3 ms·

The endlessly repeated truism that "these studies don't prove causation" isn't really saying anything, as the same could be said for every long term human drug

The endlessly repeated truism that "these studies don't prove causation" isn't really saying anything, as the same could be said for every long term human drug study. This line of thinking was the reasoning behind ignoring the link between cancer and tobacco products for decades, because the studies establishing that link were and are observational studies, considering that a controlled experimental study on humans establishing the same would be completely unethical.

jdietrich 2y ago
>This line of thinking was the reasoning behind ignoring the link between cancer and tobacco products for decades I wish people would stop repeating this falsehood. The mass consumption of tobacco only really started after 1910, with the mechanisation of cigarette manufacturing. There was a multi-decade time lag between this explosion in tobacco use and the long-term sequelae of lung cancer and cardiovascular disease. The causal link was established beyond any reasonable doubt in the late 1950s, which was really the earliest point at which we had clear data; it took only a few years for all major medical bodies to endorse this as the consensus position and start lobbying for legislative action. "We didn't know" is a propaganda line rolled out by the tobacco industry, the asbestos industry, the fossil fuel industry and others. The lack of action wasn't due to any scientific uncertainty, but naked corruption at the highest level. There was an overwhelmingly clear scientific consensus which was privately accepted as true by these industries, but they paid off politicians to protect their profits. We genuinely don't know if this is a real risk; NAION is a rare disease, which greatly increases the difficulty of making accurate inferences even from very large amounts of data. We can be extremely confident that the magnitude of risk is vastly smaller than the risks caused by Type 2 diabetes and obesity.
- TeaBrain 2y ago
  It wasn't a perfect example, which I borrowed from my recollection of The Book of Why (Pearl, 2018), but your response is besides the point that saying "these studies don't prove causation" is saying nothing. It is the most tiring over stated pointless truism, which is so often said as a response to long term observational drug studies, despite virtually none claim to be establishing causation. This phrase "correlation is not causation" is repeated ad nauseam in response to studies simply drawing correlations, mostly in order to write them off or ignore them. It has gotten to the point where I now expect at least one person to say it as a response to every single study I see posted on the internet where a comment section exists, as if there is some internet law that obligates people to state this truism if it hasn't been stated yet.

jdietrich 2y ago

>This line of thinking was the reasoning behind ignoring the link between cancer and tobacco products for decades I wish people would stop repeating this falsehood. The mass consumption of tobacco only really started after 1910, with the mechanisation of cigarette manufacturing. There was a multi-decade time lag between this explosion in tobacco use and the long-term sequelae of lung cancer and cardiovascular disease. The causal link was established beyond any reasonable doubt in the late 1950s, which was really the earliest point at which we had clear data; it took only a few years for all major medical bodies to endorse this as the consensus position and start lobbying for legislative action. "We didn't know" is a propaganda line rolled out by the tobacco industry, the asbestos industry, the fossil fuel industry and others. The lack of action wasn't due to any scientific uncertainty, but naked corruption at the highest level. There was an overwhelmingly clear scientific consensus which was privately accepted as true by these industries, but they paid off politicians to protect their profits. We genuinely don't know if this is a real risk; NAION is a rare disease, which greatly increases the difficulty of making accurate inferences even from very large amounts of data. We can be extremely confident that the magnitude of risk is vastly smaller than the risks caused by Type 2 diabetes and obesity.

TeaBrain 2y ago
It wasn't a perfect example, which I borrowed from my recollection of The Book of Why (Pearl, 2018), but your response is besides the point that saying "these studies don't prove causation" is saying nothing. It is the most tiring over stated pointless truism, which is so often said as a response to long term observational drug studies, despite virtually none claim to be establishing causation. This phrase "correlation is not causation" is repeated ad nauseam in response to studies simply drawing correlations, mostly in order to write them off or ignore them. It has gotten to the point where I now expect at least one person to say it as a response to every single study I see posted on the internet where a comment section exists, as if there is some internet law that obligates people to state this truism if it hasn't been stated yet.