Diuretic-Related Side Effects: Development and Treatment

Domenic A. Sica, MD

In This Article

Metabolic Abnormalities

Prolonged thiazide diuretic therapy can lead to glucose intolerance and may occasionally precipitate diabetes mellitus.[4,5,54,55] Short-term metabolic studies, epidemiologic studies, and a variety of clinical trials suggest a connection between ongoing thiazide diuretic use and the development of type 2 diabetes. However, it should be noted that interpretation of these studies is confounded by multiple factors including: differing definitions of new-onset diabetes, small numbers of patients, inadequate comparison groups, relatively limited periods of follow-up, selection criteria that limited the generalizability of the findings, and study designs that prohibited valid comparisons among antihypertensive drug classes.[56] Moreover, in a review of all the placebo-controlled hypertension trials with diuretics, there was only an approximate 1% increase in new-onset diabetes compared with placebo.[57]

Hyperglycemia and carbohydrate intolerance have been linked to diuretic-induced hypokalemia. K+ deficiency is known to inhibit insulin secretion by ß cells; however, diuretic-induced changes in glucose metabolism are not conclusively related to altered K+ homeostasis, and impaired glucose tolerance occurs even when thiazide-type diuretics in relatively low doses are combined with K+ -sparing agents. The glucose intolerance seen with diuretic therapy can deteriorate further with an increase in sympathetic nervous system activity, which also decreases peripheral glucose utilization. Diuretic-associated glucose intolerance appears to be dose-related, less common with loop diuretics, not present with spironolactone, and reversible on withdrawal of the agent, although the data on reversibility in HCTZ-treated patients is somewhat conflicting.[55] Of note, an overview of this issue found that glucose homeostasis was unpredictably affected by low-dose HCTZ (12.5-50 mg/d).[58]

Recently, a large, prospective, cohort study (12,550 nondiabetic adults [45-to 64-years old] who did not have diabetes) concluded (after appropriate adjustment for confounders) that hypertensive patients taking thiazide diuretics were not at greater risk for subsequent diabetes development than patients who were not receiving antihypertensive therapy. The diuretic doses were not reported in this cohort study; thus, because of the perceived variability of this effect, blood glucose should be monitored during thiazide therapy, particularly in those with either the metabolic syndrome or existing diabetes.[59] This is particularly so since the CVR risk with new-onset diuretic-related diabetes parallels that which accompanies existing diabetes.[4] Other drug classes such as ACE inhibitors and ARBs are associated with a lesser incidence of new-onset diabetes. It remains to be determined the extent to which either of these drug classes reduces the diabetogenic potential of thiazide-type diuretics.