Objective
This meta-analysis was performed to assess the effects of L-arginine supplementation on indices of glycemic control, including fasting blood glucose (FBG), hemoglobin A1c (HbA1c), serum insulin and homeostatic model assessment of insulin resistance (HOMA-IR) levels in randomized controlled trials (RCTs).

Search strategy
This study conducted a systematic review of RCTs published in PubMed, Scopus, Web of Science, Cochrane Library and Embase, up to 5 May, 2018.

Inclusion criteria
Studies were included in this meta-analysis if they were RCTs with parallel design and reported sufficient data on participants before and after intervention, and outcomes of glycemic profile parameters in both the arginine supplementation and control groups.

Data extraction and analysis
The screening of titles and abstracts was performed independently by two reviewers. Selected articles were considered if they met the study’s inclusion criteria. The quality of included studies was assessed by using the Cochrane Collaboration modified tool. From 710 articles retrieved in the initial search, only 10 trials were suitable for pooling the effects of arginine supplementation on serum glucose, insulin, HOMA-IR and HbA1c levels, with effect sizes of nine, eight, five and five, respectively.

Conclusion
Although the results of this study confirmed that supplementation with L-arginine could have significant effects on some glycemic profile indices of participants in clinical trials, the clinical importance of this reduction may not be meaningful.

Results
Pooled random-effect analysis revealed that L-arginine supplementation could significantly decrease FBG level (weighted mean difference [WMD]: 3.35 mg/dL; 95% confidence interval [CI] = [–06.55, –00.16]; P = 0.04) and serum insulin level (WMD: –2.19 μIU/mL; 95% CI = [–3.70, –0.67]; P = 0.005). However, the effects of L-arginine supplementation on HOMA-IR and HbA1c were not significant. Results of subgroup analysis showed that supplementation with L-arginine could significantly decrease serum insulin levels when the dosage of L-arginine is > 6.5 g/d (WMD: –3.49 μIU/mL; 95% CI= [–5.59, –1.38]; P = 0.001), when the duration of supplementation is ≤ 12.8 weeks (WMD: –3.76; 95% CI = [–6.50, –0.98]; P= 0.008), when the participants are not diabetic patients (WMD: –2.54 μIU/mL; 95% CI= [–4.50, –0.50]; P= 0.01) and when the baseline serum level of insulin was > 20 μIU/mL (WMD: –3.98; 95% CI = [–6.31, –1.65]; P = 0.001).

Key words: L-arginine, Fasting blood glucose, Insulin resistance, Hemoglobin A1c, Glycemic profile, Meta-analysis