Read on Twitter
Here it is. The "Why Nutritional Epidemiology of Chronic Disease (henceforth, NECD) Sucks" thread.
Let me be very clear: my animadversion has to do specifically with epidemiological studies that purport to link foods, diets, or dietary patterns to chronic diseases. 1/10
Let me be very clear: my animadversion has to do specifically with epidemiological studies that purport to link foods, diets, or dietary patterns to chronic diseases. 1/10
NECD sucks because: Confounding factors that affect associations seen in NECD studies, such as income, education, and other health-related behaviors, cannot be entirely “stripped of their metabolic consequences by sophisticated statistical methods” (Willett, 1998). 2/10
NECD sucks because: The entire field lacks a guiding theoretical framework.
[If I had to state one for NECD it would be something like, "We believe healthy food is identified by asking healthy people what they eat; what healthy people eat is therefore "healthy food."] 3/10
[If I had to state one for NECD it would be something like, "We believe healthy food is identified by asking healthy people what they eat; what healthy people eat is therefore "healthy food."] 3/10
NECD sucks because: Lack of a guiding theoretical framework means variables [in a model] may be treated by the researchers as “self-evident, requiring no analysis, or else simply a matter of idiosyncratic inspiration (or ideological proclivities)” (Krieger, 2011). 4/10
NECD sucks because: The size of associations between foods or dietary patterns and chronic disease outcomes is typically small; relative risks on the order of 0.8–1.2 are common. 5/10
NECD sucks because: NECD studies do not permit researchers to distinguish between causal and coincidental associations. 6/10
NECD sucks because: In NECD, weak associations could be causal; strong associations could be coincidental. 7/10
NECD sucks because: There is a great deal of methodological and interpretive malleability that comes with weak associations, the inability to differentiate between causal and coincidental associations, and the lack of a guiding theoretical framework. 8/10
NECD sucks because: The malleability of NECD methods & theory accommodates researchers’ own biases and inclinations, which then allows for diametrically opposed conclusions to be drawn about the same food or diet and its impact on chronic disease. 9/10
NECD sucks because: Any "science" that consistently allows diametrically opposed conclusions to be drawn doesn't deserve to be considered a science.
In fact, I'm pretty sure that makes NECD the *opposite* of science. 10/10
In fact, I'm pretty sure that makes NECD the *opposite* of science. 10/10
