The introduction of thiazolidinediones (TZDs) has been greeted with enthusiasm in part because of their ability to lower glucose by addressing an underlying pathological feature of diabetes and because TZDs have the potential to ameliorate several cardiovascular markers. This enthusiasm has been tempered by safety concerns. We, therefore, read with interest the American Diabetes Association/American Heart Association consensus statement (1) regarding TZD use, fluid retention, and congestive heart failure (CHF) .
We disagree with portions of this statement. In individuals without symptomatic heart disease, we would not start TZDs if the patient had preexisting edema or a history of heart failure, especially since no lower range for a depressed ejection fraction is stated. While this may have a salutary effect on glucose control, it introduces unnecessary complexity and confusion when considering future symptoms and signs. Likewise, we would wait to use TZDs in patients with symptomatic heart failure until ongoing clinical trials establish the safety of these medications in this patient population. We disagree strongly with components of the monitoring section. When someone who is free of CHF develops CHF after the start of a medication known to worsen CHF, a recommendation to “reconsider” or a “dosage change” is unduly restrained, especially given the uncertain time period that is required for the effect of TZDs to dissipate. Additionally, if edema develops after the start of TZDs, the statement invites consideration of alternate causes of edema. While this recommendation is intellectually defensible, it belies the reality that physicians will, with rare exception, stop a medication when a side effect develops that is clearly known to be caused by the medication. In both circumstances, we would stop the medication.
Until ongoing prospective clinical trials ascertain the safety of TZDs in the patients that we cite above, we believe that there should be a strong preference for alternate therapies.
