alt Hacker NewsStudies like this always seem to cite stats in a way that's pretty inaccessible to me. This is more clear to me:
* 217,122 participants whose data was extracted from the UK biobank database
* Out of those 217,122, 325 got early onset dementia over an average of 8.3 years
* The vast percentage of data came from exactly one blood draw per person between 2006 and 2010 at the beginning of the biobank study
Omega-3 Blood | Hazard Risk | Rate of Incidence | Percent Incidence
Level Quintiles | | Over 8.3 Years | Over 8.3 Years
-------------------|------------------|--------------------|------------------
Q1 (Lowest 20%) | 1.0 | 193 in 100,000 | 0.193%
Q4 (High) | 0.62 | 120 in 100,000 | 0.120%
Q5 (Highest 20%) | 0.60 | 116 in 100,000 | 0.116%Replies
So you can reduce your dementia incidence risk from Q1 -> Q5 by a whopping 0.08%-points. But in media you surely will read about a 40% reduction.
*edited: %-points instead of %
also keep in mind, P hacking came about as a means to try to prove racial science.
This could significantly underestimate the real impact. A single point measurement is perhaps a pretty noisy measure of long term average. If we had lifetime averages, the quintiles would be more purely differentiated by the variable of interest, and the risk would be as well.
The most interesting finding is that the non-DHA effect is much stronger than the DHA effect. This doesn't align with the mechanistic explanation. Either this this is a novel and interesting result, or it's more evidence that we're just measuring wealth and health consciousness.
Observational studies like these are useful for guiding future research, but, on their own, they're essentially useless for informing lifestyle changes.