Here is the study, I'd recommend the "methods" section it's very enlightening https://www.thelancet.com/journals/lanpub/article/PIIS2468-2667(18)30135-X/fulltext#seccestitle80 …
-
-
Replying to @GidMK @bokkiedog and
It's crucial not to adjust for mediators or their proxies, as explained by
@yudapearl below. If carbohydrates cause CVD via a mediator, e.g. excess insulin production, then diabetes, and arguably energy intake, are proxies, so adjusting for them renders the analysis impotent. 1/pic.twitter.com/B4A0ketuF4
5 replies 4 retweets 32 likes -
Replying to @KetoCarnivore @GidMK and
In so doing, they effectively engineer in some exculpation for insulinogenic foods, and some concomitant demonisation of their counterparts. Finger on the scale, so to speak. Not surprising.
1 reply 1 retweet 6 likes -
Replying to @bokkiedog @KetoCarnivore and
What total and utter nonsense. I'm sorry, but that is a deepity - using innumerable syllables to say nothing at all
3 replies 0 retweets 2 likes -
Replying to @GidMK @bokkiedog and
I understood Nick's sentence. "Engineering in exculpation for insulinogenic foods" is exactly what happens if you adjust for diabetes. It would be a bit like adjusting for traumatic brain injury in car accident deaths.
3 replies 0 retweets 10 likes -
Replying to @KetoCarnivore @bokkiedog and
Not at all. It would be like adjusting for previous traumatic brain injury, an entirely different statistic
1 reply 0 retweets 1 like -
Replying to @GidMK @bokkiedog and
The analogy is imperfect because of the time sequence, but the idea correct: this with a high carb intake and diabetes at baseline are already at risk of death from myriad diseases associated with IR and that risk is likely to be moderated by the carb intake.
1 reply 0 retweets 3 likes -
Replying to @KetoCarnivore @bokkiedog and
The time-sequence is key. If you didn't adjust for diabetes at baseline, the results would be less meaningful as they would be down to which of the groups had more people with diabetes, given that there was no randomization
1 reply 0 retweets 1 like -
Replying to @GidMK @KetoCarnivore and
It is like, for example, running your study on car deaths with a group made up largely of people with terminal mesothelioma and then concluding that this group was at the highest risk of death
1 reply 0 retweets 1 like -
Replying to @GidMK @bokkiedog and
It would be like that, if diabetes were not already a sign of metabolic syndrome that had a plausible causal pathway from excess carbohydrates. It's more like looking at ACM from smoking but adjusting for COPD.
2 replies 0 retweets 3 likes
? Diabetes is a metabolic syndrome, caused by a complex association of environment and genetics. There is no single causal pathway Regardless, that would still not obviate the need to adjust for baseline characteristics
-
-
Replying to @GidMK @bokkiedog and
I'm not saying there's a single causal pathway, only that there's a non-zero causal link between carb intake and diabetes that will skew results if adjusted for. I'm not even saying the results would change, only that it's wrong statistically, if I understand JP correctly.
1 reply 0 retweets 2 likes -
Replying to @KetoCarnivore @bokkiedog and
I think you are misapplying the theory in this case
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.