Show Me The Science is Near Impossible

#1

Long term studies on diet have so many confounding and extraneous variables, it’s hard to prove nearly anything other than short term results. Not knocking keto, but I’m not sure anything is proven yet when it comes to diets and nutrition. I think it is one of the least understood parts of health still.

(bulkbiker) #2

Until we can lock people in rooms for 20 or 30 years and only feed them specific things (and lets face it that’s never going to happen) then there will be no long term diet studies so we have to go with what we hear reported anecdotally by those we trust.
The more I hear about keto the better the results seem to be and there are no real financial incentives for people to lie about it so…

(Central Florida Bob ) #3

Recently, some big shot whose name I should remember, but don’t, was saying that yeah, epidemiological studies are crap opposed to randomized controlled trails, but we need to use them and do more of them. The simple fact is that we can’t afford (to butcher what @MarkGossage says) to lock people in rooms for 20 or 30 weeks, let alone years. Metabolic ward studies are horribly expensive, and even then you have to wonder how biased they are. After all, are you going to volunteer to be locked away from your family for a long period of time? Even if they pay you? It used to be if you wanted to do a good short term test, the cheapest volunteers were college students. As a result lots of data is based on them.

This is where coming up with ways to experiment on yourself or smallish groups like this one come into play. There’s a lot of self experimentation around here and a wealth of knowledge to tap.

Plus there are lots of supervised medical studies in the work of doctors Sarah Hallberg, Eric Westman, Jason Fung, Jeff Gerber, and many, many more. That’s certainly good data.

(Doug) #4

That keto goes a long way to fixing prevalent metabolic problems, and often does it fairly fast (in the context of entire lifetimes), reversing damage that builds up over much longer periods - quite a few decades in some cases - speaks for itself.

(Brian) #5

I’m not sure that a good study would have to lock people in a room for 20 or 30 years. I would think there would be a fairly decent group of people who would be willing to document their food intake, quite accurately, and from a rather wide spectrum of eaters. It wouldn’t surprise me to find people who would even do it without being paid as long as it wasn’t too cumbersome and didn’t cost them money to comply. Even a small incentive might help with compliance if it were possible, but even with none, I still think there would be participants.

This is a pretty modern era we’re living in. Most people have cell phones with cameras. Dave Feldman is one who has mentioned photographing every morsel that goes into his mouth, not sure whether that’s always, or just for data purposes, but it doesn’t cost anything but a little time. Yes, someone would have to sort and compile the data. And it could be rather cumbersome. But I think there could be some interesting and very legitimate studies possible today that maybe wouldn’t have been practical a decade ago.

(KetoQ) #6

Why does a long term study have to be the end all be all? And we know studies and their statistics can be gamed. Why do we give so much power to studies that are often more political than objective?

If keto works short term to reduce weight, improve lipid profile, improve insulin sensitivity (which long term studies show all those things are life extending) and live more fulfilling and active lives – why do we need a long term study to tell us what we already know?

If a vegetarian or Ornish or Mediterranean diet helped you get similar results, the “experts” would be saying you’ve extended your life and improved your health. Well, it happens with keto, too.

I had a heart attack 10 years ago. My health seemed awful over the past year until I started keto. I feel way better than I did 100 days ago. I don’t need a university professor to tell me otherwise.

I exercise vigorously everyday. If saturated fat was going to clog my arteries and strike me dead, it would have happened by now. No chest pain, no shortness of breath. I work out harder than most of the people at my gym 20-30 years younger than me.

I believe we’re trying to make health more complicated than it has to be.

(bulkbiker) #7

You’re possibly correct but the naysayers (and lets face it they are the people we would have to convince) will always claim that people lie about what they eat. This was clearly seen from the recent Harvard study where the average calorie intake was in line with the Minnesota starvation experiment so we know that was a long way from the truth. The only way an RCT would be completely acceptable would be by my method but… it won’t happen. Virta is the closest because they are testing for ketones so can tell immediately if people have cheated if they go out of ketosis but even the data from that is rejected by the doubters… I don’t have a solution I’m afraid.

(bulkbiker) #8

We don’t but if we want this to become more mainstream then unfortunately we need to convince the people who are set against it… Preaching to the converted is simple converting the unconverted (and if global health is our ultimate goal then that is what we are aiming for) is far more difficult.

(Katie the Quiche Scoffing Stick Ninja ) #9

We could do experiments on prisioners with life sentences! :stuck_out_tongue:

#10

LOL. And they would probably live longer.

(Crow T. Robot) #11

Or is it that people always lie about what they eat?

I’m inclined to believe the latter. It’s not necessarily willful, but studies on this have shown that people consciously or otherwise always seem to fudge their recollections to match whatever they think makes them look better, either to the researchers or to their own bias.

(bulkbiker) #12

very true…although looking at the food questionnaire I’m hardly surprised people mis remember… I forgot a whole pack of cooked meats that I snaffled down yesterday about 30 mins after I’d eaten it!

(Randy) #13

I think the Virta study is the best (current) way to test real low carb/high fat diets. (The recent study that claimed low carb was bad was 37% carb lol). Measuring BHB in the blood is proof of nutritional ketosis. And the results can be measured.

The fact that a place like Harvard is holding firm on the need for exogenous glucose/fructose baffles me.

But, there is money to be made huh?

(bulkbiker) #14

Maybe a slightly skewed demographic… ?

(KetoQ) #15

Mark –

I’m in total agreement and I was going to add that point, but wanted to keep my thoughts more focused. I work in brand and communication strategy and certainly understand the incredible value in being able to have a succinct and focused statement to refute some of the other “science” out there.

Just beating people over the head with “insulin sensitivity” can be ineffective, because the unwashed do not have an idea what that means or how to place it in the proper context of health and disease prevention.

(Bunny) #16

It’s all about the Leptin my dear Grinch …lol

(You've tried everything else; why not try bacon?) #17

It depends on what you mean by “long-term.” A two-year prospective study might yield very useful data, and it would be tolerably long-term, where nutrition research is concerned. If there were a way to do a randomized, controled study with accurate food-intake data, that would be ideal, and as already discussed, the advances in technology in recent years ought to simplify the data collection considerably.

Also, while epidemiological data are not usually usable in terms of establishing cause and effect, in some cases—smoking is the obvious example—the effect is large enough to constitute proof of causality, not merely correlation. Sir August Bradford Hill laid out criteria in a lecture in 1965 for assessing causal links from epidemiological data. It is on the basis of these criteria that we trust the notion that smoking causes lung cancer. Dr. Robert Lustig is convinced that his data linking sucrose in the food supply to the diseases that fall under the rubric of metabolic disease meet the Bradfor Hill criteria for establishing causation.

One thing epidemiological data can provide is proof of non-causality. For example, the fact that half of all people with familial hypercholesterolemia never develop cardiovascular disease shows that it is not cholesteral that causes the CVD in the people with FH who develop CVD. Likewise, the fact that 90% of patients with heart attacks are observed to have low cholesterol levels is another proof that cholesterol is not the cause of CVD. And so on.