Posts

Comments

Comment by Matthew Green (matthew-green) on It’s Probably Not Lithium · 2022-07-12T12:39:38.110Z · LW · GW

Isn’t the “too palatable food” theory ridiculously easy to test, once you define what “too palatable” means? Assuming that we grant the “obesity epidemic is caused by changes in diet over the 20th century” you’d just need to switch people to an unrestricted-calorie diet that mirrors the homemade foods that our ancestors ate in the early 1900s and see if their satiety plummeted. (Here in the US that would still be a pretty diverse and filling diet that includes lots of meat, potatoes and pie.) I am skeptical that this would work (at the effect size needed to explain the obesity epidemic) but I’d love to see it tested.

Comment by Matthew Green (matthew-green) on It’s Probably Not Lithium · 2022-07-12T12:26:31.309Z · LW · GW

Is there any solid evidence that walking 25k steps per day will solve the obesity epidemic? I ask this because it’s genuinely a remarkable claim, one that if verified and implemented would save huge numbers of lives and hundreds of billions in medical expenses. The literature mostly seems to indicate that increased exercise doesn’t have dramatic effects on obesity.

Comment by Matthew Green (matthew-green) on It’s Probably Not Lithium · 2022-07-12T12:16:02.732Z · LW · GW

This sounds like a pretty intense restriction diet that also happens to be unpalatable. But the palatable foods hypothesis (as an explanation for the obesity epidemic) isn’t “our grandparents used to only eat beans and vegan sausages and now we eat a more palatable diet, hence obesity.” It’s something much more specific about the palatability of our modern 20th/21st century diet vs. the early 20th century diet, isn’t it? What’s the hypothesis we could test that would actually help us judge that claim without inadvertently removing most food groups and confounding everything?