Scientific findings reflect the best knowledge available at a given time.
New data, better methods or a change in scientific interpretation may require earlier assessments to be confirmed, refined, qualified or corrected. That is why we never consider an assessment in isolation from its specific context.
What matters is not simply whether a study reports a positive result. Among other things, we examine the study design, population, substance or product form, dose, duration, comparison group, endpoints and methodological limitations. Mechanistic and clinical evidence remain distinct: a biologically convincing mechanism is not yet evidence of a health benefit in humans.
For us, scientific integrity means drawing conclusions as clearly as the data allow — and being equally clear about the limits of what they can tell us.