A practical approach to low protein diets in Sweden– 45 years of clinical use

A Swedish group led by the late Professor Jonas Bergström at Karolinska Institutet in Stockholm [2], decided to explore the possibility of improving the quality of the low protein diet by supplementing it with essential L-amino acids. They showed that nitrogen balance was neutral or even positive with the use of amino acid (AA) solutions containing the 8 essential amino acids (EAA) and histidine compared to solutions containing only the 8 EAA [3, 4]. To avoid metabolic acidosis the basic AA was provided as acetate salts, since acetate is metabolized to bicarbonate. A solution containing the essential L-amino acids in proportions as originally proposed by Rose [5] was developed in co-operation with KabiVitrum AB Sweden and commercially distributed as Aminess®. It was initially given intravenously and was successful for short-term treatment of uremic patients. Coated EAA-tablets that did not dissolve in the mouth were developed thereafter and a diet regime with 18 g protein supplemented with 20-30 tablets/day was introduced in the clinics. Later it was found that such high doses of EAA were not really necessary and that 12-20 tablets/day was sufficient for most patients [2]. The LPD supplemented with this formulation was initiated as the standard treatment for patients with near-end-stage renal failure in Sweden. The diet was also successfully applied in other countries, especially Germany.

Professor Jonas Bergström and Professor Anders Alvestrand et al. continued their work on AA-metabolism in renal failure and reported findings of low intracellular concentrations of threonine, valine, lysine and tyrosine and increased concentrations of several non-EAA in untreated uremic patients [2, 6, 7]. Some of these abnormalities persisted in spite of long-term treatment with a LPD (16-20 g protein/day) supplemented with the EAA formulation. A new EAA formula with changed contents and proportions was designed to correct these abnormalities. The new preparation (Aminess-N®) contained proportionally more valine (+70 %] and less leucine and isoleucine than prescribed by the Rose formula [5], and in addition histidine and tyrosine. Studies with this new preparation showed that abnormalities of EAA in plasma and muscle characteristic of uremia can be corrected by nutritional means, and that uremic patients require AA in other proportions than healthy subjects [8]. This preparation is still in use.

The late Professor Härje Bucht from Karolinska Hospital moved to Gothenburg in 1970 and established the treatment with LPD at the Sahlgrenska university Hospital. During the following decade much research was devoted to explore the metabolic effects of the supplemented LPD. The work was mainly carried out by Professor Per-Ola Attman and included studies of nitrogen balance and body composition in patients with uremia [9]. During this period keto-analogues to the EAA were developed in Germany to further reduce the nitrogen intake. Nitrogen balance studies indicated that keto-acids could substitute for conventional EAA and were well tolerated by the patients [10]. This practice was not strongly established in Swedish clinical practice mainly because these supplements were not subsidized. The composition of the LPD in itself, with large amounts of saturated fat and simple carbohydrates, prompted an extensive study of the effects on the deranged lipid and carbohydrate metabolism that continued into the 1990’s. Despite the composition of the LPD, there was no further deterioration of the renal dyslipidemia [11].

Until the beginning of the 1980’s the patients were introduced to the LPD as in-hospital patients during a period of 1-2 weeks and assigned protein levels of 0,3 g (very low protein diet, VLPD) or 0,6 g (LPD) of protein per kg body weight/day, supplemented with EAA, iron and multivitamin tablets daily [2]. This practice gradually changed over time. Instead of starting LPD in-hospital, the patients were referred to a renal dietician by a nephrologist. By the end of the 80’s the use of LPD declined partly because access to dialysis treatment became increasingly available. There was also an opinion among some nephrologists that LPD might expose patients to the risk of developing malnutrition and that dialysis should be preferred. The concept of early dialysis or healthy start was launched and heavily backed up by industry. Other concerns were also raised, such as feasibility and cost-effectiveness. Management of patients with advanced renal failure only with diet, the so called conservative treatment, was viewed as too time-consuming by requiring close management with frequent visits in addition to patient education and physician’s and dietician’s efforts to maintain their motivation. An on-going debate persisted on whether LPD could affect the progression of renal function and/or delay the time before starting active treatment. This hypothesis was based on animal experiments and resulted in large interventional multicentre studies among which the MDRD study [12] is best known. Although the results of the study were interpreted differently depending on attitude and knowledge among nephrologists, it spurred the interest of nutritional management in renal disease.

Despite these issues, LPD continued to be in use, as its advocators saw the diet as a clinically safe treatment with many benefits if implemented correctly according to guidelines on energy intake and protein quality. However, the use of VLPD has declined in the last years, and has not been in frequent use in Sweden since the 1990’s. Instead, an individualized LPD is often implemented at levels of 0.6 – 0.7 g/kg/day according to current national guidelines [13, 14]. The diet is supplemented with EAA if the total protein intake is less than ≤0.6 g/kg or if the requirement of at least 60 % high quality protein is not met. LPD is generally introduced to the patient at an out-patient encounter by a specialized dietician upon referral from a nephrologist. The treatment is put into practise in the context of the multidisciplinary teamwork of dieticians, nephrologists and nurses aiming to alleviate uremic symptoms, maintaining or improving nutritional status, preventing malnutrition and delaying the need of dialysis.

In Sweden, dietician is a protected professional title with three years of higher education and licensed to practice by the National Board of Health and Welfare. Dieticians working at University hospitals become specialists in renal nutrition if they hold full-time positions in renal departments. Dieticians working in regional and local hospitals usually work generically. In addition, dieticians can attend special courses and meetings on updates in renal nutrition. The Swedish dieticians working in renal units have taken an active role in the practical implementation of LPD since the 1970’s. The first cookbook with recipes and cooking recommendations for patients and their relatives/caregivers in order to comply with a LPD was published in 1973 [15]. The authors were the dieticians Marianne Ahlberg and Marianne Wessman, one nephrologist and one patient, all at St Eriks Hospital in Stockholm. This cookbook made it easier for patients to comply with the treatment, for the staff to instruct the patients and for cooking staff in the hospital kitchens to prepare LPD meals all over Sweden. The goal for the daily protein intake was 15-20 g of protein and the recommendation for energy intake was high. The diet was supplemented with EAA as detailed earlier. Unsalted butter and cream was recommended in cooking to ensure that the energy requirements were met. No sodium was added. Vegetables were included in small amounts, which probably reflect the eating habits in the Swedish society at that time. Low protein products such as pasta and bread were, and still are available for the patients and subsidized by the government.

A second cookbook was published in 1977 [16]. This contained menu suggestions and inspiration to achieve different levels of protein intake of 20, 40, and 60 g/day. It still recommended no additional sodium, small portions of vegetables and increased intake of fat from unsalted butter and cream sources. This book was widely spread throughout the country and most clinics used the recipes for planning and cooking as well as to train the patients. During a few years, at Huddinge hospital in Stockholm, patients were given the option to purchase 1 to 3-weeks of ready-made frozen low-protein dishes to bring home and ease compliance.

A third cookbook was published in 2002 by two dieticians at Sahlgrenska University hospital in Gothenburg [17]. This book is still in use and available for purchase, and two easy recipes are here provided as an example (Table 1). The cookbook was published as an addition to the development/creation of a “self-selected” LPD model for the practical implementation of LPD developed by the renal dietician Gunilla Uddebom in the late 1980’s [18]. This LPD model was initiated due to an observed desire from the patients to take a more active role in the planning and implementation of the diet. The model was introduced on full scale at the Sahlgrenska University hospital in 1988 accompanied with a patient-oriented manual to be handed out to patients prescribed with LPD. The main purpose was to enable the adaptation of the diet to the patients’ own habits at home, in a family setting, at work, on travels or in leisure time. The ”self-selected” LPD model gives considerable flexibility regarding variation in meal planning, food selection, amounts consumed, cooking methods as well as possibilities for day-to-day changes of the whole diet plan with options for most foods. The goal is to educate the patient on how to plan their diets in order to achieve an intake of 0,6 g protein /kg/day and an energy content of 35-40 kcal/kg/day. The manual includes an introduction to the diet followed by information on food items divided into ten different groups (Table 2). The food groups are identified according to Swedish food traditions and use. Cheese is found in the group of spreads and toppings as this food item is mainly eaten as a topping on bread in Sweden. Hence, there needs to be a guide on how cheese can be exchanged with other toppings and spreads such as thin slices of ham or sausage. Butter on the other hand is grouped together with fats and oils in a food group which is as good as free from protein and therefore free to eat. Included in the group “dairy products” is e.g. milk (which is a common to drink together with cooked meals), yoghurt, cream and ice cream but also oil or oat- based “dairy-like” substitutes for milk and sorbet as a substitute for ice cream.

Each group contains a fixed amount of protein per serving with the exception of two groups, which are almost free from protein but provide considerable amounts of energy, principally from unsaturated fats and to some extent from saccharides. The amount of protein per serving is equal within each food group irrespective of the food selection (see example in Table 3). The patient decides upon the number of servings from the food groups within the limits of the described daily protein intake. They are advised to distribute the protein intake over the day and to choose food items from all 10 food-groups in order to get a well-balanced diet. The manual is available for purchase through the website of the Swedish Association of Clinical Dieticians’ Reference group in Renal Nutrition. The self-selected model is schematically illustrated by photos (Fig. 1) of a normal portion size of a traditional Swedish meal and its adaptation to a LPD portion size of the same meal using the manual with suggested amounts of foods from the different food-groups.

The model has been in use for nearly 30 years in more than five hundred patients at the Sahlgrenska University hospital. It has been evaluated in a retrospective case-control study including 122 patients [19]. The evaluation showed that treatment with LPD according to the self-selected model was not inferior (with respect to the risk of morbidity during the first year in dialysis and mortality), and in several instances superior to conventional pre-dialysis care without LPD (better nutritional status achieved and less morbidity at the time of starting dialysis). A systematic review of body composition in LPD was undertaken [20] to strengthen the clinics view of LPD as a clinically safe and effective treatment for patients with CKD.

Since 2004 a handbook has been available in Sweden with the title “Living with renal failure” developed by the patient Per-Åke Zillén and supported/reviewed by various healthcare professionals. The aim of this handbook is to assist CKD patients in their everyday life. One full chapter is devoted to the protein content of different foods in portions, pieces, cups and glasses in order to facilitate a daily overview of the protein intake. New editions have been released in 2007, 2011 and in 2014 [21].

Since 2008, incident patients with CKD 4-5 followed by a nephrologist are registered into the Swedish Renal Registry Chronic Kidney Disease (SRR-CKD). Currently, non-dialysis data collection of the SRR includes more than 95 % of all out-patient nephrology care and representativeness is estimated to range from 75-90 % during 2015 (www.​snronline.​se). During data collection, it is mandatory to report on whether the patient is treated with a LPD. The definition of LPD stated as per protocol of data collection is a prescription of 0.6 g /kg body weight/ day. Other rates of protein restriction may exist, but are not currently considered in the registry.

We accessed the SRR to evaluate the extent of LPD use among CKD patients. This was performed using all patients registered with an eGFR <45 ml/min/1.73 m² during the time span ranging from June 2013 to June 2015. In total, 10 % of these patients were treated with a LPD ≤0.6 g/kg. About half of the Swedish clinics (51 %) reported to use LPD as defined by the Registry whereas the rest of the centres used it in very few patients or not at all. According to the registry, 18 % of the clinics use it in more than 20 % of the patients. The choice of LPD treatment is more available in larger hospitals, including regional hospitals (12.7 % of LPD use) and university hospital (10.2 %), as opposed to local hospital (7.0 %). The use of LPD also differed within the Swedish counties and regions, being more common in patients with CKD stage 4-5 among the counties of Västerbotten (38 %), Jämtland (45 %) and Jönköping (35 %) (Fig. 2). LPD was more commonly used in the north and southeast parts of Sweden, but the use seemed more to reflect local traditions as it could vary also within the same county or region.

As a next step we attempted to evaluate patient characteristics in centres using LPD (defined for this purpose as centres with >5 % of patients following LPD in CKD stage 4-5). This involved 25 centres and 5997 patients. We observed that LPD was more commonly followed in individuals with CKD stage 5 (39 % use), followed by CKD stage 4 (14 %) and CKD stage 3 (8 %). There were no major differences in the use of LPD overall based on sex (men 18.5 %, women 17.1 %), and LPD use was relatively similar through different age strata. If we focused on the patient group where a LPD is recommended (eGFR <20 ml/min/1.73 m², n = 2391) we observed that patients diagnosed with adult polycystic kidney disease and glomerulonephritis (41 and 38 % respectively) were prescribed LPD more often than most other patient categories, while those with non-determined renal disease were prescribed LPD less often (22 %). Also patients with a body mass index (BMI) ≤20 kg/m² were prescribed LPD less often compared with patients of normal or high BMI. Presence of comorbid diseases such as diabetes, ischemic heart disease or cerebrovascular disease was associated with a lower use of LPD in general. Adjusting for age, sex, primary renal disease, body mass index and co-morbid diseases, the only factor that was independently associated with a higher use of LPD was normal BMI (20-25 kg/m²) Risk Ratio (RR 1.2; 95 % confidence interval [CI] 1.0-1.4) while both unknown renal disease and ischemic heart disease was significantly associated with a lower use for LPD (RR 0.65; 95 % CI 0.5-0.9 and RR 0.74; 95 % CI 0.6-0.9 respectively).

The Swedish Association of Clinical Dieticians’ Reference group in Renal Nutrition is comprised of seven board members, each working in different regions of Sweden. The group has initiated and developed patient material for the nutritional care of adult pre-dialysis patients and has co-operated with the Swedish Kidney Association in developing national guidelines for good nutritional care in renal disease. Below follows a consensus view of current practical implementation processes of the LPD in Swedish practice involving:

Two patient case reports are presented below.