>> Most of the "Large N" studies are based on "diary data." Essentially, asking people via written surveys to estimate what they typically eat and extrapolating

>> Most of the "Large N" studies are based on "diary data." Essentially, asking people via written surveys to estimate what they typically eat and extrapolating that over years. Yeah, not kidding. That is one kind of study. There are also studies that measure the amounts of nutrients, or their byproducts, excreted by study subjects. These usually have small study populations or are restricted to hospital patients etc. There are also in vitro studies or in-vivo interventional studies that are most commonly conducted on animal models, like the study in the article. So like you say, many of the studies that have the largest populations are questionnaire studies, or have populations that are not very representative of the general population. But it's very difficult, if not impossible in practice, to do any other kind of study on a large population, because it's very difficult to observe what a few thousand people eat day-in, day-out. The logistics of such a study would be absolutely mad. The results of such large-scale diary studies do not have great statistical power, but you are not right to say that there is a "pervasive lack of scientific rigor". A study can be rigorous while having low statistical power. Neither does a study absolutely need to conduct double-blind RCTs to be considered rigorous. For example, cosmologists cannot conduct experiments on black holes. Anyway "rigor" is not well-defined, whereas statistical power is an objective measure that can be calculated from the elements of a study. Whether this is done right is another matter and it depends on the study, but nutrition science is not exceptional in this. Btw, I'm not a nutritional scientist. But I hear your complain often and while I understand your frustration, I think you're allowing the perfect to be the enemy of the good. We know some things about nutrition, many with low certainty. But, what are we going to do? Ignore all the knowledge and evidence we have because it's not as much or as good as we'd like, and instead ignore the important questions about the food we eat? Edit: another thing I've noticed is that your comment accuses nutritional science of not being rigorous but at the same time it freely uses concepts from nutritional science, such as "carbohydrates" and "calories". Even grouping foods together in particular groups like "carbs" is a practice that comes from nutritional science. And nutritional science will certainly tell you that carbohydrates are high-energy foods that can cause weight gain if eaten in excess. So you, too, must accept that there is something in the work of nutritional scientsts that is useful to you. In that case, if I were you (if I was trying to adress some problems in the way I eat that is) I'd be very careful not to allow myself to selectively pick only those parts of nutritional science I like, and dismiss all the rest as "sloppy science". It's important to be aware that we all have very strong confirmation biases when it comes to food (me, for example, I make cheese so I'll never, ever believe it can be bad for you).