Meta-note: Judging by the feedback I got, the last article in this series was probably a little out of my depth in terms of what I’m able to explain in a clear and at least somewhat entertaining way.
Thank you. I really appreciate the detailed analysis that you present. And I fret a lot about bias in science, so it is nice to have someone to articulate the various ways that it skews results.
But of course, I can’t comment without disagreeing with something. : )
I think your seat belt analogy missed the mark. The point about not finishing a diet is that it is hard. If a large number of people can’t finish something and if trying the activity half way and quitting is actually detrimental to your health, then I think it is reasonable to suggest to people that they shouldn’t try. If you have a 10% chance of succeeding with a +5 reward and a 90% chance of not succeeding with a -1 penalty, then from a public health perspective, it is something that should be discouraged. Of course, changing those numbers a little and having proper representative populations for diet studies could quickly change that conclusion. I offer them only as an illustration.
To strain the seatbelt analogy, it would be like a seatbelt that you had to hold on to for it to work, but if you let go, you would be more damaged than if you just let your body hang loose. Of course, seatbelts are “effective” if you use them correctly, but if most people aren’t capable of using them correctly, are they really a tool that should be promoted?
I don't think I necessarily disagree with you in terms of the scenario you present - if diets are very unsuccessful and even kind of harmful, then there's a good argument for not doing them. As I indicated a little above, I think diets are probably significantly more successful than the scientific literature thinks they are, but for reasons also listed above that's hard to prove.
To your other point, I read a little of the evidence diets are considered harmful and didn't find it very compelling, but I didn't really have the room to talk about it here. If there was compelling evidence that I didn't see that showed them to be worse than I think they are, I'd probably have to reset my view on this.
One other thing here I'd like to address is the idea of "finishing" a diet - I don't think this is necessarily the way anybody looks at it in real life. If you went up to a person and asked "So say you eat less, and lose some weight; does that weight then stay off forever even if you returned to eating as much as you ate to hit your previous peak weight?", I don't think the average person would say yes. There's some baseline understanding that you'd have to maintain a dietary modification of some kind to maintain the weight loss.
The review authors don't look at it this way. For them, if you go on a diet and lose weight and then gain the weight back for any reason at all, it's a failure of the diet, full stop, behaviors of the person factored out entirely. This would be fine if they made effort for people to understand that to them this meant "diets don't compel people to stay on them forever" as opposed to "there's no changes to your diet you can make that would cause a reduction in weight sustained as long as the changes were maintained". They don't; that wouldn't make much of a splash since everybody already knows about the "My diet was working fine until I quit doing it" failure mode.
Mostly agree. Except to push your last point a bit: How many people have to fail to continue a diet for it to be the diet’s fault rather than the individuals’ fault? Regardless of how good the results of the diet are, if no one can stay on it, doesn’t that make it a crappy way to lose weight that isn’t worth trying?
Of course, assuming no to little downside, as you mention above, it could still be worth trying. We would also have to assume there is no opportunity cost of better options, which for this argument on the review papers you mention, seems like there are no better options since they are lumping all weight loss programs together. Which, now that I think about it, does seem a bit silly.
Thanks. This thought process is a nice break from current events. Now on to think about whether a company can have values that conflict with the profit-motive...
I think you are probably right that some diets aren't followable - "starve yourself completely until you hit goal weight" would seem to be the extreme example. And I wouldn't be surprised to find a lot of "reasonable" diets have an extreme minority who are successful with them - I expect most people to quit most diets.
The distinction here for me is why they quit them - was the diet impossible to stay on, or did a person whose personal history indicates a lack of self denial revert to mean? Even if we buy all the metabolic slow-down and leptin stuff, it's still basically known that if a morbidly obese person ate, say, 2500 calories a day, they might not be as skinny as a person who had always maintained a diet of 2500 day but would at least stop being morbidly obese, perhaps settling out at something like "fairly overweight".
I actually don't have problems with "most people fail diets - do stuff to make sure you are that 5%" type messaging, and I don't judge people who are overweight (I understand fully that food is good and exercise is hard - I can tell when I look in the mirror). I mostly just want to make sure we don't revert to "don't even try" messaging, which is what most people probably get out of "diets don't work" phrasing.
Thank you. I really appreciate the detailed analysis that you present. And I fret a lot about bias in science, so it is nice to have someone to articulate the various ways that it skews results.
But of course, I can’t comment without disagreeing with something. : )
I think your seat belt analogy missed the mark. The point about not finishing a diet is that it is hard. If a large number of people can’t finish something and if trying the activity half way and quitting is actually detrimental to your health, then I think it is reasonable to suggest to people that they shouldn’t try. If you have a 10% chance of succeeding with a +5 reward and a 90% chance of not succeeding with a -1 penalty, then from a public health perspective, it is something that should be discouraged. Of course, changing those numbers a little and having proper representative populations for diet studies could quickly change that conclusion. I offer them only as an illustration.
To strain the seatbelt analogy, it would be like a seatbelt that you had to hold on to for it to work, but if you let go, you would be more damaged than if you just let your body hang loose. Of course, seatbelts are “effective” if you use them correctly, but if most people aren’t capable of using them correctly, are they really a tool that should be promoted?
I don't think I necessarily disagree with you in terms of the scenario you present - if diets are very unsuccessful and even kind of harmful, then there's a good argument for not doing them. As I indicated a little above, I think diets are probably significantly more successful than the scientific literature thinks they are, but for reasons also listed above that's hard to prove.
To your other point, I read a little of the evidence diets are considered harmful and didn't find it very compelling, but I didn't really have the room to talk about it here. If there was compelling evidence that I didn't see that showed them to be worse than I think they are, I'd probably have to reset my view on this.
One other thing here I'd like to address is the idea of "finishing" a diet - I don't think this is necessarily the way anybody looks at it in real life. If you went up to a person and asked "So say you eat less, and lose some weight; does that weight then stay off forever even if you returned to eating as much as you ate to hit your previous peak weight?", I don't think the average person would say yes. There's some baseline understanding that you'd have to maintain a dietary modification of some kind to maintain the weight loss.
The review authors don't look at it this way. For them, if you go on a diet and lose weight and then gain the weight back for any reason at all, it's a failure of the diet, full stop, behaviors of the person factored out entirely. This would be fine if they made effort for people to understand that to them this meant "diets don't compel people to stay on them forever" as opposed to "there's no changes to your diet you can make that would cause a reduction in weight sustained as long as the changes were maintained". They don't; that wouldn't make much of a splash since everybody already knows about the "My diet was working fine until I quit doing it" failure mode.
Mostly agree. Except to push your last point a bit: How many people have to fail to continue a diet for it to be the diet’s fault rather than the individuals’ fault? Regardless of how good the results of the diet are, if no one can stay on it, doesn’t that make it a crappy way to lose weight that isn’t worth trying?
Of course, assuming no to little downside, as you mention above, it could still be worth trying. We would also have to assume there is no opportunity cost of better options, which for this argument on the review papers you mention, seems like there are no better options since they are lumping all weight loss programs together. Which, now that I think about it, does seem a bit silly.
Thanks. This thought process is a nice break from current events. Now on to think about whether a company can have values that conflict with the profit-motive...
I think you are probably right that some diets aren't followable - "starve yourself completely until you hit goal weight" would seem to be the extreme example. And I wouldn't be surprised to find a lot of "reasonable" diets have an extreme minority who are successful with them - I expect most people to quit most diets.
The distinction here for me is why they quit them - was the diet impossible to stay on, or did a person whose personal history indicates a lack of self denial revert to mean? Even if we buy all the metabolic slow-down and leptin stuff, it's still basically known that if a morbidly obese person ate, say, 2500 calories a day, they might not be as skinny as a person who had always maintained a diet of 2500 day but would at least stop being morbidly obese, perhaps settling out at something like "fairly overweight".
I actually don't have problems with "most people fail diets - do stuff to make sure you are that 5%" type messaging, and I don't judge people who are overweight (I understand fully that food is good and exercise is hard - I can tell when I look in the mirror). I mostly just want to make sure we don't revert to "don't even try" messaging, which is what most people probably get out of "diets don't work" phrasing.
Request for comment: SlimeMoldTimeMold and their explanation on the obesity epidemic?