Is observational science good for nutrition study?

SHBoss1673

I'd like to post a hypothesis, and let's discuss.

Firstly, please read the following description of observational science.
http://en.wikipedia.org/wiki/Observational_science

now, Think about what that means with regards to health topics and nutrition.

Do you think that observational science is a good method to predict things such as health risk factors, success rates for nutrition plans, and exercise philosophies? What areas of health and nutrition ARE good areas to use observational science in?

An example of observational science, would be the studies that are done on wine drinking. Since these are long term studies with humans, and you can't create a control group for this type of thing, so studies are done on people who drink lots of wine, no wine, and some wine. This is all well and good, the problem lies in that in order to be able to make any statements work, they would all need to have similar genetics, lived very similar lives, be of similar age...etc. The variables can't be controlled so conclusions cannot be considered very strong.

The health and wellness industry is rife with these kinds of inconclusive conclusions. It's my opinion that we would be better off without them in many instances as they can lead to a lot of these crazy diet plans that are all over the world.

What do you guys think? I think there's very little validity to this kind of science, and it just perpetuates myths and nutritional dogma that we could all afford to do without.

iplayoutside19

I get what you're saying about observation science. Each human being is vastly unique and a prodoct of their genetics, diet, culture, and attitude.

An example like the wine drinkers group is wrought other variables that it's not even worth looking at.

My biggest aggravation come from is the media's propensity to over state what the study said. Even if it were a well done study.
Upon learning that weight loss is mostly nutrition they flash huge headlines that say: "EXERCISE is USELESS, stop doing it"
That's not what that study said. But because it appeared in a headline people listen to it without any research on their own.

Or the biggest most misquoted study... the starvation mode study.

Paige1108

Okay, I read the description. I'm also half way through my other homework, the great studies you posted the other day. And it all leaves me with the same question....With human beings when does the study begin and end? By this I mean, the studies you shared were controlled studies, not observational, but I still wondered, using the the lipolytic response in obese women study, how similar or dissimilar were these women? What were their lives like and would those differences effect what was being studied. If one was wealthy and had access to better healthcare prior to the study could that effect her results? If one woman is raising a terminally ill child, I've read studies about the health effects of this so it comes to mind, her basic body chemistry would start out out of whack?

So I wonder, can you really set up "controled study" when you are studying people/humans or are we stuck with things always on the borders of "Observational Studies" no matter what we do anyway? I know, not an answer to the question you're posing, just another question. But an honest one. I'm truly not playing devils advocate here. I just wonder.

SHBoss1673

Iplay, completely agree, although I think that borders on being a separate topic, I.E. the voracity of the media, and the willingness for the public to accept the media as quasi-experts just because they report on something. It's made even worse these days with the prolific nature of blogs, which make every tom, ****, and harry an expert with a soap box.

Paige, I'm happy you bring that up. For I feel there is a point where we can take the concept to far. At some point the minutia can get in the way of "good enough"

There was an old saying in my engineering class, it went like this.

What's the difference between an chemical engineer and a chemist? If you tell a chemist and a chem E. to start out 20 feet from an attractive woman, and tell him he can 1/2 the distance once a minute for 6 minutes. The chemist won't bother because he can never reach his goal, the chem E. will do it because he may not reach her, but he'll come "close enough to do everything he wanted".
OK so that's a bit sexist, and a bit perverted, but it's from the late 70's (it was told to me by a professor who was pretty old, don't shoot the messenger). But it gets to the point. At some point, if the controls are done well, the variables can be minimized to a point where they no longer significantly affect the outcome. In true observational science, this cannot be achieved. And with certain fields, this can't be helped (such as the examples in the wiki), but with human anatomy, many things CAN be controlled (such as the blood work of a person who drinks wine, what happens to the hormones and other biological markers in a body), these are what we should focus on, not the long term bogus facts that coincidentally may occur between large groups of dissimilar people.

So all that said, yeah, you can pick apart almost any study, even one with impeccable control groups, but unless an experiment is performed in a vacuum, you will always have variables. The point is to make those variables small enough to be inconsequential, which observational science does not do, nor does it even try to do. If you can repeat the controlled experiment with a high degree of accuracy, you are then able to prove, or disprove a hypothesis, which is the ultimate goal I think. The more you can repeat research, the better that research is. Which is why many of those topics I posted are based on prior studies in which they use similar measures and controls, and test similar things, and maybe add a few extra tests to it to give more up to date data.

Paige1108

So all that said, yeah, you can pick apart almost any study, even one with impeccable control groups, but unless an experiment is performed in a vacuum, you will always have variables. The point is to make those variables small enough to be inconsequential, which observational science does not do, nor does it even try to do.

I like the analogue, got to the point perfectly.

With that as more of the guide in my mind, I think maybe there are two possible things in play in "observational science" that is problematic. Who is paying the bills and what do they want to hear? If an observational science study adheres to some riggers of science they may actually draw a some conditional conclusion. But if the real sciences at work is statistical data mining, you just can't trust the research. Like all the wonderful research that proves that High Fructose Corn Syrup, done by the High Fructose Corn Syrup industry, is actually great for you and will help you get a tan and lose 10lbs. Okay the last part is silly, but if you can't laugh while you discuss what's it all about. The problem is there's no way to tell the difference most of the time. I hope that doesn't sound to paranoid. But I think it makes a difference.

SHBoss1673

I like the analogue, got to the point perfectly.

With that as more of the guide in my mind, I think maybe there are two possible things in play in "observational science" that is problematic. Who is paying the bills and what do they want to hear? If an observational science study adheres to some riggers of science they may actually draw a some conditional conclusion. But if the real sciences at work is statistical data mining, you just can't trust the research. Like all the wonderful research that proves that High Fructose Corn Syrup, done by the High Fructose Corn Syrup industry, is actually great for you and will help you get a tan and lose 10lbs. Okay the last part is silly, but if you can't laugh while you discuss what's it all about. The problem is there's no way to tell the difference most of the time. I hope that doesn't sound to paranoid. But I think it makes a difference.

It's not paranoid, it's pretty much exactly the point I'm trying to make. With all the variables in observational science, you can choose any one of them and come to conclusions. Also with as large as those types of studies are, you can pull (clandestinely) subsets out that have similar characteristics, and still make assumptions that seem perfectly valid cast in a certain light, but when taken as a whole are just silly, of course they won't tell you the whole.

It's like taking a group of 5000 people that all ate 1 snickers bar a day for 10 years, and say 300 of them also ate very healthy and lived long lives, you could take that 300, and say "In a study, a group of 300 adults ate snickers bars for 10 years and all were extremely healthy...blah blah blah" The statement itself is not false, but it's not the whole truth either. Not saying that statistically, that 300 is less than 10% of the whole group, or giving any details on the health of the rest is wrong. It's bull crap science, and it makes me mad. It's done to promote something usually, or to prove a point, and it's trashy.

Paige1108

In other debates it's called junk science. I the weight loss world, it's the next best seller. I know it makes me so mad. Dunk your head in ice water and lose weight. What shock your system and your metabolism may go up also? That's amazing, I think I'll write a book. Too soon to bash this diet???

millerll

Great topic, and one most folks don't understand. As an engineer, I can tell you that there are lies, damn lies, and statistics. In other words, you can pretty much manipulate data to achieve any outcome you want. It takes pretty rigid peer review to weed out the wheat from the chaff. Which is why I don't trust most studies unless they include some explanation on methodology, and most don't. I'm a natural-born skeptic.

Keep 'em coming, Banks!

SHBoss1673

Great topic, and one most folks don't understand. As an engineer, I can tell you that there are lies, damn lies, and statistics. In other words, you can pretty much manipulate data to achieve any outcome you want. It takes pretty rigid peer review to weed out the wheat from the chaff. Which is why I don't trust most studies unless they include some explanation on methodology, and most don't. I'm a natural-born skeptic.

Keep 'em coming, Banks!

that's why I posted all that research the other day, it was all experimental. Sure the groups were small, but they were tightly controlled, and the methodology on all of them was superb. Well, except for the one that was a summary of a bunch of other studies.

TrainingWithTonya

Agreed. I hate observational studies. But unfortunately, we're not allowed to treat people like lab rats so we can't get actual human data in most cases. So, we use what we can and, hopefully, we weed out the junk. Unfortunately, the media seem to lack the filter of common sense or any actual nutritional or physiological knowledge. My opinion is that there should be some sort of regulation for reporting and writing books and even personal training and any other form of exercise and nutrition information so that the junk science isn't propagated.

SHBoss1673

I don't know, a lot of the experimental studies that I read have very precise controls. I mean I get the meaning, I.E. we can't keep some test humans locked up from birth to make sure they all have the same composition. But that's what I meant about minimizing variables. There's always going to be some kind of variability involved. But If you can control the beginning, the ending, and what the experimental patient experiences during the experiment, then you are minimizing variability. Some of those studies basically just poll people once, come back in 6 months, poll them again, take a blood sample and they're done. There's just no credibility to that kind of research, 1/2 the people could be lying.