Hyponatremia, hypokalemia and thiazide diuretics, part 2

Click here for part 1.


“Thiazide diuretic prescription and electrolyte abnormalities in primary care”, published in the January issue of the British Journal of Clinical Pharmacology, has several important lessons on the use of thiazides. Thiazide induced hypokalemia is generally known to be dose related, and the better safety profile with chlorthalidone in ALLHAT compared to MRFIT is believed to reflect the lower doses used in ALLHAT. However, low doses of thiazides were used in the current British study. No patient whose electrolyte disturbance was categorized as severe exceeded the usual starting dose of the thiazide, and 75% of patients with severe disturbances received only half the starting dose. So, while lower doses of thiazides are safer they do not eliminate the problem. Unfortunately the British study does not address mortality, nor does it tell us how many patients were hospitalized as a result of electrolyte disturbances.

Hyponatremia deserves special mention. Unlike hypokalemia, thiazide induced hyponatremia is less predictable. It appears to be an idiosyncratic reaction. This complication of thiazide treatment can have severe clinical consequences: lasting neurologic sequelae and deaths have been documented. This is the subject of a fascinating editorial from last August’s issue of Nephrology Dialysis Transplantation which reviews several unique clinical characteristics and risk factors for the syndrome. The principal mechanism of thiazide induced hyponatremia is probably not mineral depletion. There is evidence to suggest that thiazides promote retention of free water by up regulation of aquaporin channels in the renal collecting ducts.

While the British study informs us about electrolyte disturbances we desperately need data on the frequency with which such disturbances lead to clinical consequences. Such information is elusive in the case of hypokalemia given the difficulty in knowing the prevailing potassium concentration at the time of sudden cardiac death. The link between severe hyponatremia and neurologic sequelae is more easily established. Although multiple instances have been documented in the literature we don’t know the true frequency because we lack a denominator. Clinical events related to electrolyte disturbances, though they must have occurred, were not reported in ALLHAT.

These observations have implications for clinical practice and research.

1) As DB pointed out, electrolytes should be monitored regularly in patients taking thiazides.
2) Because thiazide induced hyponatremia is idiosyncratic and may occur early, the first electrolyte panel should be obtained within two weeks of the start of treatment.
3) Though normal electrolyte levels would seem a reasonable goal for most patients this recent paper from the Journal of the American College of Cardiology suggests that for patients with underlying heart disease normal is not enough. The authors recommend a potassium target of 4.5.
4) If the data are available the ALLHAT investigators should report the frequency of clinical sequelae and hospitalization related to thiazide induced electrolyte disturbances.