...Are rare to nonexistant, are they not? Well, I've got an epic idea that is full of holes, and I want some feedback, or I want someone to steal it.
I'm in school for nutrition, and work full time at a clinical trial review center. All of which has led me to believe that nutrition is much less amenable to randomized trials, cohort studies, or case control studies, compared to medical interventions. There's a few papers out on this, but the reasons range from nutrient interactions to difficulty in maintaining diets, up to long timeframes necessary for impacts.
Here's my idea: a website which doles out food/nutrient modifications to willing paleo guinea pigs, who record set qualitative observations (weight gain, bowel movements etc) or quantitative observations (LDL, etc, if available intermittently from a physician). Randomization and adherence would be difficult to impossible, but something is better than nothing! Then you could say something like "There's no peer reviewed studies on coconut oil vs butter for depression, but that weirdo website has deidentified info on ten people who made that comparison..."
Poke this idea full of holes, please. I likely don't have the time to do this, and the logistics are a nightmare, but it does intrigue me.
I heard of something like this on one of Rob Wolf's podcasts. It would be totally awesome, but you would have to control for environments which would be impossible unless you kept people in a hospital like setting for the duration. You could manage adherence this way, but people would have to abandon their lives for awhile. My work stress is different than yours and may cause my body to react differently than yours. So many things are factors that could be confounding variables. It would cost a lot of money to have a decent sample size because people would have to commit to it for a long period of time and all the medical tests you would want to run would get expensive. Great idea though! Maybe someday.
This would be great, but unfortunately no matter how big the study, unless it was a proper controlled clinical study, it would be difficult for it to gain much scientific merit.
I'd love to see studies done in prisons/ jails - a completely controlled enviroment with large populations to compare and study.