Clinical Trials and Supplements
Jokes aside, having research papers that support your dietary supplement’s health claims is a huge boon in most global markets. In many cases, it allows for product specific health claims that rise above generic ones. It gives extra confidence and reassurance to consumers that what they’re investing in for their health will actually work for them. There are many different kinds of research articles that might be whipped out in support of a product or ingredient so we’re going to examine some of the different ones and what sets apart a good study from a poorly designed one.
First up – when it comes to scientific research papers, we can divide them into two kinds: observational and interventional.
- Case Control: 2 existing groups with different outcomes (case vs control) are compared to see what causal factors may exist for a certain disease/condition. Always retrospective, past looking. Present: disease. Past: exposure.
- Cohort: follows a group of linked (due to something shared – e.g. exposure to a disease, toxin, workplace environment, participation in a program, diet, etc.) people over time. May compare them to another group not affected/exposed. Prospective, future leaning. Stronger than Case Control studies. Present: exposure. Future: disease? The Fukushima nuclear disaster and future incidence of thyroid cancer is a recent example.
- Observational studies are sometimes the only way to study or examine things that would be unethical to intentionally do in a clinical trial but they aren’t as strong to show causation and may have many confounding factors. They are great for generating hypotheses to test in randomized, controlled trials (RCTs), even if they don’t always pan out (e.g. vitamin E and coronary heart disease)
- Many nutritional and dietary studies are in this category and rely a lot on recall questionnaires, which have limitations due to memory and honesty
- Interventional (RCTs): Randomized Controlled Trials– comparing two groups – placebo and control – blinded or double-blinded. The randomized part is who gets the placebo and who gets the active “ingredient” (or whatever it may be) – this is done so the two groups are as equal as possible in terms of the people and their characteristics. Double-blinded means even the researchers don’t know who gets the placebo and who doesn’t. These are the best in terms of showing cause and effect but also the most expensive to do.
- Meta-Analysis – a statistical approach to combine data across similar studies. They should state the selection criteria for the included studies and search term criteria in finding them – to avoid selection bias – this is the opposite of ‘cherry picking’. It can be challenging to make a good one if the studies on a given topic all have wildly different parameters from each other.
- Review – thorough review and summary of literature to examine and answer a certain question or topic. It doesn’t necessarily involve using stats and combining clinical trial data like a meta-analysis.
Things to look for in a paper:
- is the design appropriate and has the right controls?
- Is the number of people (subjects) really small or more significant (at least 50+)?
- Was it properly randomized? Is the sample of people representative?
- Sample size (n=): must be large enough to detect a real clinically significant effect. This is determined through statistical analysis before the trial starts. Statistical results will usually be reported with a probability value (‘p-value’). A p-value of less than or equal to 0.05 is usually considered ‘significant’, meaning likely a real effect or result. It means that there’s a 5% or less chance that the results were just a coincidence rather than actually due to whatever the intervention (drug, herb, vitamin, etc.) was.
- Do other studies agree or disagree with it?
- Do the conclusions follow from the results or are things overstated?
- Are there facts in the full paper not mentioned in the abstract (summary)?
- What in the design was poorly done (was the dosage too low/high, was the duration too short/long, were there confounding factors not accounted for, etc.)?
- Randomization – researchers should be blinded to control vs. active groups as well. Proper randomization helps to evenly distribute confounding factors.
Clinically Proven Flora Products
Some Flora products with clinically proven ingredients include:
As an example, the ashwagandha root extract we use in Stressveda™ was the subject of this study for stress. We can see that it’s an interventional study involving 64 adults, who were randomly assigned either a placebo or the herbal extract and that the study was double-blinded, so no one, not even the researchers, knew who was getting what until the big reveal at the end of the study.
We can see the dosage (300 mg twice daily) and duration (60 days) and then the results that the people taking ashwagandha each day had a significant reduction in stress, based on a standard stress questionnaire. The p-value is < 0.0001, meaning there’s a 99.999% likelihood that the results were due to the ashwagandha and not just chance.
Something that makes this study even stronger is that they do not rely on just a questionnaire to gauge stress reduction – they also tested cortisol levels, a key stress hormone. They also found a highly significant reduction here for the people taking ashwagandha, with a p-value of <0.0006. The authors concluded that the ashwagandha “safely and effectively improves an individual’s resistance towards stress and thereby improves self-assessed quality of life.”
As the saying goes ‘knowledge is power’, so hopefully the next time you read ‘clinically proven’ on a product or ingredient, you’ll have a better sense of what may be involved and, even better, can check out the study details and see if you think the product meets the hype.
About the Author:
Robert Dadd is a Master Herbalist (Dominion Herbal College) with a BA in Communications from Simon Fraser University. His areas of research include adaptogens, probiotics, and essential fatty acids. He is currently the Product Information Supervisor for Flora Manufacturing and Distributing.