I'm curious what you'd say, @ethanjweiss. The "reverse causality" hypothesis doesn't really work in this case, as these are all very elevated LDL-C levels 16 years prior... and likely not a coincidence they match up with high HDL and low TG, no?
-
-
Replying to @DaveKeto @ethanjweiss and
The obvious answer would be survivorship bias of some description. People who survive to 100 are weird in many ways!
1 reply 0 retweets 3 likes -
Replying to @GidMK @ethanjweiss and
... and yet, they are surprising near uniform in the triad thus far. I’m excited to see upcoming centenarians and how much/little we see this phenomenon continue.
1 reply 0 retweets 0 likes -
-
Replying to @GidMK @ethanjweiss and
I want to see how common the triad is in centenarians (High LDL, high HDL, low TG) when tested a decade and a half earlier.
1 reply 0 retweets 0 likes -
Replying to @DaveKeto @ethanjweiss and
But you'd still be seeing survivorship bias there. You're seeing the people who, for whatever reason, can be in that state and live for a long time. Doesn't say much about the rest of the population imo
1 reply 0 retweets 0 likes -
Replying to @GidMK @ethanjweiss and
Survivorship bias makes sense if we were testing these 100/101 year olds a year earlier. But I'm emphasizing these markers (high LDL, high HDL, low TG) were taken 15 years earlier when they were age 84/85. The other 118 of those 84/85 taken at that time are not centenarians.
1 reply 0 retweets 1 like -
Replying to @DaveKeto @ethanjweiss and
Well, the sample size is still tiny. But even so, the case-control you're suggesting would definitely be open to survivorship bias, because you're selecting people who have lived a very long time. More appropriate to start at, say 50, and see how many make it to 100
2 replies 0 retweets 2 likes -
You could do this with a retrospective cohort design rather than the case-control you're thinking of. This tells you how many people have the triad and die vs the number that survive to 100
1 reply 0 retweets 2 likes -
Replying to @GidMK @ethanjweiss and
Would love to. If you've followed me for a while, I know just how much I'd love to get ahold of more raw datasets to test this out directly. Obviously, Framingham Offspring would be especially useful given followups, etc.
1 reply 0 retweets 0 likes
Yeh we should all move to Scandinavia it's the mecca of long-term linked data. Would be amazing to work with the datasets they have in Sweden and Denmark
-
-
Replying to @GidMK @ethanjweiss and
Believe me, I could have a very un-twitter-like rant about access to data in this industry. Coming from technology, we are awash in free, robust data — not so in nutrition/medicine. The irony is that the latter would benefit far more in impacting our lives if it were open source.
1 reply 0 retweets 2 likes -
Replying to @DaveKeto @ethanjweiss and
That we can definitely agree on, but I also understand it. As a society, we are far too free with our non-medical data, even if it's incredibly private. Conversely, we're incredibly cagey about anything medical, even if it's entirely deidentified and boring
1 reply 0 retweets 2 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.