Glossary · Methodology
Observational study
In short
An observational study does not intervene but watches: it records what people do anyway and compares how they fare later. It finds associations, but can only with difficulty show what is the cause.
Its strength is size. Following hundreds of thousands of people for decades is hardly feasible with a randomized trial, but it is with registries and biobanks. Many important risk factors, from smoking to blood pressure, were discovered this way.
Its weakness is confounding: people who regularly go to the sauna or take supplements often also differ in income, exercise and pre-existing conditions. Statistical adjustment helps, but only captures what was measured.
Observational studies include cohort studies, case-control studies and cross-sectional studies. Mendelian randomization is an attempt to get closer to the cause using gene variants.
Occurring together is not the same as causing.
Where this comes up in the fact checks
- Sauna & cold – The strongest numbers come from an observational cohort; where randomized, controlled testing was done, the added benefit did not appear.
- Pentadecanoic acid (C15:0) – People with more C15:0 in their blood develop type 2 diabetes less often — as a supplement, the fatty acid has only been tested in two small studies.
- Cutting down on alcohol – The seemingly protective U-curve of moderate drinking is a well-known example of a biased comparison.
Sources
Related terms: Cohort study · Mendelian randomization · Randomized controlled trial (RCT)
Information only, not medical advice. The BK-Score is a subjective assessment according to disclosed rules – not a scientific evaluation. Scoring rules