This study was designed to see if lowering A1c to 6.5%, instead of the ADA's recommended 7.0%, could prevent heart attacks. The study was stopped early when analysis of preliminary data showed a slight excess of heart attack deaths in the subjects in the group who were striving to lower their A1cs.
This is all most doctors ever heard about ACCORD--that lowering A1c led to an increased risk of heart attack. What they didn't hear about was the methodology used in the study. That methodology makes it very clear that it wasn't the lowering of blood sugars that caused the deaths, but the way the study attempted to lower A1c.
ACCORD studied only people with long-standing Type 2 diabetes who had been diagnosed with heart disease before the start of the study. These patients were put on a statin drug (which we now know can further raise blood sugar) and a fibrate drug.
Then the researchers set out to lower blood sugar by putting their subjects the discredited high carbohydrate, low fat diet--which a large body of research has shown not only raises blood sugar but worses triglycerides and LDL. To counteract the blood-sugar-raising effect of this diet, the ACCORD researchers put the study subjects trying to lower blood sugar on a cocktail of every diabetes drug available at the time.