Hacker News Reader

Comment by waterlooalex | Hacker News Reader

It doesn't look like they had a control group, did they?

Nope, and no blinding either. This should be considered a preliminary study at best. The quantity of participants doesn't make much of a difference without proper scientific controls.

I'm not trying to quibble, but am genuinely curious: What would a control group consist of? There were plenty of patients who didn't receive the trial program, at large in the general population. And it couldn't have been a blind study, because the patients knew what they were eating.

A control group goes through the same amount of attention and procedure as the experimental group, controlling for as many independent variables as possible. "Someone in a lab coat is watching my diet" all by itself is probably a pretty strong motivator for change, and one that the general public does not have. What people eat when they think someone's watching can be markedly different to what they'd eat in private.

> What would a control group consist of?

It would be an equivalent population who weren't asked to change their diets.

There are multiple reasons the patients in this study might not be comparable to participants in other studies or to the general population. Age, gender, average disease severity all need to be controlled for. And if nothing else the simple fact the patient knows they're being monitored could have an impact on all sorts of behaviors.

Yeah, these are interesting results, but without a control group you can't say for certain what is causing an effect or, actually, if an effect exists.

Is that not standard advice when you have a heart attack. Give up smoking, eat healthier, do a bit of exercise. It's going to be difficult to find a willing and valid control group under the circumstances.

> It would be an equivalent population who weren't asked to change their diets.

Which is actually an ethical quandary when you think about it. A doctor's ethics are really tested when he has to say to some patients "sure, I think you should change your diet and you'll probably die if you don't, but you're part of the control group!"

One more ethical (and still statistically somewhat interesting) way of administering such a study would be to bring in doctors advocating different sorts of diets. The Cleveland Clinic hasn't done such a thing but they probably could: they've got a group that advocates the rigid vegan diet and another that advocates the Mediterranean diet. Not as good as having a control group that continues to subsist on cheeseburgers and coke but it actually can test a few specific ideas, like the vegan claim.

> Which is actually an ethical quandary when you think about it.

You're right, of course. But whether a study is scientifically valid and whether it is ethical are two orthogonal questions. Similar concerns are why most cancer therapy trials end up looking more like A/B testing than classic experimental/control experiments. The point I was trying to make, though, is that you need at least two groups which differ only by a small set of (ideally one) known variables to draw any real conclusions

Not exactly a control group, but among the group of 22 patients who did not follow the diet, 13 or 62% had some sort of adverse event (stroke/heart-attack etc), compared with just 1 out of the group of 177 (0.6%) who did follow the diet. That seems to be pretty compelling evidence that the approach is effective.

It's not. For all we know, those that were unable to follow the diet were those who had the worst cases and/or least healthy diets.

One immediate issue is that conscientiousness, e.g. the ability to maintain a difficult diet for an extensive period of time, is quite correlated with longevity.

This is a very commonly known effect, which makes me more than a bit suspicious of this particular study.

This is also why a control group with random assignment is necessary, particularly in the social sciences. There are simply too many confounding factors to make epidemiology useful in basically any way.

Reply on news.ycombinator.com