ABSTRACT
CONTEXT: Vascular dysfunction and insulin resistance precede atherosclerosis in type 2 diabetes (T2DM). Better knowledge of the interaction between these is of considerable clinical interest. OBJECTIVE: The objective of this study was to examine the association between inflammation, glucose, and lipid metabolism and vascular dysfunction. DESIGN AND SETTING: We conducted a randomized, double-blind, controlled trial of pioglitazone vs. placebo and other therapies aimed at equal glycemic control for 24 wk at an academic tertiary referral clinic. PATIENTS AND INTERVENTIONS: Mexican-American subjects with T2DM and no complications were randomly assigned to pioglitazone 45 mg daily (PIO, n=16) or placebo (CON, n=15) and matched for age, gender, body mass index, diabetes duration, and glycemic control. All subjects completed the study. MAIN OUTCOME MEASURE: We looked for improved vascular reactivity independent of glycemic control but closely related to plasma adiponectin, lipids, and insulin sensitivity. RESULTS: After 24 wk, there was an equal decrease in fasting plasma glucose (approximately 135 mg/dl), glycosylated hemoglobin (approximately 7.0%), and glucose production (approximately 15%). The decrease in free fatty acids (30 vs. 10%) and increase in glucose disposal (40 vs. 25%) were greater in PIO vs. CON (P<0.05). In PIO, plasma high-density lipoprotein rose by 15% (P<0.05), and low-density lipoprotein and high-density lipoprotein particle size rose significantly (P<0.01). Plasma adiponectin doubled in PIO (from 6.1+/-0.8 to 12.7+/-2.1 microg/ml). Forearm blood flow rose equally (approximately 130%) during reactive hyperemia in both groups, although after therapy, the increase was greater (P<0.001) in PIO (153%) than in CON (137%); vasodilation was greater (P=0.01) in PIO (92, 160, and 204%) than in CON with acetylcholine (74, 130, and 144%) and with sodium nitroprusside (PIO=164 and 253% vs. 116 and 230%; P=0.04). The elevation in diameter was also greater in PIO (13 vs. 10%; P<0.05). Vascular responses correlated with plasma free fatty acids, adiponectin, and low-density lipoprotein particle size but not with glycemic control. CONCLUSION: These data indicate that pioglitazone improves vascular reactivity irrespective of glycemic control and suggest a close association with changes in fat cell metabolism.