Prevalence and predictors of rhabdomyolysis in patients with hypophosphatemia

Abstract

Purpose: We undertook this study to determine the prevalence and predictors of rhabdomyolysis in the hypophosphatemic state.

Patients and methods: To identify patients with hypophosphatemia, we reviewed medical admissions for the period of January through December 1989. The hypophosphatemic state was considered whenever the serum phosphate was less than or equal to 2.0 mg/dL. Rhabdomyolysis secondary to hypophosphatemia was defined when serum creatine kinase levels were greater than or equal to 224 IU/L; it occurred within 72 hours of the hypophosphatemic episode; and it subsequently normalized. Patients who had any other independent etiology for rhabdomyolysis were excluded. Clinical and biochemical characteristics of patients with rhabdomyolysis (Group I) and patients without rhabdomyolysis (Group II) were compared. Variables that predicted rhabdomyolysis in hypophosphatemia were identified by stepwise logistic regression using a backward elimination procedure.

Results: One hundred twenty-nine patients were found to have hypophosphatemia. Forty-six (Group I) of 129 patients (36%) showed biochemical evidence of rhabdomyolysis. There was no difference in serum phosphate and potassium concentrations between Group I and Group II patients. Patients in Group I showed higher values for serum osmolality (p less than 0.05), serum glutamic oxaloacetic transaminase (p less than 0.001), chloride (p less than 0.01), and blood urea nitrogen (less than 0.05) when compared with Group II patients. When biochemical profiles of patients with rhabdomyolysis were evaluated on the day of their peak creatine kinase level, only 16 patients were hypophosphatemic, and the majority of patients showed a transient increase in serum phosphate levels because of ongoing muscle cell injury. Of 17 potential predictors, six variables emerged including sodium, chloride, glucose, blood urea nitrogen, uric acid, and osmolality. These variables provided high sensitivity (0.88) as well as moderate specificity (0.79) for predicting the occurrence of rhabdomyolysis in hypophosphatemia.

Conclusion: We conclude that rhabdomyolysis commonly occurs in the hypophosphatemic state and that at times severe hypophosphatemia as an etiology may be masked because of ongoing rhabdomyolysis. Serum sodium, chloride, glucose, blood urea nitrogen, uric acid, and osmolality have a predictive role for the occurrence of rhabdomyolysis in the hypophosphatemic state that shows a high specificity and a moderate sensitivity.