Correction of serum sodium for glucose concentration in hemodialysis patients with poor glucose control.

A s a rule of thumb in clinical medicine, the serum sodium concentration decreases by 1.6 mEq/l for every 100 mg/dl increase in glucose concentration due to water shifts from the intracellular to the extracellular compartment (1). This correction factor is based on theoretical considerations and has not been well validated. An experimental study in healthy subjects found significantly greater decreases in sodium concentration than expected when using the standard correction factor, especially when glucose concentrations were above 400 mg/dl (2). It was suggested that a correction factor of 2.4 mEq/l per 100 mg/dl increase in glucose concentration might be a better overall estimate of the association between the serum sodium and glucose concentrations (2). Exact knowledge of this correction factor for hemodialysis (HD) patients is increasingly important in view of recent efforts to align the sodium concentration in the dialysis fluid with that of the serum during HD treatment (3,4). A sodium concentration in the dialysis fluid higher than the serum sodium may lead to sodium loading and subsequent thirst, increased interdialytic weight gain, and hypertension. On the other hand, a sodium concentration in the dialysis fluid lower than the serum sodium may lead to a net sodium loss and has been associated with intradialytic symptoms such as hypotension and muscle cramps. Dialysis related morbidity can be significantly reduced when the sodium concentration of the dialysis fluid is aligned with the serum sodium (3). Although serum sodium measurements are stable over time, alignment of the dialysate sodium concentration in HD patients with highly variable glucose concentrations is challenging, and an adjustment may be needed (5). Therefore, we recently evaluated the relation between serum sodium and glucose concentrations in HD patients with poor glucose control. From a cohort of 4,532 prevalent patients receiving HD at one of our facilities between January and December 2009, we identified 208 patients with at least three available monthly serum sodium and glucose measurements and with a difference between the lowest and highest glucose value of more than 300 mg/dl. The mean ( SD) age of these patients was 55 13 years, 52% were male and 55% were black. Glucose values ranged from 81 to 549 mg/dl (5th and 95th percentiles, respectively). We calculated the slope for the relation between serum sodium and glucose for each individual. The mean slope was 1.47 0.82 mEq/l per 100 mg/dl increase in glucose level, which was significantly lower than the standard correction factor of 1.6 (P 0.02). In other words, each increase in the glucose level by 68 mg/dl lowered the serum sodium concentration with 1 mEq/l. The slope was independent of age, sex, and race. Although the interpretation of these data is limited due to their retrospective character, they do represent routine clinical practice. We conclude that the reduction in sodium concentration of 1.5 mEq/l per 100 mg/dl increase in glucose concentration found in our study is slightly lower than 1.6 mEq/l generally used in clinical practice. However, for clinical purposes our data validate the established correction factor of 1.6 mEq/l. E. LARS PENNE, MD STEPHAN THIJSSEN, MD JOCHEN G. RAIMANN, MD NATHAN W. LEVIN, MD PETER KOTANKO, MD