU. Generally, reminders to check prescriptions for aspartame came from patients/parents' instruction rather than the GP or pharmacist. Aspartame is commonly used as a sugar replacement in antibiotics, chewable tablets and sugar-free liquids. The European PKU guidelines [11,17] recommend that for immediate and short-term treatment of infections, if only aspartame containing medicines are available, it may be better to use these until aspartame-free medication is sourced rather than leave a person with PKU without treatment (for a concurrent illness) as blood phenylalanine levels will rise with infection. However, for chronic long-term use of medications, it is better to find alternative aspartame free medications. Aspartame can be identified from the list of excipients in the medication instruction leaflet or the EMC summary of product characteristics. The amount of estimated phenylalanine in a drug may also be listed and can vary from 1 to 25 mg per dose of medication. There is usually no aspartame warning on the outside packaging of medication and there is no legal obligation to include this [26]. However, it is considered important to have mandatory legislation to identify aspartame on the outer packaging for people with PKU, otherwise it is challenging to recognize its presence at the point of prescription or purchase, and it can be a cause of frustration, inconvenience and distress for carers or people with PKU.

The impact of aspartame in food and drinks on inhibiting socialisation, increasing the incumbrance of dietary management and decreasing autonomy for children and teenagers is evident. Respondents were particularly dissatisfied with the choice of suitable drinks at many venues including fast-food restaurants, leisure centres, tourist venues and even hospital clinics. Respondents were angry that waiters/waitresses or sales vendors convey little understanding or empathy. They were displeased with the lack of aspartame free soft drinks at their hospital, as they considered this to be one location that above all others should demonstrate understanding of their condition. For NHS England hospital trusts, the Commissioning for Quality and Innovation (CQUIN) offer a financial incentive if they provide healthier food and drinks. This includes that 80% of drinks provided/sold must not be sugary. If a hospital trust adheres to the CQUIN for healthy food for NHS staff, visitors, and patients, they receive additional funding worth 0.1% of the trust's overall budget [27]. Unfortunately, there are no exceptions for vulnerable groups who are unable to tolerate aspartame for medical reasons.

There was much anger and despondency concerning the sugar tax by the respondents to this survey. Although the sugar tax has been implemented to reduce national overweight/obesity, it will not necessarily change unhealthy lifestyle practices. Overall people with PKU and their caregivers felt marginalised by this government policy. The sugar tax has led to diminished choice of favourite branded drinks and increased the cost of sugar containing drinks. For many adults, most available soft drinks in bars now contain aspartame, so the freedom of choice and the ability to enjoy a drink with friends has been withdrawn, which is hard to endure when there are so many other dietary restrictions to contend with. Many people with PKU have a functional approach to food; they eat for necessity rather than pleasure. However, drinking 'normal' branded drinks brought normality and choice. Almost 60% of respondents considered that the sugar tax led to more dietary errors and 33% felt fatigued or unwell with aspartame consumption, although no other information was collected about symptoms. Sugar is one of the few foods that is protein free and can be eaten without adversely affecting blood phenylalanine control in PKU. Giving adequate energy intake from very low protein sources is essential to meet energy requirements and to minimise catabolism that can lead to poor blood phenylalanine control [11], so sugar is not an 'unhealthy' food for people with PKU when eaten in moderation. Although it is unlikely there will be any reversal of the sugar tax, and it is expected to be extended to other foods, it is disappointing there is little consideration about the impact of the sugar tax on PKU by Public Health England or HMRC. Promoting healthy eating and exercise habits in the general population should be the key to solving obesity rather than focusing on one food component. Taking a balanced approach, offering many healthy choices without compromising the aspartame-free options for people with PKU would be a better policy.

Confusion and regular recipe changes with the addition of aspartame to manufactured foods/drinks affect a child's ability to self-manage their diet. For foods such as fresh meat, fruit and vegetables there is clear guidance on whether these are either permitted or forbidden in a low phenylalanine diet; but the ingredients, particularly in popular manufactured sweetened products, may change without notice, adding aspartame, with no clear warning to the consumer. It is, therefore, difficult to give pragmatic advice about suitable foods and drinks. Aspartame may be added to many children's foods such as ice lollies, soft drinks and iced 'slush' drinks that may be purchased from an ice cream van or local shop. Consequently, an adult with dietary knowledge should always check the suitability of these foods and the continual checking of food labels is time consuming and endless.

It is incomprehensible that alcoholic beverages with added sweeteners with an alcohol by volume content of 1.2% or more, do not have to declare the type of sweetener on the label. Moreover, legally no nutrition information needs to be supplied on the label of alcohol although appropriate allergen information and relevant quantitative ingredient information should be given [28]. This renders it unmanageable for people with PKU to be confident that any alcoholic drinks with unnamed sweeteners are safe for consumption. Fortunately, it is likely this situation will improve in the next 2 years. A memorandum of understanding (Self-regulatory proposal from the European alcoholic beverages sectors on the provision of nutrition information and ingredients listing) was presented as a joint voluntary commitment to the EU Health Commissioner in June 2019. It committed that by the end of 2022 the list of ingredients on alcohol will be provided according to the

EU 1169/2011 law. This law asserts that aspartame should be identified on the list of ingredients and it must state that it contains a source of phenylalanine [29].

There are several limitations to this study. The participants were not randomly selected and individuals without internet access may have been unable to participate. The survey was also promoted on the NSPKU Twitter and Facebook page, meaning participants were more likely to be NSPKU members who may be more proactive and informed about PKU. Therefore, the survey population may not be representative of the entire PKU population for which it is estimated that there are around 2000 UK patients in hospital follow up. Some surveys were completed by caregivers on behalf of patients with PKU and therefore responses to some questions may have been the caregiver's opinion rather than the actual experiences of those with PKU. It may be that aspartame was consumed more often but respondents did not realise this. Some respondents were unable to remember how many times they had consumed aspartame over the three-year period. The survey was not validated and therefore has not been checked for reliability, however expert opinion was used to develop it. This study should be repeated and expanded in the future using a validated survey that is piloted and carefully applied by health professionals to further improve the accuracy of the data collected.

#### **5. Conclusions**

It is important that health care professionals and policy makers understand the impact of aspartame and policies affecting the increased use of aspartame such as the sugar tax on the lives of people with PKU. Aspartame addition to food and drinks introduces social constraints, impacts on metabolic control as well as providing a source of frustration, guilt and distress to people with PKU and their carers. It is difficult to adhere to the PKU diet when all ingredients are not readily declared on labels at the point of purchase or issue. This applies to food, drinks, and medicines. It is essential that that industry gives clear and 'front of package' labelling about aspartame presence and the amount of phenylalanine that the product contains. Manufacturers should also consider using alternative sweeteners that would be a suitable option for people with PKU.

**Author Contributions:** Conceptualisation, A.M., E.N., A.P. and S.E.; methodology, A.M., E.N., A.P. and S.E.; formal analysis, M.O., E.N., A.M. and A.P.; writing—original draft preparation, A.M., E.N., A.P. and A.D.; writing—review and editing, A.M., E.N., A.P., S.E., S.F., M.O., C.A., A.D., and supervision, A.M., A.P. and S.E. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Birmingham City University ethics committee with approval number 6085 (project title "The accidental consumption of aspartame in PKU: The experiences of patients and their caregivers").

**Informed Consent Statement:** Informed consent was given by all subjects when filling in the questionnaire.

**Data Availability Statement:** The data will be made available from the authors upon reasonable request.

**Acknowledgments:** We would like to acknowledge and thank all the patients and families that have taken their time to fill in this survey.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

