What's that you say? That doesn't sound anything like the headlines? I see you've spotted one of the MASSIVE GLARING ISSUESpic.twitter.com/NL2XpFgPY1
-
Show this thread
-
The study used telomere shortening as a proxy for biological age Basically, shorter telomeres means 'older', because shorter telomeres are associated with aging
1 reply 0 retweets 0 likesShow this thread -
(Telomeres are part of our DNA that stop it from breaking when it replicates. You have less of them as you grow older, which some people think is part of the reason that we get less healthy as we age)pic.twitter.com/6iLjGFhzIh
1 reply 0 retweets 2 likesShow this thread -
FIRST PROBLEM: telomeres are NOT age. They are ASSOCIATED with age In other words, you can't predict much from shorter telomeres. It might just be that they get shorter as we age naturally, rather than causing our aging processes per sepic.twitter.com/9gn5y4Zy1e
1 reply 0 retweets 5 likesShow this thread -
So making any claims about adding years to your life from this research is baldly wrong. The study made some weird extrapolations from shortened telomeres to 4.5 years of aging, but that's not very realistic
1 reply 0 retweets 3 likesShow this thread -
SECOND PROBLEM: The analysis was cross-sectional Basically, the researcher (yes just one) took a sample of people at one point in time and looked at whether full-fat milk was associated with telomere issues
1 reply 0 retweets 3 likesShow this thread -
This is a problem, because we haven't measured people over time All we can say is that, in the NHANES sample from 1999-2002, people with shorter telomeres drank more full fat milkpic.twitter.com/p5ZowpcOoI
1 reply 0 retweets 4 likesShow this thread -
Which brings us to the THIRD PROBLEM: this study was observational in nature All the analysis did was look at the association between full/low fat milks and telomere length
1 reply 0 retweets 3 likesShow this thread -
The study didn't even control for many covariates. It's not hard to see how your telomere length (a complex measure of DNA) could be influenced by things that this study failed to control for
4 replies 0 retweets 3 likesShow this thread -
Replying to @GidMK
They claimed they controlled for many of these and didn't see any differences in the correlations- but there were many others probably that weren't even measured so couldn't be controlled for. Observational studies are always weak for this reason.
1 reply 0 retweets 0 likes
Depends on the study. Drawing causal inferences of a treatment effect can be challenging with these datasets, but most of my research is observational and not impacted in any way (i.e. monitoring diabetes rates)
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.