Bret Scher, MD - Figure 26
Lower quality evidence
We can dig deeper and find some other concerns with this type of data. This is nutritional epidemiology – observational studies – which in the hierarchy of evidence is lower quality. That doesn't mean we should totally disregard it, but we have to understand that it is lower quality. For one, these individuals filled out a food frequency questionnaire maybe twice over a 25-year period. That's not the most accurate data collection. Second, the hazard ratios are very low at 1.2. For reference, if you remember studies that looked at smoking and the risk of cancer, those were epidemiological studies with hazard ratios of 15-20 or more, depending on the study. Not 1.20, but 20. And here we're talking about 1.20 – a very, very small increased risk in the context of observational studies.