#472: Compared To What? – Understanding Food Substitution Analysis & Adjustment Models

Links: Subscribe to Premium Go to this episode's page (with links) Live event: London - March 18th Receive 'Sigma Synopsis' emails Description: When thinking about the effect of eating or not eating a certain food or nutrient, we can't consider this in isolation. Meaning, we need to evaluate the impact within the context of what such an inclusion/exclusion does to an individual's overall diet pattern. Thinking about this concept, the phrase "compared to what?" has been colloquially used. And while this is an important idea, there has been some misapplication of this principle. In nutrition science, this is related to the concept of food or nutrient "substitution". And this concept is crucial to understanding the issues that can arise in nutrition studies, particularly when it comes to single food analyses in nutritional epidemiology. This concept of substitution is quite intuitive in controlled feeding studies. However, it is not as obvious when considering nutrition epidemiology studies. As noted by Ibsen & Dahm (2022): "Whereas studying the effects of eating one food instead of another is typically explicit in interventional study designs, it is often implicit and sometimes hidden in analyses of observational studies." However, in nutrition epidemiology substitution is still happening, but it typically emerges as a consequence of adjustment models . In nutritional epidemiology, it is essential to adjust for confounders. E.g., one vital adjustment is often for total calorie intake. However, when our exposure is a specific food/nutrient, we must think about confounding by other foods. So knowing what, and how, a study is adjusting for variables helps us interpret it better. In this episode, Dr. Alan Flanagan and Danny Lennon discuss these crucial ideas of food substitution, adjustment models, and "compared to what?". Go to this episode's page (with links)