RESUMO
OBJECTIVE: Hypophosphatemia might cause respiratory and heart failure and even death. We aimed to evaluate risk factors for hypophosphatemia and refeeding-related hypophosphatemia in patients requiring parental nutrition (PN). METHODS: This was a single-center, retrospective study. Clinical parameters were obtained from medical records. Serum phosphate (inorganic phosphorus) was measured by photometric analysis. Hypophosphatemia was confirmed when serum phosphate level was less than 0.8 mmol/L (≈2.5 mg/dl). Refeeding related hypophosphatemia was confirmed if serum phosphate level had a decrease of 0.16 mmol/L or more from baseline and if the final assessment was below 0.65 mmol/L. RESULTS: A total number of 655 (426 men and 229 women, aged 62.8 ± 14.8 years) hospitalized patients requiring PN were included in the study, and 60.6% of them were patients with cancer. The average body mass index (BMI) was 21.1 ± 4.1 kg/m2 and the median of serum phosphate was 0.9 mmol/L (quartile range: 0.68 mmol/L, 1.11 mmol/L). The prevalence of hypophosphatemia was 37.6% (246/655). Older age (≥ 65 years vs. < 65 years), lower serum level of pre-albumin (< 160 mg/L vs. ≥ 160 mg/L), calcium (< 2.11 mmol/L vs. ≥ 2.11 mmol/L), and magnesium (< 0.75 mmol/L vs. ≥ 0.75 mmol/L) were associated with high risk of hypophosphatemia by multivariate logistic regression (OR ranged from 1.43 to 3.06, all p < 0.05). Refeeding related hypophosphatemia was 9.5% (16/168). Serum level of calcium at baseline was significantly lower in participants with refeeding related hypophosphatemia than those without it. Total calorie and nitrogen delivered during first week of PN period showed no obvious difference between patients with and without refeeding related hypophosphatemia. CONCLUSIONS: Hypophosphatemia is common (37.6%) in hospitalized patients requiring PN. Monitoring of serum level of phosphorus is necessary to facilitate early treatment of hypophosphatemia.