-
-
Show this thread
-
Basically, this was a case-control study, where researchers took a group of people with diabetes, another group without, and then looked back to see how much of different types of coffee they drankpic.twitter.com/8XUJoDK5dU
Show this thread -
This type of study is common for environmental risks - you take people who have a relatively rare issue, and see what exposures might have caused it
Show this thread -
In this case, the researchers used two measures of how much coffee people drank: 1. asking them questions 2. metabolites of coffee in their blood
Show this thread -
They found two different things. For the self-reported coffee intake, there was no association between any coffee drinking and diabetes In other words, the opposite of the headlinespic.twitter.com/UxeswpjKDW
Show this thread -
For the metabolite analysis, there was a reduced risk of diabetes for those who had a lot of filtered coffee metabolites in their blood This equated to an odds ratio of 0.42, which is where we get the "60% reduced!" figure from
Show this thread -
But...metabolites aren't necessarily intake, especially since these metabolites were associated with coffee through a complex set of assumptions and statistical analyses
Show this thread -
So what did the study conclude? Well, have a read. Notice how it doesn't mention the wonders of filtered coffee AT ALL?pic.twitter.com/X4FS9POY8B
Show this thread -
So where did the headlines come from? Well, it appears that the senior author of the paper gave what I can only describe as an astonishing interview in the press release
Show this thread -
Here's some quotes. Remember THIS PAPER DIDN'T LOOK AT WHETHER FILTERED COFFEE CAUSED REDUCED DIABETES RATESpic.twitter.com/HG7AzbOv9S
Show this thread -
The aim of the study appears to be mostly to use a new technique to investigate consumption, which is where the metabolites come from Everything to do with coffee reducing diabetes is basically complete nonsense
Show this thread -
The high filtered coffee drinkers - the people least likely to get diabetes - were better educated, younger, thinner, and smoked less than people who didn't drink filtered coffee It's a textbook case of "probably residual confounding"
Show this thread -
There are so many things that might've caused the association between filtered coffee and diabetes that simply weren't measured in this study! It's almost certain that it wasn't the coffee
Show this thread -
It is SO FRUSTRATING when academics are appropriately cautious in studies but then give absurd interviews with total overreach of their resultspic.twitter.com/lCH7zPQZ87
Show this thread -
Anyway, TL;DR: - vague association between filtered coffee and diabetes - probably not causal - drink your coffee however you want



Show this thread
End of conversation
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.