This isn't new. From the Pill to ibuprofen, every time a large observational study comes out, the media goes wild https://medium.com/@gidmk/why-is-the-pill-so-depressing-1b429d271d33 …
-
-
Replying to @GidMK
And yet not a single media outlet identifies these as observational. Ever. They act as if these are perfect proof of causation.
1 reply 0 retweets 0 likes -
Replying to @GidMK
Health Nerd Retweeted Health Nerd
This is the first huge problem; YOU CAN'T PROVE CAUSATION WITH AN OBSERVATIONAL STUDYhttps://twitter.com/GidMK/status/857091846438367232 …
Health Nerd added,
1 reply 3 retweets 4 likes -
Replying to @GidMK
You can get a huge amount of information, and with successive studies you can get a very good idea of causation, but one study?
1 reply 0 retweets 1 like -
Replying to @GidMK
After this we have the tired old canards of shoddy reporting; poor understanding of risk ratios, over-reliance on p-values etc
1 reply 0 retweets 2 likes -
Replying to @GidMK
Health Nerd Retweeted Health Nerd
In this particular study's case, there WAS no effect of diet drinks on dementia. All the articles were wronghttps://twitter.com/GidMK/status/857092472379506688 …
Health Nerd added,
1 reply 0 retweets 0 likes -
Replying to @GidMK
Health Nerd Retweeted Health Nerd
When the authors controlled for other risks (such as diabetes), the significance disappearedhttps://twitter.com/GidMK/status/857092472379506688 …
Health Nerd added,
1 reply 0 retweets 2 likes -
Replying to @GidMK
That leaves us with stroke, for which the authors did find a significant association with diet drinks
1 reply 0 retweets 0 likes -
Replying to @GidMK
And whilst this is interesting, there was sufficient information in their study to infer that it wasn't actually due to the diet drinks
1 reply 0 retweets 0 likes -
Replying to @GidMK
For example, the authors didn't control for BMI (even though they had this information), but BMI was higher in diet drinkers
1 reply 0 retweets 2 likes
So it's very likely that it was actually BMI (a known risk factor) that was the major influence on stroke, not diet drinks
-
-
Replying to @GidMK
This is backed up by the fact that there was no biological gradient; the risk from drinking diet drinks didn't increase the more you drank
1 reply 0 retweets 1 like -
Replying to @GidMK
The confidence intervals (a stats measure) were all also huge, meaning that a lot of these findings are just *barely* significant
1 reply 0 retweets 0 likes - Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.