Monday, May 07, 2018

Why Do We Report on Diet Studies As If Patients Adhered To Their Prescribed Diets?

It happens all the time.

Researchers recruit participants for a diet study and randomize them to particular diets.

The participants don't adhere to said diets particularly well (and this is putting aside the known and glaring inaccuracy of dietary recall), and the macronutrient breakdowns of what they report eating are nearly always different from the macronutrient breakdowns of what they were instructed to eat.

And yet the researchers write up their studies as if their prescribed diets were followed, and consequently so too do journalists.

So here are my two asks.

1. Unless you're writing or reporting on diet studies conducted in metabolic wards, please treat their results with tempered enthusiasm.

2. And if the difference between two dietary outcome arms isn't likely to be clinically meaningful (like say the difference between weight losses of 3.3 kg, 4.6 kg, and 5.5 kg at two years (and bonus points if you get this reference right out of the gate)), maybe hold off on declaring one diet to be the world's best?