The current recommendation for patients with diabetes is to maintain A1C levels of 8 percent or less to reduce long-term complications of type 2 diabetes. Most patients with type 2 diabetes eventually need to take more than two agents to achieve and maintain this blood glucose goal. More than 50 percent of patients who have type 2 diabetes for more than 15 years require insulin therapy. Although insulin therapy has been shown to be beneficial in patients in research studies, it is not as effective in clinical settings, perhaps because general patients are less motivated and have fewer resources.

Thiazolidinediones, a new class of oral hypoglycemic agents, have been shown to improve glucose control when used as add-on therapy in patients already taking metformin and a sulfonylurea. Aljabri and associates evaluated the efficacy of adding pioglitazone, a thiazolidinedione, or bedtime NPH insulin to maximal dosages of metformin and an insulin secretagogue in patients with type 2 diabetes who have poor glucose control.

The trial was a nonblinded, open-label, randomized controlled study involving 62 patients with poorly controlled type 2 diabetes, defined as an A1C level greater than 8 percent. Eligibility criteria included age between 30 and 85 years, having type 2 diabetes for at least one year, taking maximal doses of metformin and an insulin secretagogue, and undergoing stable treatment for diabetes for more than 12 weeks. Exclusion criteria included previous treatment with insulin or a thiazolidinedione, or class III or IV New York Heart Association heart failure, myocardial infarction, or stroke within the past six months.

Before the study, patients had baseline measurements of fasting blood glucose and A1C levels, as well as a lipid panel and assesment of liver and kidney functions. Participants were supplied with home glucose monitors and instructed to keep written records of their measurements.

Patients were assigned randomly to receive NPH insulin or pioglitazone at bedtime in addition to their usual doses of diabetes medications for 16 weeks. The NPH insulin therapy was started at 0.3 units per kg and titrated in an attempt to achieve a fasting blood glucose level of less than 108 mg per dL (6 mmol per L). Pioglitazone therapy was started at 30 mg per day and, after four weeks, the dosage was increased to 45 mg per day if the fasting glucose goal had not been achieved. The main outcome measurements included glycemic control, hypoglycemia, blood pressure, lipid levels, microalbuminuria, and quality of life.

Insulin therapy reduced the A1C level by 2.3 percent, and pioglitazone therapy reduced it by 1.9 percent during the study. This was not a significant difference in reduction. A significant number of patients receiving insulin were more likely to develop hypoglycemia than those who received pioglitazone. Pioglitazone significantly increased high-density lipoprotein (HDL) cholesterol levels, while insulin therapy had no impact on this cholesterol factor. There were no differences between the two groups with regard to other lipid values, weight, blood pressure, and urine microalbumin levels. Patients in both groups had similar total scores on a quality-of-life survey.

The authors conclude that the addition of pioglitazone or bedtime insulin to maximum dosages of metformin and insulin secretagogue improves glycemic control during a 16-week trial. Compared with insulin therapy, pioglitazone causes fewer hypoglycemic episodes and improves HDL cholesterol levels.