Comment by ed
2 years ago
> Even worse he never tells us the number of humans being examined
You might want to actually open the paper. N = 282,541.
They have data for 98% of the population of wales. Of the ~3m people there, 282,541 were in the relevant age bracket.
I did quickly peruse the paper now. If I'm understanding it right the probability of being affected by dementia goes from 16% to 12% for the 2 different populations, and it also only seems to work for women. For women it decreases from 17% to 12% while for men it stays exactly the same. With these numbers I rest my case, if this is being taken seriously by the academic community I do not know what to say. What if more women were diagnosed of dementia before that time period (because say women were generally considered hysterical) and a definitional change in dementia reduced diagnosis cases?. What if hospitals were getting funded for taking more psychiatry patients and then that reduced?. What if there was a change in a popular birth control supplement? I cannot see how this can lead to any clean conclusive hypothesis.