3 ms·

Maybe I am not understanding it correctly, but it seems to me that an initial analysis does show a correlation between dementia and consumption of diet drinks,

Maybe I am not understanding it correctly, but it seems to me that an initial analysis does show a correlation between dementia and consumption of diet drinks, but once you adjust for factors that could cause dementia (such as diabetes) that correlation disappears. But doesn't that imply a correlation. Eteeen diet drinks and diabetes? So maybe diet drinks don't cause dementia, but they do lead to diabetes, which then increases the odds of dementia. How do you conclude "drink away" from that? That would also fit other studies that have shown that drinking diet soda instead of sugar soda leads to no weight loss, likely because people compensate by getting their sugars elsewhere. Which would match well with the idea that diet soda consumption correlates with diabetes.

szemet 10y ago
A causal relation definitely exists in the other way around: people with diabetes are encouraged to consume less sugar... This may explain large part of the correlation.
URSpider94 9y ago
That's an interesting idea, but there is no proof for this in the data, and in fact that's a common logical fallacy. Let's assume a population that lives on an island that sits on the equator. One side of the island is polluted with radiation. It also features a kind of purple fruit that doesn't grow anywhere else and is not exported. Now let's assume that the fruit contains a cumulative toxin that causes people to die very young of brain cancer. A study without adequate controls might conclude: -- Living near the equator causes brain cancer -- living on this island causes brain cancer -- radiation causes brain cancer -- eating purple fruit causes brain cancer -- radiation causes people to eat more purple fruit Only one of these is true, but one could certainly invent reasons to explain nearly all of these correlations as causal. In the case of this study, an alternate explanation that makes sense is: -- diabetes causes dementia -- diabetics are strongly encouraged to drink sugar-free soda to avoid sugar intake -- therefore, diet soda intake and dementia are linked only through their common root cause, diabetes

szemet 10y ago

A causal relation definitely exists in the other way around: people with diabetes are encouraged to consume less sugar... This may explain large part of the correlation.

URSpider94 9y ago

That's an interesting idea, but there is no proof for this in the data, and in fact that's a common logical fallacy. Let's assume a population that lives on an island that sits on the equator. One side of the island is polluted with radiation. It also features a kind of purple fruit that doesn't grow anywhere else and is not exported. Now let's assume that the fruit contains a cumulative toxin that causes people to die very young of brain cancer. A study without adequate controls might conclude: -- Living near the equator causes brain cancer -- living on this island causes brain cancer -- radiation causes brain cancer -- eating purple fruit causes brain cancer -- radiation causes people to eat more purple fruit Only one of these is true, but one could certainly invent reasons to explain nearly all of these correlations as causal. In the case of this study, an alternate explanation that makes sense is: -- diabetes causes dementia -- diabetics are strongly encouraged to drink sugar-free soda to avoid sugar intake -- therefore, diet soda intake and dementia are linked only through their common root cause, diabetes