An avid reader of this site (yes, there is apparently such a thing…) writes:
evidence-based medicine… was wondering your opinion on this guy’s theory,
methodology, and empirical support? it’s nothing if not
it, and i hate to say I think it works.
My Response:Your lead-in to this was "since you’re into evidence-based medicine…" – so I thought I’d view this from that point of view. I have lots of other points of view after looking at this web site, but I’ll try to stick to the question.
I’m aware of many who think there is an obesity "set point" and there
are lots of different people working on this type of thing – also
possibly debunking the idea of calories all being the same.
One of the
more convincing speakers I’ve heard recently talks about obesity as a
vestige of an evolutionary advantage – the ability to hold onto
calories and store energy may have
conferred evolutionary advantage while there were scarce resources, but
now things are different…this makes a lot of sense to me…probably because it fits in with my preconceived world view about how all this medicine and genetic stuff works anyhow…but it doesn’t really help anyone lose weight.
Anyway, I digress. From my admittedly brief perusal of his web site,
this theory and his evidence for it gets a Level 5. Now, that in and of
itself should not direct anyone to do anything…the context of this "level of evidence" is
much more important.
One of the great recent stories in medicine that reminds me of this issue has
an Australian scientist who was convinced of the bacterial origin of
peptic ulcers swigging a cocktail of Helicobacter pylori bacteria and developing
some nasty gastritis from it indeed in partial fulfillment of Koch’s posulates for
determining if an infectious agent actually causes a disease. This is the "self-experimenation" that Roberts touts as part of the "science" behind his "theory."
There’s other theory and some experimental research behind the notion of an obesity/calorie set point that Roberts quotes, but I’d have to see the book. I have absolutely no inclination right now to pore over his 73 page PDF that outlines his theory. I also happen to worry about a theory that takes 73 pages to explain, but I promised to stick with the EBM stuff.
So his theory is interesting, and may well have elements of truth. However, all a theory does is point people in the direction to start performing the necessary research to back it up…to gather more and more evidence that is consistent with the theory so that that theory is the best explanation we have for a certain phenomenon. He has lots of theories – one of the PDF’s on his site is about three things statistics books don’t tell you that can lead to new ideas. He has another whole paper on the value of self-experimentation. He seems wonderfully prolific, creative and certainly energetic and passionate – all extremely important qualities in a scientist and researcher.
But the empirical evidence that this is a better way to lose weight than all the others simply is not there. How would you get that evidence? You would eliminate all other sources of confounding and test the intervention against other interventions…i.e. the randomized controlled trial. There’s absolutely no reason to NOT do such a thing. In fact, if his intervention is easy enough to be written in a book for lay consumption (no pun intended…), then it is amenable to an RCT. There are lots of people out there who will say that not everything must be subjected to an RCT in order to say it works. I even think that’s true in certain rare circumstances – but it should NOT be used to avoid the cost and difficulty associated with properly testing an intervention before foisting it upon a desperate public as the "best" intervention without actually PROVING it.
Yes, RCTs have been proven wrong…yes, there’s something to human intuition and the doctor-patient relationship that cannot be explained by a rational, scientific approach, but let’s not throw out the baby with the bath water. RCTs and empirical evidence have brought us wonderful medical innovations and powerful new tools to fight disease. If they have gone wrong, it’s usually because we don’t play by the rules in interpreting the evidence. We adopt interventions on a wide scale BEFORE they’re proven (class Ia antiarryhthmics, HRT, etc.), or we believe evidence that has been produced and too-heavily edited by those with an interest in the financial success of the intervention (cox-2-inhibitors, e.g.).
Will the "Shangri-La Diet" work for obese people in general? I don’t know…it might, but the owner of "SethRoberts.net" certainly hasn’t proven it yet.