Are we really sure about the pycnogenol antihypertensive effect?

We carefully read the meta-analysis by Malekahmadi et al. recently accepted for publication in Pharmacological Research with a little criti- cism [1]. Our concerns regard the fact that not every included study meets the inclusion criteria of the meta-analysis. As a matter of fact, the authors considered 9 non-randomized clinical studies (non-RCTs) (references 26, 28, 33, 34, 36, 38, 41, 44, 45 in the paper), and a clinical trial (reference 24 in the paper) testing the e ect of Oligopin® supple- mentation, which is characterized by a higher content in low molecular weight oligomeric procyanidins than Pycnogenol® (∼20 % vs. 5 %). Including these clinical studies poses a signi cant risk to the in- tegrity of data quality, especially since they enrolled more than half of the subjects overall considered in the meta-analysis. In particular, after a careful look at the forest plots, considering the percentage weight of each individual study, we have reason to think that this methodological weakness has substantially in uenced the e ect sizes of pycnogenol supplementation on the investigated parameters and de nitely a ected the validity of the ndings. Of course, authors claim that the quality of the meta-evidence for the e ect of pycnogenol on cardiometabolic risk factors ranges to high- to-moderate for almost all the signi cant ndings, in agreement with the NutriGrade scoring system [2]. Passing on the improper use of this method -which has been speci cally validated for systematic reviews of RCTs- [3], it has been incorrectly applied and led to misleading results. Similarly, in the Cochrane Collaboration’s tool for assessing risk of bias in meta-analysis (as it is reported by the authors in Supplementary table 1) there are several inaccuracies. To give an example, the study by Belcaro et al. (reference 38 in the paper) has been evaluated at low risk of selection bias (allocation concealment), performance bias (blinding of participants and providers) and selection bias (random sequence generation), though it is not a RCT but has an observational design. Consequently, it is not surprising that the results by Malekahmadi et al. are not entirely consistent with the previous evidence in humans [4,5], and they have also been substantially disavowed by a meta- analysis of double-blind placebo-controlled RCTs recently published by our research group, showing that pycnogenol supplementation has no detectable e ect on blood pressure [6]. Certainly, larger well-designed, longer-term and adequately pow- ered RCTs are needed in future to determine the e ect of pycnogenol supplementation on cardiovascular risk factors. However, until further data become available, it is necessary to be very cautious in drawing denitive conclusions.