4 ms·

10. Trouble distinguishing correlation from causation.

dsego 9y ago
This! Hearing loss could mean those nerves started dying out before other (cognitive) brain regions. Usually the sense of smell dulls first (peanut-butter test).
- euyyn 9y ago
  I have a poor sense of smell (used to have a good one as a child) and my grandmother had dementia. What is the peanut-butter test?
  - r00fus 9y ago
    Possibly this? https://en.wikipedia.org/wiki/Peanut_butter_test https://en.wikipedia.org/wiki/Peanut_butter_test "The researchers concluded that patients with Alzheimer's were not able to smell the peanut butter as well through their left nostril as their right one"
    - dsego 9y ago
      That's the one. My grandfather had dementia. And my father now has signs of brain atrophy, although still no cognitive decline. But he is in his 60s and has had hearing issues for the last decade, combined with ptsd, depression, anxiety and alcoholism. All of it (including mental illness in my family) is most probably somehow connected. I've also found some resources on how having bad teeth (which I do) is also correlated. For now I'm trying to be fitter and take bigger doses of B-vitamins and some other "nootropics". Also, looking forward to the new generation of meds now being tested, such as aducanumab.
    - euyyn 9y ago
      It says that it could not be replicated :/
- pwthornton 9y ago
  The authors seem to think that early hearing loss contributes to isolation and lack of socialization. If properly treated, this shouldn't happen to a person.
matt_wulfeck 9y ago
Having trouble with dementia? Try going back and completing secondary school! Our "study" shows it's the second largest factor in cause of dementia!
munificent 9y ago
If you read the linked paper (access is free), you will note that the authors are quite clear that they have a limited ability to determine causation. They say there seems to be some evidence that some of these risk factors are forward causes, but acknowledge it may be correlation based on another cause, or bicausal where the risk factor increases dementia which in turn reinforces the risk factor. As usual, the popular news article goes straight to the causality argument because calls to action are more effective at getting people to read something.
tuxracer 9y ago
"No, correlation does not imply causation, but it sure as hell provides a hint." http://www.slate.com/articles/health_and_science/science/2012/10/correlation_does_not_imply_causation_how_the_internet_fell_in_love_with_a_stats_class_clich_.html http://www.slate.com/articles/health_and_science/science/201...
dang 9y ago
"Please avoid introducing classic flamewar topics unless you have something genuinely new to say about them." https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- Will_Parker 9y ago
  Correlation vs causation isn't a flamewar topic. It's a pervasive problem in science journalism and public understanding that leads to bad decisions and bad policy. Blatant mistakes in understanding, very clear in the news article (not the original paper) deserve to be called out.

This! Hearing loss could mean those nerves started dying out before other (cognitive) brain regions. Usually the sense of smell dulls first (peanut-butter test).

euyyn 9y ago
I have a poor sense of smell (used to have a good one as a child) and my grandmother had dementia. What is the peanut-butter test?
- r00fus 9y ago
  Possibly this? https://en.wikipedia.org/wiki/Peanut_butter_test https://en.wikipedia.org/wiki/Peanut_butter_test "The researchers concluded that patients with Alzheimer's were not able to smell the peanut butter as well through their left nostril as their right one"
  - dsego 9y ago
    That's the one. My grandfather had dementia. And my father now has signs of brain atrophy, although still no cognitive decline. But he is in his 60s and has had hearing issues for the last decade, combined with ptsd, depression, anxiety and alcoholism. All of it (including mental illness in my family) is most probably somehow connected. I've also found some resources on how having bad teeth (which I do) is also correlated. For now I'm trying to be fitter and take bigger doses of B-vitamins and some other "nootropics". Also, looking forward to the new generation of meds now being tested, such as aducanumab.
  - euyyn 9y ago
    It says that it could not be replicated :/
pwthornton 9y ago
The authors seem to think that early hearing loss contributes to isolation and lack of socialization. If properly treated, this shouldn't happen to a person.

matt_wulfeck 9y ago

Having trouble with dementia? Try going back and completing secondary school! Our "study" shows it's the second largest factor in cause of dementia!

munificent 9y ago

If you read the linked paper (access is free), you will note that the authors are quite clear that they have a limited ability to determine causation. They say there seems to be some evidence that some of these risk factors are forward causes, but acknowledge it may be correlation based on another cause, or bicausal where the risk factor increases dementia which in turn reinforces the risk factor. As usual, the popular news article goes straight to the causality argument because calls to action are more effective at getting people to read something.

tuxracer 9y ago

"No, correlation does not imply causation, but it sure as hell provides a hint." http://www.slate.com/articles/health_and_science/science/2012/10/correlation_does_not_imply_causation_how_the_internet_fell_in_love_with_a_stats_class_clich_.html http://www.slate.com/articles/health_and_science/science/201...

dang 9y ago

"Please avoid introducing classic flamewar topics unless you have something genuinely new to say about them." https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html

Will_Parker 9y ago
Correlation vs causation isn't a flamewar topic. It's a pervasive problem in science journalism and public understanding that leads to bad decisions and bad policy. Blatant mistakes in understanding, very clear in the news article (not the original paper) deserve to be called out.