EFSA’s Panel on Dietetic Products, Nutrition and Allergies (NDA) last week rejected Malta-based ICP Ltd’s claim that a marine-derived herbal extract called Padina pavonica could increase bone mineral density (BMD) despite the presence of four, peer-reviewed human intervention studies plus other research.
The brown algae extract was present in a prototype product called Dictyolone.
In calling out EFSA as bullies, ICP (Institute of Cellular Pharmacology) managing director Charles Saliba told us this morning that he could not understand the logic behind the NDA’s rejection of the statistical methods employed in the studies, despite those studies being accepted by “respected” peer review journals.
“We had several stop-the-clock procedures with EFSA and were asked the same questions over and over that delayed the initial publication of this by eight months but the reasons for the statistical method problem were not clarified,” Saliba said of the dossier that cost 100s of 1000s of euros to compile.
“So we are contesting the reasoning of this opinion.”
The company was working with its lawyers and consultants to prepare the appeal and has 30 days from January 10 to submit it to the European Commission.
Scientific bullying
Saliba said his firm was not put off by the near 100% appeal rejection rate of appeals to negative NDA health claim opinions.
“The problem with EFSA is that they are scientific bullies and they ride over everyone and we feel we have a duty to stand up to them. Businesses in the US, in Asia do not have to deal with these problems when perfectly legitimate science is in play; Europe is shooting itself in the foot in this way.”
“And don’t get me wrong – we support the principles of this regulation and we know there are many bad products making bad claims on the market and they should be removed. The regulation is good for that. But we had a dossier weighing 6kg in paper and a product backed by serious science.”
“It’s making it impossible to do business in Europe. People look elsewhere and in a global economy that is easy to do.”
Consistent?
Saliba said the fact EFSA rejected studies focused on the active Padina pavonica extract contradicted earlier opinions where EFSA had backed particular extracts over whole products if the evidence showed their activity in the body.
It had done that in an opinion where the NDA said there was sound scientific data to demonstrate short-term constipation alleviation from hydroxyanthracene root and rhizome derivatives in a food supplement called Transitech made by French firm, Vivatech.
NDA reasoning
In its opinion the NDA wrote off two of the four human intervention trials because the studied formulation did match Dictyolone.
Another randomised, open label, parallel human study placed 40 post-menopausal women split into two groups to receive 200 mg of Dictyolone plus 450 mg calcium carbonate or 450 mg calcium carbonate only. These were taken daily for a year but the panel found no statistical difference in various bone outcomes.
Another trial following the same protocols over 12 months split 40 post-menopausal women to receive either 200 mg of Dictyolone or 450 mg calcium carbonate. BMD measurements were taken at baseline, six and 12 months.
The NDA observed: “Absolute values of BMD were transformed to percent changes of BMD prior to hypothesis testing. When using the Mann Whitney test, after 12 months the BMD was increased (as percent change from baseline) in the group which had consumed the Padina pavonica-extract in Dictyolone in lumbar spine and in the femoral neck, when compared with the control group.”
But it then discounted the results by adding: “When an analysis of variance (ANOVA) was performed on such transformed (i.e. percent changes) data, a statistically significant difference between the placebo and the Padinapavonica-group was found at the femoral neck only and not at the lumbar spine.”
“When ANOVA was performed with absolute values of BMD, there were no statistically significant differences between the placebo and the Padina pavonica group at any time point. The repeated measures design was not taken into account in any of the provided analyses.”
“The Panel notes the limitations of the statistical analyses performed by the applicant (i.e. transformation of data, lack of baseline adjustments in the analysis of absolute values, lack of consideration of the repeated measures design of the study) and that the requested re-analysis of data addressing these limitations was not presented by the applicant.”
“The Panel also notes that the results of the different statistical analyses, i.e. percent change in BMD analysed with the Mann Whitney test and ANOVA, and absolute values of BMD analysed with ANOVA, were inconsistent both between the different statistical tests and between bone sites investigated.”
It therefore concluded: “The Panel considers that this study did not show an effect of the Padina pavonica-extract in Dictyolone on BMD.”
The NDA opinion can be found here.