RESUMO
INTRODUCTION: Strict glucose control using multiple doses of insulin is the standard treatment for type 1 diabetes mellitus (T1DM), but increased risk of hypoglycemia is a frequent drawback. Regular insulin in multiple doses is important for achieving strict glycemic control for T1DM, but short-acting insulin analogues may be better in reducing hypoglycemia and postprandial glucose levels. OBJECTIVE: We conducted a systematic review and meta-analysis of randomized controlled trials (RCTs) to assess the effects of short-acting insulin analogues vs regular human insulin on hypoglycemia and postprandial glucose in patients with T1DM. METHODS: Searches were run on the electronic databases MEDLINE, Cochrane-CENTRAL, EMBASE, ClinicalTrials.gov, LILACS, and DARE for RCTs published until August 2017. To be included in the study, the RCTs had to cover a minimum period of 4 weeks and had to assess the effects of short-acting insulin analogues vs regular human insulin on hypoglycemia and postprandial glucose levels in patients with T1DM. Two independent reviewers extracted the data and assessed the quality of the selected studies. The primary outcomes analyzed were hypoglycemia (total episodes, nocturnal hypoglycemia, and severe hypoglycemia) and postprandial glucose (at all times, after breakfast, after lunch, and after dinner). Glycated hemoglobin (HbA1c) levels and quality of life were considered secondary outcomes. The risk of bias of each RCT was assessed using the Cochrane Collaboration Risk of Bias table, while the quality of evidence for each outcome was assessed using the GRADEpro software. The pooled mean difference in the number of hypoglycemic episodes and postprandial glucose between short-acting insulin analogues vs. regular human insulin was calculated using the random-effects model. RESULTS: Of the 2897 articles retrieved, 22 (6235 patients) were included. Short-acting insulin analogues were associated with a decrease in total hypoglycemic episodes (risk rate 0.93, 95% CI 0.87-0.99; 6235 patients; I2 = 81%), nocturnal hypoglycemia (risk rate 0.55, 95% CI 0.40-0.76, 1995 patients, I2 = 84%), and severe hypoglycemia (risk rate 0.68, 95% CI 0.60-0.77; 5945 patients, I2 = 0%); and with lower postprandial glucose levels (mean difference/MD - 19.44 mg/dL; 95% CI - 21.49 to - 17.39; 5031 patients, I2 = 69%) and lower HbA1c (MD - 0,13%; IC 95% - 0.16 to - 0.10; 5204 patients; I2 = 73%) levels. CONCLUSIONS: Short-acting insulin analogues are superior to regular human insulin in T1DM patients for the following outcomes: total hypoglycemic episodes, nocturnal hypoglycemia, severe hypoglycemia, postprandial glucose, and HbA1c.