Association of diet alone, insulin, sulfonylureas, metformin, and thiazolidinediones with the severity of coronary artery disease in patients with diabetes mellitus.

Coronary angiography was performed in 152 men and 163 women with diabetes mellitus, mean age 55 +/- 8 years, because of chest pain. Of 67 patients with 3-vessel or 4-vessel coronary artery disease (CAD), 17 (25%) were treated with diet alone, 29 (43%) with insulin, 18 (27%) with sulfonylureas, 12 (18%) with metformin, and 6 (9%) with thiazolidinediones. Of 76 patients with 2-vessel CAD, 20 (26%) were treated with diet alone, 36 (47%) with insulin, 21 (28%) with sulfonylureas, 21 (28%) with metformin, and 11 (14%) with thiazolidinediones. Of 40 patients with 1-vessel CAD, 15 (38%) were treated with diet alone, 11 (28%) with insulin, 8 (20%) with sulfonylureas, 12 (30%) with metformin, and 4 (10%) with thiazolidinediones. Of 132 patients with 0-vessel CAD, 18 (14%) were treated with diet alone, 21 (16%) with insulin, 7 (5%) with sulfonylureas, 75 (56%) with metformin, and 35 (26%) with thiazolidinediones. Cochran-Armitage trend tests were used to examine whether the use of treatment significantly increases or decreases as the number of arteries with CAD increases (P = 0.036 for diet alone; P < 0.0001 for insulin, for sulfonylureas, and for metformin; P = 0.002 for thiazolidinediones).