Association of l-Arginine Supplementation with Markers of Endothelial Function in Patients with Cardiovascular or Metabolic Disorders: A Systematic Review and Meta-Analysis

Systematic review and meta-analysis of 13 randomized controlled trials in adults with cardiovascular disease, obesity, impaired glucose tolerance/metabolic syndrome, and/or type 2 diabetes. Included studies used crossover or parallel placebo-controlled designs, with 723 participants overall and 358 assigned to L-arginine across the active arms.

Oral L-arginine supplementation was tested across 13 randomized placebo-controlled trials in doses ranging from 3 g/day to 15 g/day, using tablets, capsules, powder, bars, or effervescent tablets; treatment durations ranged from 3 days to 18 months. One active regimen also combined L-arginine 1200 mg/day with N-acetylcysteine 600 mg twice daily for 6 months, and one trial paired L-arginine 5 g/day with a hypocaloric diet enriched in legumes.

Overall, oral L-arginine did not improve endothelial function versus placebo across the pooled trials. In the main meta-analysis, blood flow showed no overall benefit (SMD 0.30; 95% CI -0.85 to 1.46; N = 469; p = 0.60), and neither NOx (mean difference 4.41 µmol/L; 95% CI -0.50 to 9.32; N = 336; p = 0.08) nor ADMA (mean difference -0.04 µmol/L; 95% CI -0.15 to 0.08; N = 290; p = 0.53) changed significantly. Sensitivity analyses suggested a possible positive blood flow effect after removing highly heterogeneous studies (SMD 0.59; 95% CI 0.10 to 1.08; N = 323), and NOx improved in obese/type 2 diabetes and cardiovascular subgroups. The authors concluded that larger, higher-quality trials in specific patient subgroups are needed.

The evidence base was small and highly heterogeneous, with substantial variation in dose, duration, formulation, and patient population. Several outcomes were based on crossover studies and subgroup/sensitivity findings, increasing the risk that apparent benefits are unstable or indirect. Risk of bias was present across included studies, and some individual trials had missing or inconsistently reported arm-level details.