ACCORD

In the years after the DCCT established tight glycemic control as the standard for type 1 diabetes, a widespread assumption in diabetology held that the same principle should apply to type 2 disease. The UKPDS reported in 1998 that intensive control in newly diagnosed patients slowed microvascular disease and might also reduce heart attacks, and guidelines progressively pushed HbA1c targets lower. By the mid-2000s, many specialists recommended targets below 7% or even below 6.5% in patients with established cardiovascular disease, the group presumed to have the most to gain.

ACCORD was designed to test that assumption directly. The trial assigned 10,251 adults with type 2 diabetes and either established cardiovascular disease or multiple risk factors to an intensive glucose control arm targeting HbA1c below 6%, or to a standard arm targeting 7 to 7.9%. Hertzel Gerstein co-led the glycemia component. The multicenter ACCORD Study Group enrolled patients across North America, with a median baseline HbA1c of approximately 8.1% and median diabetes duration of around ten years.

The intensive arm was halted in February 2008, after a mean of 3.5 years of follow-up, when the Data Safety Monitoring Board identified that all-cause mortality was higher in the intensively treated group, with a hazard ratio of 1.22. Cardiovascular death accounted for most of the excess. The announcement came at a time when both ADVANCE (published the same month) and the VADT trial in veterans (published soon after) were also reporting results, and together the three trials reframed the glycemia question for high-risk patients.

The mechanism behind the mortality signal was never definitively established. Severe hypoglycemia occurred more frequently in the intensive arm, and hypoglycemia-driven arrhythmia was a proposed explanation, but analyses trying to link hypoglycemic events directly to mortality were inconclusive. Other proposed explanations included the rapid rate of HbA1c lowering in the intensive arm, the use of combination pharmacotherapy with multiple agents simultaneously, or possible detrimental effects of specific drugs. No single mechanism was confirmed.

ACCORD did not negate tight control broadly. It specifically demonstrated harm in a population of high-risk, longer-duration patients pushed to near-normal glycemia through intensive pharmacotherapy. Combined with ADVANCE and VADT, it shifted guidelines toward individualized targets and away from a single HbA1c goal for all patients with type 2 diabetes. Current guidelines from the ADA and other bodies explicitly recommend loosening targets in older patients, those with long disease duration, established cardiovascular disease, or history of severe hypoglycemia. In 2012 the ADA and the European Association for the Study of Diabetes called for a patient-centered approach in a joint position statement.