4 ms·

Here’s one of the fundamental problems with studying human diet and psychology, from the article: > The best way to assess the effects of alcohol would be to r

Here’s one of the fundamental problems with studying human diet and psychology, from the article: > The best way to assess the effects of alcohol would be to randomly assign people to drink it or not in childhood and then monitor their health and drinking over the rest of their lives. Since such studies cannot be done, researchers instead have to ask people about their drinking habits and follow them over much shorter periods of time. Studying humans properly is nigh-on impossible.

blackbrokkoli 2y ago
The other being the myriad of influencing factors you can never really isolate. Say you do the setup from the quote, then one group would also be in a special situation where researchers told them to do this thing, and they would tell their friends, or maybe they would keep it secret, or maybe it would make them a bit weird, or maybe it would pave the way for more addictions, or make them more conscious of it, and they would have another item on their usual shopping list, and they may meet their spouse in a bar, and so on — all of which influences outcome.
- FredPret 2y ago
  That’s right. Studies will say that they “control for age and income” but even doing just that is probably not possible, unless a) you break a number of laws and ethical rules and b) you have a population of hundreds of billions of humans to study. Humans are very different from one another in important but subtle ways.
silverquiet 2y ago
Not to mention that humans are very different including their metabolic physiology. It's very hard to predict what level of drinking will cause systemic issues for any given individual. That said, I'm inclined to believe the studies saying that there is no health benefit to any level of drinking.
4gotunameagain 2y ago
It is not impossible, it is abhorrent and unethical. See Nazi twin studies, Japanese Unit 731, and who knows what is going on right now which will take years to be uncovered..
goda90 2y ago
I hope someday we could develop computers that can do a molecularly accurate simulation of the human body. Then we could try all the inputs we can imagine, with a bunch of different sets of genes, and figure out exactly what's good or not.
- FredPret 2y ago
  True - but then the simulation will have human rights
  - psd1 2y ago
    Not a problem, just gerrymander the computers running the simulations.
- coldcode 2y ago
  People are working on it, but its a long way from being complete. See, for example, https://www.quantamagazine.org/with-digital-twins-the-doctor-will-see-you-now-20240726/ https://www.quantamagazine.org/with-digital-twins-the-doctor... on modeling the circulatory system.
n4r9 2y ago
It will get much easier in the future when the likes of Meta collect and publish personal data down to the level of drinks choice. Researchers can then simply ask their institution to apply for a Meta-science subscription, pay for access to Meta-API to extract the data, hire a Meta-integration consultant to get it into a usable format, upgrade to Meta-science-plus in order to cross-reference Meta-pref with Meta-health, scramble to perform data analysis with the remaining two weeks active period, discover that there's a fatal bug in Meta-API because it was rushed to release by an ambitious product manager, wait two years for a bug-fix, and start over.
dynm 2y ago
Frustratingly, we came very close to getting a study that should have come pretty close to giving a conclusive answer to this. It was funded and approved and started. But then it was cancelled after people found out that the lead researcher lied about his ties to the alcohol industry: https://dynomight.net/alcohol-trial/ https://dynomight.net/alcohol-trial/
dpkirchner 2y ago
> Studying humans properly is nigh-on impossible. This is why I find the proclamation that "correlation != causation" so annoying. Of course it doesn't, but we cannot make a causative study of humans!

blackbrokkoli 2y ago

The other being the myriad of influencing factors you can never really isolate. Say you do the setup from the quote, then one group would also be in a special situation where researchers told them to do this thing, and they would tell their friends, or maybe they would keep it secret, or maybe it would make them a bit weird, or maybe it would pave the way for more addictions, or make them more conscious of it, and they would have another item on their usual shopping list, and they may meet their spouse in a bar, and so on — all of which influences outcome.

FredPret 2y ago
That’s right. Studies will say that they “control for age and income” but even doing just that is probably not possible, unless a) you break a number of laws and ethical rules and b) you have a population of hundreds of billions of humans to study. Humans are very different from one another in important but subtle ways.

silverquiet 2y ago

Not to mention that humans are very different including their metabolic physiology. It's very hard to predict what level of drinking will cause systemic issues for any given individual. That said, I'm inclined to believe the studies saying that there is no health benefit to any level of drinking.

4gotunameagain 2y ago

It is not impossible, it is abhorrent and unethical. See Nazi twin studies, Japanese Unit 731, and who knows what is going on right now which will take years to be uncovered..

goda90 2y ago

I hope someday we could develop computers that can do a molecularly accurate simulation of the human body. Then we could try all the inputs we can imagine, with a bunch of different sets of genes, and figure out exactly what's good or not.

FredPret 2y ago
True - but then the simulation will have human rights
- psd1 2y ago
  Not a problem, just gerrymander the computers running the simulations.
coldcode 2y ago
People are working on it, but its a long way from being complete. See, for example, https://www.quantamagazine.org/with-digital-twins-the-doctor-will-see-you-now-20240726/ https://www.quantamagazine.org/with-digital-twins-the-doctor... on modeling the circulatory system.

n4r9 2y ago

It will get much easier in the future when the likes of Meta collect and publish personal data down to the level of drinks choice. Researchers can then simply ask their institution to apply for a Meta-science subscription, pay for access to Meta-API to extract the data, hire a Meta-integration consultant to get it into a usable format, upgrade to Meta-science-plus in order to cross-reference Meta-pref with Meta-health, scramble to perform data analysis with the remaining two weeks active period, discover that there's a fatal bug in Meta-API because it was rushed to release by an ambitious product manager, wait two years for a bug-fix, and start over.

dynm 2y ago

Frustratingly, we came very close to getting a study that should have come pretty close to giving a conclusive answer to this. It was funded and approved and started. But then it was cancelled after people found out that the lead researcher lied about his ties to the alcohol industry: https://dynomight.net/alcohol-trial/ https://dynomight.net/alcohol-trial/

dpkirchner 2y ago

> Studying humans properly is nigh-on impossible. This is why I find the proclamation that "correlation != causation" so annoying. Of course it doesn't, but we cannot make a causative study of humans!