Absorption of Orally Ingested Phosphate in Refeeding Syndrome

The study was designed as a non-blinded, randomized, controlled crossover intervention study.

A total of 12 hospitalized adult head-neck cancer patients, 11 men and one woman, gave informed consent to participate. Six patients with a p-phosphate between 0.30-0.80 mmol/l were included as well as six patients (controls) with a p-phosphate in the normal range of 0.80-1.50 mmol/l. The control patients matched the hypophosphatemic patients regarding sex, age (± 10 years) and alcohol habits. Patients with diagnosed renal failure, edema, diabetes, liver failure, gastrointestinal disorders, pregnant and lactating women were excluded. None of the patients had phosphate supplementation within the previous 24 h. Randomization determined the order of the phosphate supplements divided into low-doses of five mmol followed by high-dose 20 mmol both for two day periods. Accordingly, four days intervention with a different phosphate supplement every day.

In the low-dose period supplements consisted of a daily dose of five mmol phosphate from skimmed milk (97 mg phosphorus and 0.1 g fat/100 ml) or Phosphate with black currant flavor (PBC) (disodiumphosphatedihydrate and potassiumdihydrogenphosphate, 1 mmol phosphate/ml). In the high-dose period supplements consisted of 20 mmol phosphate from Addiphos® (Fresenius-Kabi) (disodiumphosphate, potassiumdihydrogenphosphate, potassiumhydroxide, 2 mmol/ml) and PBC. The patients fasted six hours prior to baseline blood- and urine samples and during the four hours of the trial. Blood samples were repeated after two and four hours, urine after four hours. The interval between blood samples was determined on the basis of an animal experiment, which described phosphate absorption measured in plasma 1 hour after administration (10). The interval was extended to 2 hours due to the risk of gastric retention.

The primary outcome was the changes in p-phosphate. The secondary outcomes were changes in u-phosphate and changes in p-potassium, p-magnesium and p-sodium. The paired data were tested by the Wilcoxon test. Non paired data were tested by the Mann-Whitney test. Spearman-rank correlation test was used, and the statistical analyses were performed with STATA version 13.1 (StataCorp LP, USA, Texas).