4 ms·

This seems like a better interpretation (linked and mentioned in the article): https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2798908 https://jam

This seems like a better interpretation (linked and mentioned in the article): https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2798908 https://jamanetwork.com/journals/jamanetworkopen/fullarticle...

graymatters 4y ago
Thank you for sharing this interpretation!
defrost 4y ago
Cheers - better technical discussion being an orig paper on the subject: > Zhang et al performed a meta-analysis of 19 observational studies with more than 3.9 million participants followed from 3 months to 9.5 years (median follow-up, 1.5 years) and found no association between use of medications to treat attention-deficit/hyperactivity disorder (ADHD) and any cardiovascular disease (CVD) outcome. Caveats (my thoughts on what to look into): * Meta analysis of other studies can be a real mixed bag, they almost always deserve a closer look. * Median Follow-up 1.5 years means half of the patients were tracked for less than 18 months ... (hmmm) * With no distribution it's possible that literally one patient was tracked for 9.5 years, and the vast bulk of patients for under three years ... * "Found no association" is a statement that is consistent with none of the studies showing a significant above expected number of heart issues; this may be because the bulk of the studies were too short term and (possibly) not especially focused on such issues - hence the need to look deeper. I have no opinion either way, just the above notes on where to look further.
- lukas099 4y ago
  > Meta analysis of other studies can be a real mixed bag, they almost always deserve a closer look. My understanding was that meta analyses were better than individual studies, since they are 'samples of samples,' so to speak --- admittedly my understanding of the matter comes from the picture at the top of [0] though. [0] https://flexiblelearning.auckland.ac.nz/philson/51_5.html https://flexiblelearning.auckland.ac.nz/philson/51_5.html
  - defrost 4y ago
    > My understanding was that meta analyses were better than individual studies, since they are 'samples of samples,' so to speak In principle and in theory .. yes. In practice some can be "publish or perish" pumped papers that are thrown out at speed at lack deep insight when comparing apples, pears, and oranges. They're also rather notorious for having been quoted during the recent pandemic years to support every COVID claim you care to mention (ie. vaccines good, vaccines bad, masks good, masks bad, horse dewormer good, etc. ) A good meta study paper is a mine of useful infomation, a poor one is grist for conspiracy mills.

graymatters 4y ago

Thank you for sharing this interpretation!

defrost 4y ago

Cheers - better technical discussion being an orig paper on the subject: > Zhang et al performed a meta-analysis of 19 observational studies with more than 3.9 million participants followed from 3 months to 9.5 years (median follow-up, 1.5 years) and found no association between use of medications to treat attention-deficit/hyperactivity disorder (ADHD) and any cardiovascular disease (CVD) outcome. Caveats (my thoughts on what to look into): * Meta analysis of other studies can be a real mixed bag, they almost always deserve a closer look. * Median Follow-up 1.5 years means half of the patients were tracked for less than 18 months ... (hmmm) * With no distribution it's possible that literally one patient was tracked for 9.5 years, and the vast bulk of patients for under three years ... * "Found no association" is a statement that is consistent with none of the studies showing a significant above expected number of heart issues; this may be because the bulk of the studies were too short term and (possibly) not especially focused on such issues - hence the need to look deeper. I have no opinion either way, just the above notes on where to look further.

lukas099 4y ago
> Meta analysis of other studies can be a real mixed bag, they almost always deserve a closer look. My understanding was that meta analyses were better than individual studies, since they are 'samples of samples,' so to speak --- admittedly my understanding of the matter comes from the picture at the top of [0] though. [0] https://flexiblelearning.auckland.ac.nz/philson/51_5.html https://flexiblelearning.auckland.ac.nz/philson/51_5.html
- defrost 4y ago
  > My understanding was that meta analyses were better than individual studies, since they are 'samples of samples,' so to speak In principle and in theory .. yes. In practice some can be "publish or perish" pumped papers that are thrown out at speed at lack deep insight when comparing apples, pears, and oranges. They're also rather notorious for having been quoted during the recent pandemic years to support every COVID claim you care to mention (ie. vaccines good, vaccines bad, masks good, masks bad, horse dewormer good, etc. ) A good meta study paper is a mine of useful infomation, a poor one is grist for conspiracy mills.