Selective TZD Use Won't Exacerbate HF Symptoms in Diabetics With Heart Failure, Says Analysis

November 13, 2006

November 13, 2006 (Chicago IL) – The use of thiazolidinedione (TZD) agents is common in selected diabetics with heart failure and, contrary to concerns in the absence of definitive clinical trials, neither increases nor decreases the risk of mortality or HF-related hospitalization, suggests a retrospective cohort study [1].

Diabetes is common in patients with HF and has a negative impact on prognosis, but there are as yet no completed, prospective randomized trials of TZDs for diabetes control in patients with established HF, observed the lead author of the current analysis, Dr David Aguilar (Baylor College of Medicine, Houston, TX). His group's findings in a multicenter Veteran Affairs population, he told heartwire , suggest that the TZDs in selected diabetics with HF "can still be used cautiously as we await further data." Any HF patient getting one of these drugs for diabetes, he said, should be carefully followed for signs of volume overload.

The TZDs, which include pioglitazone (Actos, Takeda Pharmaceutical Company) and rosiglitazone (Avandia, GlaxoSmithKline), aren't recommended for patients in NYHA functional class 3-4, out of concern that they could exacerbate HF signs and symptoms. Edema is a long-recognized side effect of these agents. Despite their prowess for diabetes control, pioglitazone and rosiglitazone--as covered previously by heartwire --were both recently associated with small but significantly increased risks of new HF in prospective, randomized trials of diabetics without established heart failure.

As Aguilar presented here at the American Heart Association 2006 Scientific Sessions, the cohort of more than 4500 patients with both diabetes and HF seen at two years' worth of outpatient visits included 818 who were taking thiazolidinediones and 4700 who weren't on any insulin-sensitizing drugs. Patients taking biguanides alone had been excluded.

The rate of HF hospitalizations was significantly increased (p<0.01) and absolute mortality was significantly reduced (p<0.05) over two years from the index outpatient visit among those on thiazolidinediones compared with those not taking the drugs. For both end points, however, neither group showed an increased risk in an analysis that controlled for a range of demographic, clinical, and comorbidity- and treatment-related features.

Two-year outcomes, TZD therapy vs no TZDs

End point

TZD, n=818

No TZDs, n=4700

HR* (95% CI)

HF hospitalization (%)



1.03 (0.84–1.27)

Mortality (%)



0.98 (0.81–1.17)

*Adjusted for age, sex, body-mass index, LVEF, HF severity, hypertension, renal function, glycosylated hemoglobin levels, prior diabetic complications, prior MI, chronic obstructive pulmonary disease, peripheral vascular disease, dementia, liver disease, and therapy with ACE inhibitors, beta blockers, statins, and other diabetic medications

There were suggestions in the analysis that the TZDs may have been given preferentially to HF patients with more severe diabetes or ischemic heart disease or to patients judged at lower risk of HF exacerbation. Those who received the drugs were significantly younger, had a significantly increased body-mass index, and were significantly more likely to have a history of diabetic complications, normal or only mildly reduced LVEF (as compared with more severe LV dysfunction), a history of MI, and more likely to be on statins. That might explain why the raw, unadjusted numbers suggested that the TZDs may have reduced mortality. "But when we adjusted for age, we lost a lot of that significance, and when we adjusted for the gamut of risk factors, the results approached unity," Aguilar said.

  1. Aguilar D, Deswal A. Thiazolidinediones are not associated with increased risk of heart failure hospitalizations in ambulatory patients with diabetes mellitus and heart failure. American Heart Association 2006 Scientific Sessions; November 12-15, 2006; Chicago, IL. Abstract 1894.

The complete contents of Heartwire , a professional news service of WebMD, can be found at, a Web site for cardiovascular healthcare professionals.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.