Headlines about food additives and their potential health risks pop up regularly, often citing the French NutriNet Santé study—for example this week about additives and diabetes risk. This observational cohort is very well designed and provides a huge amount of great data—but it is mainly the data on additives that are widely used to argue against ultra-processed foods (UPF). But does it really tell us that additives are bad? Or does it just tell us that certain dietary patterns are associated with increased disease risk.

The first question to ask here is: do they know how much additive each participant consumes? Or is it an estimate? After all, as I’ve explored before ad nauseam, shaky data can lead to shaky conclusions (see here).

In the EU and UK, food additives are tightly regulated. The European Food Safety Authority (EFSA) for example reevaluates them regularly. These evaluations are thorough, diving into everything from toxicology to exposure assessment—i.e. estimating how much of an additive people consume at a population level.

Read the full post on Substack →

Originally published on Substack.