Relation of severe coronary artery narrowing to insulin or thiazolidinedione use in patients with type 2 diabetes mellitus (from the Bypass Angioplasty Revascularization Investigation 2 Diabetes Study)

Abstract

Patients with diabetes continue to die of coronary artery disease (CAD) at rates 2 to 4 times higher than patients without diabetes, despite advances in treatment of cardiovascular disease. The role of glycemic control therapies, independent of their glucose-lowering effects, on cardiovascular disease is a recurring question. We examined the association of glycemic control therapies with extent of CAD as measured by coronary angiogram obtained at baseline in 1,803 subjects in the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial who had type 2 diabetes mellitus, documented moderate to severe CAD, and no previous cardiac revascularization procedures. The association between glycemic control therapy use recorded at baseline and percent coronary artery stenosis and myocardial jeopardy index was analyzed by multiple regression models. Insulin use at study entry was associated with 23% fewer highly stenotic lesions (> or =70%) (p <0.001) and a significantly lower myocardial jeopardy index compared with subjects not on insulin, despite a worse cardiac risk factor profile, more unstable angina, and increased inflammatory markers in insulin users. Subjects taking thiazolidinediones (TZDs) for > or =6 months had 17% fewer highly stenotic lesions (p = 0.02) and significantly lower C-reactive protein, fibrinogen, and plasminogen activator inhibitor-1 levels compared with those not taking TZDs. In conclusion, this cross-sectional study of patients with type 2 diabetes mellitus and CAD showed that treatment with insulin or TZDs was associated with fewer highly stenotic lesions, independent of disease duration, glycemic control, and other risk factors.

Figures

Figure 1

Glycemic control therapies and 50% to 69% (black bars) and ≥70% (white bars) stenosis of coronary arteries. The 1,803 patients in the BARI 2D study without previous revascularization were divided into 5 nonoverlapping groups based on their history of glycemic control therapies at baseline—no diabetes medication; insulin sensitizers (ISs), metformin and/or a TZD only; insulin providers (IPs), a sulfonylurea and/or repaglinide and/or nateglinide only; combination of IS and IP medications (IS + IP); and insulin alone or together with any other medication. Degree of stenosis in coronary arteries was determined on entry coronary angiogram at the BARI 2D core angiography laboratory. Data shown are means ± SEMs. For intermediate lesions, the insulin group was significantly different from the no-diabetes medication group (p = 0.03). Overall difference among the 5 nonoverlapping medication groups for highly stenotic lesions was significant (p <0.001).

Figure 2

Joint association between TZD and insulin use and adjusted (Adj.) mean ratios for highly stenotic lesions. Adjusted mean ratios for number of highly stenotic (>70%) lesions were calculated for patients based on insulin and TZD use at baseline. The TZD group included patients who reported TZD use ≥6 months and any other diabetes medications, except insulin. The insulin group included patients who reported insulin use ≥3 months and other diabetes medications, except TZDs. The TZD + insulin group included patients with reported insulin use ≥3 months and TZD use ≥6 months and any other diabetes medications. Data were adjusted for age, gender, tobacco use (current, previous), glycosylated hemoglobin, diabetes duration, hypertension (blood pressure >130/85 mm Hg, >1 antihypertension drug), lipids, and microalbuminuria; 95% confidence intervals (error bars) are displayed. The p values for the insulin and TZD + insulin use groups were <0.001 compared with the no-insulin, no-TZD group.