Objective: Recent data support the hypothesis that microvascular dysfunction may be a potential mechanism in the development of insulin resistance. We examined the association of microvascular dysfunction with incident type 2 diabetes mellitus (T2DM) and impaired glucose metabolism by reviewing the literature and conducting a meta-analysis of longitudinal studies on this topic.
Methods and results: We searched Medline and Embase for articles published up to October 2011. Prospective cohort studies that focused on microvascular measurements in participants free of T2DM at baseline were included. Pooled relative risks were calculated using random effects models. Thirteen studies met the inclusion criteria for this meta-analysis. These studies focused on T2DM or impaired fasting glucose, not on impaired glucose tolerance. The pooled relative risks for incident T2DM (3846 cases) was 1.25 (95% confidence interval, 1.15; 1.36) per 1 SD greater microvascular dysfunction when all estimates of microvascular dysfunction were combined. In analyses of single estimates of microvascular dysfunction, the pooled relative risks for incident T2DM was 1.49 (1.36; 1.64) per 1 SD higher plasma soluble E-selectin levels; 1.21(1.11; 1.31) per 1 SD higher plasma soluble intercellular adhesion molecule-1 levels; 1.48 (1.03; 2.12) per 1 SD lower response to acetylcholine-mediated peripheral vascular reactivity; 1.18 (1.08; 1.29) per 1 SD lower retinal arteriole-to-venule ratio; and 1.43 (1.33; 1.54) per 1 logarithmically transformed unit higher albumin-to-creatinine ratio. In addition, the pooled relative risks for incident impaired fasting glucose (409 cases) was 1.15 (1.01-1.31) per 1 SD greater retinal venular diameters.
Conclusions: These data indicate that various estimates of microvascular dysfunction were associated with incident T2DM and, possibly, impaired fasting glucose, suggesting a role for the microcirculation in the pathogenesis of T2DM.