The type of study you run determines the strength of the claims you can make. An observational study and a randomised controlled trial are both "clinical studies," but they produce fundamentally different evidence, and regulators and retailers treat them differently.
For most supplement brands seeking to make commercial claims, the study design question has a straightforward answer. This guide explains why.
What an observational study is
In an observational study, researchers observe participants without intervening. No one is randomly assigned to a group, and there is no placebo comparison. Researchers collect data on what participants naturally do or experience and look for patterns.
Examples in a nutrition context: asking a group of people about their supplement habits and measuring their health outcomes over time; tracking how customers report feeling after using a product; examining the dietary patterns of a population and correlating them with health markers.
What observational evidence tells you: that there is a relationship between two variables. What it cannot tell you: whether one variable caused the other. This is the foundational limitation of observational research. Confounding factors (other differences between people who take a supplement and those who do not) mean you can never be certain the supplement is responsible for any observed difference.
What a randomised controlled trial is
An RCT randomly assigns participants to one of two (or more) groups. In a placebo-controlled RCT, one group takes the active product and the other takes a placebo that looks and tastes identical. Neither group knows which they received (this is the "double-blind" element). At the end of the intervention, both groups are compared on the pre-specified outcomes.
What the RCT evidence tells you: because participants are randomly assigned, the groups should be similar in all respects except the product they received. Any statistically significant difference in outcomes between groups can therefore be attributed to the product. An RCT produces causal evidence: it tells you the product caused the outcome, not just that the two are associated.
This is why the RCT is the gold standard in clinical research, and why regulators and retailers treat it differently from observational studies.
What the ASA accepts
For claims that imply your product caused a specific outcome ("shown to improve cognitive focus," "clinically demonstrated to support energy levels"), the ASA expects randomised controlled trial evidence. The rationale is the same as above: only an RCT can establish causation, and only causal evidence supports an efficacy claim.
Observational study evidence may support lower-level claims (for example, describing what a product is associated with in a customer sample), but it is generally not sufficient to substantiate a "clinically shown to..." style claim in paid advertising. An ASA upheld ruling on a claim backed only by observational evidence is a real risk, particularly in a sector the ASA is actively monitoring.
What retailers accept
Boots and Holland & Barrett, in line with the ASA standard, expect RCT evidence for specific efficacy claims. An observational study may support inclusion of a product in a range, but it is unlikely to support the kind of premium positioning or on-pack claim language that differentiates your product from generic alternatives.
Variations on the RCT design
Not all RCTs look the same. The most important variations for nutrition brands are:
Parallel group design
The most common RCT design. Participants are randomised to either active or placebo groups and stay in that group for the full intervention. Simple to explain and widely accepted by journals and regulators.
Crossover design
Each participant takes both the active product and the placebo in sequence, with a washout period in between. Because each person acts as their own control, crossover designs need fewer participants than parallel group designs, making them cost-effective for nutrition studies. They work well for interventions measuring reversible outcomes (cognitive function, energy, blood glucose, sleep quality) where the effect can wash out between periods.
Open-label design
Participants know they are taking the active product (there is no blinding). This is acceptable for some studies but produces weaker evidence for subjective outcomes, because participants knowing they are taking the active product can influence how they report feeling. Open-label designs are appropriate where blinding is impractical (for example, studies involving exercise or dietary interventions).
When observational evidence is useful
Observational studies are not without value. They are useful for:
- Preliminary research. Before investing in an RCT, a small observational study can check whether there is a signal worth pursuing, which reduces the risk of an RCT producing null results.
- Long-term outcomes. Outcomes that take years to develop (long-term cardiovascular health, chronic disease risk) are impractical to study in a controlled intervention of reasonable duration.
- Real-world use data. Observational post-market studies can describe how customers actually use your product and what they report experiencing, which is useful for marketing context without making causal claims.
For most supplement brands at the point of wanting to make commercial claims, however, an observational study alone is not a substitute for an RCT.
If you want to say your product works, run an RCT. If you want to understand who is buying it and why, run an observational survey. The two serve different purposes and neither replaces the other.
Nathan Phillips, CEO, StudySetGoFrequently asked questions
An RCT randomly assigns participants to active product or placebo and controls for all other variables, allowing you to conclude the product caused the outcome. An observational study watches what happens without randomisation, so it can show association but not causation. For commercial claims, the distinction matters significantly: the ASA expects RCT evidence for efficacy claims.
Not for efficacy claims. If your ad says your product "clinically improves" or is "shown to" cause a specific outcome, the ASA expects RCT evidence. Observational study evidence may support lower-level descriptive claims but is generally insufficient for causal or efficacy-type claims.
A crossover trial has each participant take both the active product and placebo in sequence, making them their own control. This requires fewer participants than a parallel group design and is commonly used in supplement studies measuring reversible outcomes like cognitive function, energy, or blood glucose. It is as rigorous as a parallel group RCT when properly designed.
The right design depends on your specific claim, the outcome being measured, your budget, and your timeline. A CRO will recommend the appropriate design based on your objectives. Book a discovery call and we will scope the most suitable study type for your product and commercial goals.