Difficulty of Designing Nutrition Studies

[matriarco]

Lately I've been thinking about nutrition studies. My sister is getting her PhD in nutrition and we've been talking a lot about how difficult it can be to find/do nutrition studies.

So we all know there are a lot of bogus claims out there about nutrition. Soy gives you boobs! Blueberries cure everything! And there's a general consensus about basic nutrition stuff. But in the case of most individual nutrients, we really don't know exactly what they do. It's incredibly hard to design an accurate study of nutrients, or even specific foods, because it means regulating everything people eat, which is really difficult.

If we're trying to eat as heathfully (healthily) as possible or just to discuss nutrition based on good science, how do we balance that with the possibility that we may not know for decades (if ever) what that means in terms of individual nutrients? In this case, absence of proof doesn't always mean something isn't true. Sometimes it just means it hasn't been tested. If we say, "There's no evidence of that!" when it hasn't been tested are we representing it as definitely untrue? And if we make our nutrition decisions solely on what HAS been studied, are we missing important stuff?

In no way am I saying we should fall for pseudoscience. I mean something like that there's no study definitively showing that a cup of leafy greens a day is beneficial but that doesn't mean I should decide not to eat a cup of greens a day.

[Dandelion]

In no way am I saying we should fall for pseudoscience. I mean something like that there's no study definitively showing that a cup of leafy greens a day is beneficial but that doesn't mean I should decide not to eat a cup of greens a day.

Well, we know about the nutrient composition of foods and deficiency (or over abundance) of those nutrients are correlated with health issues so recommendations of what to eat are guided by that, no? My RD recommended (certain) leafy greens and I trust her to do her job in keeping up with the latest nutrition information so it's credible enough for me.

I'm not sure if I understood your question though.

[matriarco]

Yeah, I'm not sure how to word it clearly. We have a lot of correlational evidence regarding nutrients/food but not a lot of causal evidence and we may never have that. How do we best represent nutrition in this light considering that lack of causal evidence doesn't mean lack of causation?

[Dandelion]

Good questions and it strikes close to the core of the matter of how science works and epistemology especially in a field as young as nutrition. This is probably more a question for someone smart in these matters like Lentil to weigh in on. It's hurting my tiny brain, in a good way!

[Lentil]

Yeah, I'm not sure how to word it clearly. We have a lot of correlational evidence regarding nutrients/food but not a lot of causal evidence and we may never have that. How do we best represent nutrition in this light considering that lack of causal evidence doesn't mean lack of causation?

You're entirely correct that establishing causation is very difficult and can take a long time. Scientists consider new information with varying degrees of uncertainty -- if it hangs together with other well-established evidence, they may feel pretty confident that the new evidence will be reproducible (but would still wait for more data before accept it as fact). If it doesn't seem to be a logical extension of any existing knowledge or goes against well-established evidence, they will likely wait for more evidence before incorporating the new information into their thinking but will keep the new info in the back of their minds. I'm not sure if that makes sense or answers your question at all, but that's at least my interpretation of how scientists consider new and uncertain data.

As far as specific applications to nutrition, it isn't really appropriate to be making decisions based on the latest research. I think our current knowledge of nutrition is enough to make smart choices -- most of us have a hard enough time getting enough fruits and veggies and maintaining calorie balance without laboring over whether to eat a cup of kale or a cup of cucumbers. At this point, the best idea is to hedge your bets and eat a varied diet because we just don't know.

Logic does certainly play a role in bringing research from basic science to clinical practice -- I know MD's do this, and I would guess RD's do too. Although there may not be a clinical trial looking at the benefits of kale, we know the nutritional composition of kale, and can recommend eating it based on the fact that it is high fiber, low calorie, and high in various vitamins, minerals and antioxidants that we don't necessarily *know* are going to have an affect (there may be some limited data such as in vitro or animal studies), but are probably not going to be harmful. There is certainly a large degree of uncertainty in applying science this way, but with limited information it's the best we can do.