Does tight glycemic control decrease the rate of cardiovascular complications in patients with type 2 diabetes mellitus?

If the authors of this systematic review have captured all the relevant studies, tight glycemic control decreases the rate of cardiovascular complications in patients with type 2 diabetes mellitus. There is strong reason to believe, however, that by excluding studies in which cardiovascular outcomes were secondary, and by relying solely on the published outcomes (and not contacting authors for additional data), the authors had incomplete data from which to draw their conclusions. (LOE = 1a-)

Ray KK, Seshasai SR, Wijesuriya S, et al. Effect of intensive control of glucose on cardiovascular outcomes and death in patients with diabetes mellitus: a meta-analysis of randomised controlled trials. Lancet 2009;373(9677):1765-1772.  [PMID:19465231]

Meta-analysis (randomized controlled trials)

Self-funded or unfunded

Various (meta-analysis)

These authors searched several databases to identify randomized trials of patients with type 2 diabetes mellitus in which intensive glycemic control was compared with standard care. They only included studies in which the primary end point included cardiovascular events and the reporting of results was sufficient to calculate individual event rates. Please note that they excluded 6 large trials because cardiovascular events were not part of the primary study outcomes or because the original papers didn't report individual event rates. Since there were only 5 studies included (33,000 patients), this is a big concern. Some analysts would have contacted authors to obtain missing data. Additionally, if cardiovascular event rates are measured in a comparable manner, it shouldn't matter if these were primary or secondary end points. In the studies the authors chose to include, the overall rate of nonfatal myocardial infarctions was 10 per 1000 patient-years in intensively treated patients compared with 12.3 in control patients. The rate of coronary events was 14.3 and 17.2 , respectively, per 1000 patient-years. The rate of strokes (approximately 7 per 1000 patient-years) and the total death rate was comparable regardless of treatment (approximately 18 deaths per 1000 patient-years). The average difference between glycohemoglobin levels in patients treated with intensive care and with usual care was approximately 1%. In these studies, blood pressure control and control of lipids resulted in greater reductions in cardiovascular outcomes than glycemic control.