Effect of Oral L-arginine Supplementation on Blood Pressure

A Meta-analysis of Randomized, Double-blind, Placebo-controlled Trials

Jia-Yi Dong, BSc; Li-Qiang Qin, MD, PhD; Zengli Zhang, MD, PhD; Youyou Zhao, PhD; Junkuan Wang, PhD; Fabrizio Arigoni, PhD; Weiguo Zhang, MD, PhD


Am Heart J. 2011;162(6):959-965. 

In This Article


Search Strategy

We attempted to follow the Preferred Reporting Items for Systematic Reviews and Meta-Analyses[5] guidelines in the report of this meta-analysis. We searched PubMed, Cochrane Central Register of Controlled Trials, and the ClinicalTrials.gov databases through June 2011 for relevant studies, using terms of "arginine" and "L-arginine" in combination with "blood pressure" and "hypertension." Our search was limited to randomized controlled trials of oral L-arginine supplementation in humans. In addition, we systemically searched the reference lists of obtained articles. No attempt was made to identify unpublished studies.

Study Selection

Studies were included if they: (1) were randomized, double-blind, placebo-controlled trials; (2) used oral L-arginine supplementation as intervention; and (3) reported the net changes of BP and the associated standard deviations (or data to calculate them). Studies were excluded if they: (1) had a short duration of intervention (<1 week); (2) had L-arginine administered by infusion; (3) were single-blind or open-label; (4) used L-arginine as part of intervention; (5) or lacked a concurrent placebo-controlled group.

Data Extraction

We recorded study characteristics as follows: first author's last name, publication year; design details, including whether parallel or crossover; study duration; number of participants; antihypertensive medication use; daily dose of L-arginine treatment; and adverse effects. Participant characteristics including health status, mean age, and baseline BP were also recorded. Further, we assess the methodological quality of each included trial using the Jadad scale, which assigned scores for reported randomization, blinding, and withdrawals.[6] Two of the authors independently performed the literature search, data extraction, and bias assessment, with disagreements resolved by discussion.

Statistical Analysis

For parallel trials, the net changes were calculated by the difference (intervention minus control) of the changes (final values minus baseline values) of the mean values. For cross-over trials, the net changes were calculated as the difference of mean values at the end of the intervention and control periods. Where necessary, standard errors, CIs, and P values were converted to standard deviations for the analysis. Standard deviations for changes from baseline in each group were obtained. If not specified, we computed the missing standard deviations using the method proposed by Follmann et al[7] in which a correlation coefficient of 0.5 was assumed.

The homogeneity of the effect size among studies was tested using the Q test at the P < .10 level of significance. We also calculated the I2 statistic,[8,9] a quantitative measure of inconsistency across studies. An I2 value > 50% was considered to indicate substantial heterogeneity between trials. Either a fixed-effects or, in the presence of heterogeneity, a random-effects model was used to calculate the combined effect size. We did not conduct subgroup analysis because of the small number of included studies. Rather, we performed sensitivity analyses to explore potential sources of heterogeneity across studies and to test the robustness of the results based on various criteria regarding the magnitude of BP reduction, the duration of intervention, and the use of antihypertensive medication. We also investigated the influence of a single study on the overall effect estimate by omitting one study in each turn. Furthermore, we conducted meta-regression analyses to assess whether BP reductions were related to study or subject characteristics, including L-arginine dose, intervention duration, and baseline BP levels. Potential publication bias was assessed by Begg's test[10] and Egger's test[11] at the P < .10 level of significance. All analyses were performed using STATA version 11.0 (StataCorp, College Station, TX). P < .05 was considered statistically significant, except where otherwise specified.

The study was sponsored by Nestec Ltd, Vevey, Switzerland. No other extramural funding was used to support this work.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.