Background: Hypokalemia is a common and sometimes serious clinical problem, whose etiological diagnosis can frequently be based on the patient's history and the clinical setting. Measurement of urinary indices such as excretory rate of K+, random urine K+ concentrations and blood acid-base parameters have been employed in the pathophysiological diagnosis, though with some pitfalls.
Methods: To investigate the diagnostic usefulness of the transtubular potassium concentration gradient (TTKG) and urine ammonium in the differentiation of hypokalemia, we measured serum K+ and osmolality, random urine electrolytes, osmolality and ammonium, the urinary [Na]/[K] ratio (U(Na)/K), plasma aldosterone and TTKG in 7 patients with diarrhea, 6 with vomiting, 7 with mineralocorticoid excess, 6 with diuretic usage, and compared them with those of 7 overnight fasted and acid-loaded healthy volunteers.
Results: The urine K+ concentrations did not reflect urinary loss of potassium according to the subjects' hydration status. U(Na)/k in the hypokalemic patients with mineralocorticoid excess (1.4 +/- 0.5) was lower than in normal subjects (2.3 +/- 0.4) (p<0.05). TTKG was higher in hypokalemic patients with mineralocorticoid excess (13.3 +/- 4.4) and diuretic usage (8.6 +/- 1.3) and lower in those with diarrhea (1.6 +/- 0.3) than in the normal controls (5.0 +/- 0.7) (p<0.5). TTKG in the patients with vomiting (3.5 +/- 0.6) was the same as in normal controls. TTKG was stronger correlated with the plasma aldosterone levels in the hypokalemic patients due to renal potassium loss. Urine ammonium concentrations of the acid-loaded normal subjects (73.3 +/- 5.0 mEq/L), patients with diarrhea (74.4 +/- 2.0 mEq/L) and patients with mineralocorticoid excess (68.7 +/- 6.9 mEq/L) were higher than in overnight-fasted normal subjects (31.3 +/- 4.9 mEq/L).
Conclusion: TTKG and random urine ammonium were useful in the pathophysiological differential diagnosis of hypokalemia.