A Randomized, Double-Blind, Crossover Pilot Trial of Rice Endosperm Protein Supplementation in Maintenance Hemodialysis Patients

In maintenance hemodialysis (MHD) patients, low protein intake is associated with protein-energy wasting, a risk factor that affects outcome. However, increased protein intake may lead to hyperphosphatemia and hyperkalemia, which are also mortality risk factors. Here, we evaluated the safety and effects of purified rice endosperm protein (REP), which contains less phosphorus and potassium than soy and casein proteins, as a supplemental protein source for MHD patients. This randomized, double-blind, placebo-controlled, crossover pilot study of REP supplementation (5 g/day × 4 weeks) was carried out in 50 Japanese adult MHD patients (1 dropped out); the primary outcome was the change in the urea kinetic-based normalized protein catabolic rate (nPCR), an indicator of protein intake in MHD patients. Intention-to-treat analyses of 24 patients in the REP-first group and 25 in the placebo-first group showed that REP supplementation increased nPCR significantly by 0.07 g/kg/day (95% confidence interval, 0.03–0.11), whereas changes in serum phosphorus and potassium concentrations were not different from the placebo. REP supplementation did not show a significant effect on other nutritional or metabolic parameters and no specific complications. In conclusion, purified REP with efficient bioavailability may be safe and useful for dietary supplementation in MHD patients.

(nPCR), which is an index of dietary protein intake in this population, was reported to be suboptimal across countries, with a median value ranging from 0.91 g/kg/day in Japan to 1.10 g/kg/day in Spain 3 . This suggests the need for protein supplementation for the majority of maintenance hemodialysis (MHD) patients.
However, increased protein intake may lead to hyperphosphatemia and hyperkalemia, which are risk factors for mortality 4,5 . In addition, high acidogenic food intake, high salt and fluid intake, and lower vegetable intake may be other risk factors associated with increased protein intake 6 . A previous study reported a U-shaped association between nPCR and mortality in MHD patients; where the best survival was observed between 1.0 and 1.4 g/ kg/day, and both lower (<0.8 g/kg/day) and higher (>1.4 g/kg/day) nPCR were associated with greater mortality 7 . Another study also suggested greater survival in MHD patients whose serum phosphorus decreased but whose nPCR increased, when compared with those whose serum phosphorus and nPCR rose over 6 months 8 . Therefore, it is critical to establish dietary protein sources at least containing less phosphorus and potassium to prevent protein-energy wasting by enhancing protein intake in MHD patients.
Rice, one of the most important cereals in the world, is a major plant source of both energy and protein 9,10 . Recently, a simple, mass preparative procedure for rice endosperm protein (REP) was developed using an alkali extraction method 11 . This isolate is characterized by >80% crude protein, with lower amounts of phosphorus and potassium compared with soy 12 and casein proteins 13 (Table 1). The digestibility of native REP in humans is 88% 14 , with the major indigestible component being prolamin. The alkali extraction procedure changes the digestibility of prolamin markedly 15 and improves the bioavailability of whole REP 11,16 .
In this clinical trial, we investigated the effects of REP supplementation on nPCR, mineral metabolism, and nutritional and metabolic parameters in MHD patients with insufficient protein intake and inadequate nutritional status.

Results
Participant Characteristics. The outline of this single-center, double-blind, placebo-controlled crossover pilot trial is shown in Fig. 1. We screened 415 MHD patients and enrolled 50 participants satisfying the inclusion criteria who were allocated randomly to receive REP first or placebo first (Fig. 2). The composition of the REP powder and the constituents of the intervention foods are shown in Tables 1 and 2, respectively. One participant dropped out of the REP-first group after completing the REP supplementation period because of hospitalization for a skin ulcer on the right arm, leaving 25 in the REP-first group and 24 in the placebo-first group for analysis (Fig. 2). Table 3 shows the clinical characteristics of the 49 participants at baseline. The mean age was 68 ± 10 years, 41% were female, mean serum albumin was 3.3 ± 0.3 g/dL, mean body mass index was 20.6 ± 1.3 kg/m 2 , and mean nPCR was 0.88 ± 0.16 g/kg/day. The participants had been on MHD for a median duration of 10 years (interquartile range, 4-17 years; range, 2-44 years). They were all anuric, so their residual kidney function was minimal or negligible. Phosphate binders, vitamin D receptor activators, cinacalcet, and potassium binders had been prescribed to 40, 27, 7, and 17 participants, respectively, and the medications were constant from 1 month  (Fig. 3). Table 4 shows the "within-group" changes in other parameters from baseline to the end of each intervention period and their "between-group" differences. Of note, phosphate metabolism-related parameters (i.e., intact parathyroid hormone [PTH], 1-84 PTH, and intact fibroblast growth factor-23 [FGF23]) did not change in the REP group. Single-pool Kt/V, an index of dialysis dose, was not significantly different between the groups. There were also no significant between-group differences in relation to changes in anthropometric data (e.g., dry weight, skeletal muscle, fat mass, lean body mass, mid-arm circumference, and skinfold thickness) or changes in markers of nutritional status (e.g., serum total protein, albumin, prealbumin, retinol-binding protein [RBP], and transferrin). There were no differences in the other parameters  of blood analysis between the groups. During the intervention and follow-up periods, there were no specific complications associated with REP or placebo supplementation.

Self-Administered Diet History Questionnaire Analysis. The composition of nutrients analyzed from
the self-administered diet history questionnaire (DHQ) at the beginning and end of this study is shown in Table 5. There were no significant differences in the intake of energy, carbohydrate, fat, protein, phosphorous, or potassium between before and after the study.

Discussion
In this randomized, double-blind, crossover pilot trial, oral REP supplementation (5 g/day) for 4 weeks significantly increased nPCR without significant changes in nutrient intake and serum concentrations of phosphorus and potassium in MHD patients with insufficient protein intake. There were no specific complications associated with REP or placebo supplementation. In addition, there were no significant changes in anthropometric data or markers of nutritional status by REP supplementation, probably due to the relatively short-term study period. Values of nPCR may be underestimated among MHD patients with residual kidney function 17 . However, most of our participants were long-term MHD patients with a median duration of 10 years (interquartile range, 4-17 years; range, 2-44 years) and they were all anuric. Therefore, they were likely to have little or no residual kidney function. Additionally, the primary outcome of this study was the change in nPCR, and hence, the influence of residual kidney function is considered to be minimal given the nature of this randomized cross-over trial design with a relatively short-term study period of 4 weeks.
Some randomized controlled studies of protein supplementation in MHD patients have been reported. Moretti et al. reported that liquid hydrolyzed collagen protein supplement given to patients treated with MHD or peritoneal dialysis (45 g for MHD and 105 g for peritoneal dialysis patients per week) for 6 months significantly    studies did not report changes in serum phosphorus or potassium concentration. Casein protein supplementation (6.6 g/day) in MHD patients for 4 months improved prealbumin levels, but increased serum phosphorus levels 20 . Meanwhile, a 6-week single-arm pilot study involving 13 MHD patients was conducted with 225 g pasteurized liquid egg white as a supplemental protein source 21 , which contains biologically high-value protein but relatively low phosphorus and cholesterol 22,23 . The study showed a decrease in serum phosphorus and an increase in serum albumin, requiring further controlled studies. In addition, it remains to be tested whether increasing nPCR by protein supplementation would decrease the mortality of MHD patients.  The source of dietary protein (animal or plant) is an important nutritional issue. Excess animal protein intake was found to be associated with an increased risk of diabetes in a European prospective cohort study with 38,094 participants 24 . A diet with a higher proportion of protein from plant sources was associated with improved metabolic acidosis and blood pressure reduction 25 as well as lower all-cause mortality in patients with CKD 26 . Plant-sourced protein is also known to have beneficial effects on phosphorus metabolism in CKD patients 27,28 , although there may be other unknown favorable factors involved. However, as shown in Table 1, purified soy protein was reported to contain relatively high levels of phosphorus (776 mg/100 g) and potassium (81 mg/100 g) 12 . In contrast, the potassium content of our REP preparation is negligible, which is another beneficial factor in addition to its low phosphorus content for its use as a supplement for MHD patients.
Protein supplementation may not be solely sufficient to improve protein-energy wasting in MHD patients. Indeed, protein supplementation with ricotta cheese alone did not increase muscle mass in sarcopenic elderly men 29 . Kim et al. suggested that exercise and amino acid supplementation together may be effective in enhancing not only muscle strength but also the combined variables of muscle mass and walking speed and of muscle mass and strength in sarcopenic women 30 . In MHD patients, Majchrzak et al. reported that resistance exercise augments the protein anabolic effects of oral nutritional supplementation 31 .
The limitations of this pilot study are the shortness of the trial period and the small number of participants. In addition, this was a single-center study. The clinical relevance of an increase in nPCR also remains to be investigated. The placebo intervention food was prepared with the same recipe and packaging as the test food, but without REP. However, the taste of both types of intervention food-REP-containing jelly and placebo jelly-may not have been completely the same, although it would have been difficult for the participants to distinguish one from the other.
In conclusion, as it has low phosphorus and potassium content as well as efficient bioavailability, REP may be a safe and useful source of protein supplementation for MHD patients. In the future, long-term studies with adequate sample size, as well as with combined exercise intervention, are needed to evaluate whether REP supplementation improves the clinical outcome of MHD patients.

Methods
Participants and Study Design. This was a single-center, double-blind, placebo-controlled crossover pilot trial, with change in the nPCR as the primary outcome. Serum minerals, nutritional and metabolic parameters, and adverse effects were also evaluated. The study participants were enrolled from July 29 to August 31, 2013, and were followed up for 12 weeks by M.H., H.S., and S.M. in Shinrakuen Hospital, Niigata, Japan. The inclusion criteria were: (1) age ≥ 20 years, (2) on maintenance HD ≥ 2 years, (3) serum albumin < 3.8 g/dL, (4) body mass index ≥ 19 and <23 kg/m 2 , (5) change in dry weight during the last 6 months < 5%, (6) nPCR < 1.2 g/kg/day, and (7) ability to provide informed consent and come for all visits. The exclusion criteria were: (1) severe heart disease (New York Heart Association III or IV), (2) severe hepatic insufficiency, (3) apparent signs of current systemic infection or sepsis requiring the active use of intravenous antibiotics, (4) perioperative status, (4) severe traumatic injury, and (5) any food allergy. The study participants were all anuric, that is, their urine volume was less than 100 mL/day, ascertained by questionnaires at baseline. A random allocation table was generated separately by computer at the central study facility, using a block size of 4 with stratification by gender. Y.O. allocated the participants in a 1:1 ratio to receive initially either placebo or 5 g REP-containing jelly supplement to eat once daily for 4 weeks. This amount was determined by the water intake restriction for MHD patients and a previous preliminary safety evaluation. The assigned allocation was concealed from the participants and care providers throughout the study period. After 4 weeks of a given intervention and after a wash-out period of 4 weeks, the participants received the other intervention. The participants continued their usual diet and physical activities during the study period. Pre-and post-HD blood samples were obtained at the beginning and end of each intervention period (Fig. 1). This study was approved by the institutional review boards of Niigata University and Shinrakuen Hospital in accordance with the principles embodied in the Declaration of Helsinki, and all participants gave written informed consent. The study was registered with the University Hospital Medical Information Network-Clinical Trials Registry (UMIN000010876; June 10, 2013).
Preparation of REP. REP was extracted from rice flour of regular Japonica rice Koshihikari using the alkaline extraction method 11 with modifications. Rice flour (100 kg) was mixed with a 0.2% NaOH solution (400 L, approximately pH 12.5) and extracted overnight at room temperature. The mixture was then centrifuged to collect the supernatant. The supernatant was heated at 50 °C and HCl was added to adjust the pH to 7.0 ± 0.2. REP Preparation of the Intervention Food. As MHD patients generally restrict their water intake, the intervention food was prepared with a limited quantity of water to produce a thick jelly. Two types of intervention food were prepared-REP-containing jelly and placebo jelly-by Kameda Seika Co., Ltd. (Niigata, Japan). The composition of each intervention food is shown in Table 2 week. All materials were mixed with REP and boiled. Then, the jelly mixture was packaged into film-covered opaque cups (volume, 50 mL) without any indication to ensure double-blindness. Placebo food was prepared with the same recipe but without REP. Finally, the jellies were sterilized using a retort machine. Intervention foods used for 2 weeks were prepared every other week and were provided to each participant based on the random allocation within 4 days of sterilization.
Laboratory Methods. Pre-HD venous blood samples were obtained at the beginning of the week. Routine biochemical parameters and RBP, transferrin, 25(OH)D, 1,25(OH) 2 D, 1-84 PTH, and intact FGF23 were analyzed. Serum phosphate and albumin were measured by the molybdate direct method and modified bromocresol purple method, respectively. RBP and transferrin levels were measured by latex agglutination turbidimetry and turbidimetric immunoassay (Nittobo Medical Co., Ltd., Tokyo, Japan), respectively. Double-antibody radioimmunoassays were used to measure the levels of 25(OH)D (DiaSorin, Inc., Stillwater, MI, USA) and 1,25(OH) 2 D (Immunodiagnostic Systems Ltd., Boldon, UK). Levels of 1-84 PTH were measured by a chemiluminescent enzyme immunoassay (Fujirebio, Inc., Tokyo, Japan) and those of intact FGF23 were measured using an enzyme-linked immunosorbent assay (Kainos Laboratories, Inc., Tokyo, Japan). nPCR and Kt/V were calculated using a formal single-pool model of urea kinetics 32 . Body composition was assessed by segmental multiple frequency bioelectric impedance measurements at baseline and at the end of each intervention period using tetra polar 8-point tactile electrodes (InBody S20; BioSpace, Seoul, South Korea).
Dietary habits were assessed by a validated DHQ at the beginning and end of the study. Details of the structure, calculation of dietary intake, and validity of the DHQ for commonly studied macronutrient intake are available elsewhere 33 . Statistical Analysis. This was a pilot study without sample size calculation. The intention-to-treat principle was employed for primary analyses, and a two-sided P-value of <0.05 was considered significant. Summary data of all variables are expressed as mean ± standard deviation and were compared using a paired t-test or signed-rank test according to their distribution. For 25(OH)D (P = 0.002), FGF23 (P = 0.001), and transferrin (P = 0.019), we found significant differences in the effects of REP supplementation between the intervention orders, suggesting incomplete waning of the treatment effect during the washout period. Therefore, these variables were analyzed using linear mixed-effects models with participant-level random intercepts to test the effect of REP supplementation, adjusting for the period in which REP supplementation was administered. Sensitivity analyses were conducted after excluding patients who dropped out of this study due to severe adverse events. All analyses were conducted using Stata/IC 12.1 for Windows (StataCorp LP, College Station, TX, USA). Data Availability. The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.