May 9, 2013
In Europe, achieving an approved health claim is no easy featonly 222 of more than 2,000 submitted Article 13 EFSA health claims have been passed. The reason given for the rejection of many applications is that the scientific evidence was not substantial enough or did not adhere to EFSA guidelines, which call for trials with very specific parameters and outcomes. This has caused contention in the industry, with some people suggesting the regulations are unfair and disregard perfectly sound science. At the Vitafoods Europe Conference, health claims and regulation will be put under the spotlight; Stoffer Loman, Ph.D., who is speaking at the event, offered INSIDER a look at designing trials and research in light of the EFSA regulation.
INSIDER: When a company is interested in investing in clinical research on a nutritional ingredient, what are a couple of the top considerations they might not immediately know to address?
Stoffer Loman, Ph.D.: First, it is very important to get a clue about how the ingredient affects the parameter of interest (primary endpoint), e.g., weight loss, in terms of effect size (the difference in primary endpoint values between intervention and placebo/control) and the variation that occurs in the measurement of the primary endpoint. This important information will determine the sample size of a well-powered full-scale intervention study. Sufficient statistical power, or the likelihood of detecting a difference between intervention and control that is physiologically relevant, is directly related to the number of subjects needed in the trial. The more accurately the number of participants can be determined, the more cost-effective your trial is likely to be. Moreover, detecting a statistically significant number combined with a physiologically relevant effect in a well-powered study is a pre-requisite for EFSA to be able to draw scientific conclusions from the results.
A second and also highly important consideration is to get a clue about the dose-dependency of the relationship between the ingredient and the parameter of interest (e.g., blood pressure). Showing dose dependency is very convincing evidence of a cause-and-effect relationship. Moreover, dose-dependency may also provide information about the minimal effective dose (that may differ per health benefit for the same ingredient).
A third, but by no means last, consideration is to come to a clear definition of the primary aim of the study, especially in light of the appropriate selection of the control compound.
INSIDER: How has EFSA's new claims paradigm affected how researchers design clinical trials; if it has not, should it have had an effect?
Loman: Frankly, I wouldnt know whether the way EFSA evaluates scientific studies in the context of health claim applications has really affected how researchers currently design clinical trials, but I assume it has. In some areas (e.g., probiotics) there have been initiatives among researchers to come to a consensus on relevant primary endpoints and the methods to measure those. Nevertheless, I think it should have an effect because, from my perspective, generating data that EFSA will accept as pertinent to a claim requires a health claim-focused study design, which is likely to be different from the aim of getting your data published in a peer-reviewed journal. The "health-claim" approach could serve both purposes. However, the opposite has already proved itself insufficient in many cases.
INSIDER: When designing a research trial, does it help to have a target health claim in mind to ensure the data will support the expected claim?
Loman: Of course. Nevertheless, if a company is still in the phase of exploring potential health benefits of its ingredient, it may adopt the approach of designing a study with multiple endpoints, rather than one specific primary endpoint. This study will most likely be of limited scale (pilot trial), but may still turn out to be a fruitful approach to quickly get a clue about the health potentials of the ingredient.
A pilot study will enable the researcher to monitor the effect within various aims without taking power into account too much. After this initial phase, a fully powered main study, based on the outcome of the pilot study, provides a clear focused primary target, However, if it comes to the point of filing of a health claim application, the pilot trial cannot, in most instances, be considered "pertinent" to the claim, even when the outcome of the pilot study bears physiological meaning.
INSIDER: If clinical trial results fall short of the level of desired substantiation, how can companies look at statistical analysis to gain maximum value from their research investment?
Loman: There are various approaches to this issue:
Outlier analysis: trying to identify, on a statistically sound basis, those observations that are completely out of range.
Testing to which extent the distribution in the acquired data requires a change in the appropriate statistical testing.
Co-factor analysis. Establishing the impact of a covariate on the outcome of the trial. It might well be that the results are partially influenced by another factor unknown at the start. By taking into account the impact of this factor the outcome might be totally different.
Again, it should be stressed that an optimal design of the trial upfront will enable much more powerful analyses of the data at the end.
For expert opinion on statistical issues the reader is recommended to contact my co-author and statistical expert Dr. Wim Calame, StatistiCal BV, The Netherlands ([email protected]).
You May Also Like
CRN petition to FTC: RCTs aren’t required to substantiate ‘health-benefit’ claimsSep 22, 2023
Collagen peptide ingredient solutions for seniors’ changing needs – infographicSep 19, 2023
Radicle Insights—Covid Eris and dietary supplements: separating fact from fictionSep 21, 2023
More bioavailable hyaluronic acid complex for today’s beauty-from-within consumer – snapshotSep 18, 2023