Am Fam Physician. 2012;85(11):1099-1100
Background: Recent randomized clinical trials comparing intensive blood glucose control with conventional control have not shown a reduction in cardiovascular disease or mortality in patients with type 2 diabetes mellitus. The Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial in 2008 was stopped early because of increased all-cause and cardiovascular mortality in the intensive control group. However, other studies have shown a reduction in microvascular complications with intensive control. Although overall glycemic control likely helps prevent morbidity and mortality, debate persists about the optimal glycemic target and the benefits and risks of achieving an intensively lowered glycemic goal. Hemmingsen and colleagues performed a systematic review and meta-analysis on the effects of intensive glycemic control compared with conventional control on all-cause mortality, cardiovascular mortality, cardiovascular disease, and microvascular disease in patients with type 2 diabetes.
The Study: The authors searched for randomized controlled trials published in any language in the Cochrane Library, Medline, Embase, Science Citation Index Expanded, LILACS, and CINAHL, and reviewed unpublished data. Trials that compared a strict glycemic target with a more relaxed one were eligible, although “strict” or “intensive” control was defined differently in each study. The primary outcomes included all-cause mortality, cardiovascular mortality, nonfatal myocardial infarction, composite microvascular complications, retinopathy, nephropathy, and severe hypoglycemia. The data on each outcome were statistically summarized as relative risks with 95% confidence intervals. To determine clinically relevant results, trial sequential analysis was used to identify information (sample) sizes that could reflect a 10 percent relative risk reduction for mortality (equivalent to a number needed to treat of 100). The meaningful increased relative risk of severe hypoglycemia was set at 30 percent, equivalent to a number needed to harm of 50.
Results: Fourteen trials met the inclusion criteria. These trials included data from 28,614 participants (15,269 randomized to intensive glycemic control and 13,345 to conventional control), and dealt exclusively with glycemic control in the usual care setting. Target A1C levels varied between trials; the lowest was less than 6 percent in the ACCORD and Veterans Affairs Diabetes Trial studies. Some trials used fasting glucose concentrations as a treatment target instead of a predefined A1C value. The conventional care targets varied from achieving an A1C level less than 7 to 8 percent, to avoiding hyperglycemia.
In the meta-analysis, all-cause mortality was not reduced with intensive glycemic control, and this finding was confirmed by trial sequential analysis. Cardiovascular mortality was not reduced in the meta-analysis, but trial sequential analysis showed too little evidence to conclude risk, benefit, or no difference. Although there seemed to be a small but statistically significant risk reduction for nonfatal myocardial infarction in the intensive control group, trial sequence analysis showed a lack of sufficient evidence to support it. Similarly, the small reductions in relative risk for composite micro-vascular complications with intensive control were not confirmed by trial sequential analysis. Conversely, the risk of severe hypoglycemia was significantly increased in the intensive glycemic control groups, a result confirmed by trial sequential analysis.
Conclusion: In this large meta-analysis, intensive glycemic control did not reduce mortality and was associated with a significantly increased risk of severe hypoglycemia.