To be clear on this study, it's a longitudinal study of 300k people who were given the shingles vaccine based on a very basic criteria: if you were born before September 2 1933 you could not get the shingles vaccine, and if you were born after that date you got to have the vaccine.
The explicit date removes the vast amount of confounding factors the usually occur in large studies. They are to compare the outcomes for people born in the same place in the week preceding Sep 2 '33, and the week after, and see an 8.5% difference in rates of dementia.
The study isn't _just_ that week, it's over every person that received the shingles vaccine since 2013 in a program that said "if you are born after 9/2/33 you can get the shingles vaccine, and if before then you cannot". This study is over all people in that program (~300k people), regardless of exactly when they were born. The study found an 8% reduction in dementia diagnoses that correlates entirely to whether the person's DoB was before or after that date. Presumably the only reason they don't say "to the day" is because the number of people involved makes statistical variance too high (and lets be honest there's probably some fudging of people's DoB if you're one day off no matter the rules).
> Using our regression discontinuity approach, we find that being eligible for the zoster vaccine caused a 1.3 (95% CI: 0.2 – 2.7; p=0.022) percentage point absolute, and 8.5% relative, reduction in the probability of a new dementia diagnosis over our seven-year follow-up period[0]
Although the magnitude is small, the interesting thing here isn't that you could get a shingles vaccine to marginally reduce your dementia risk, but that if one virus can cause dementia, it's probably not the only one.
Additionally, it could make understanding the mechanism much easier. Viruses are only so complicated, so learning what they're doing to the brain to cause dementia might reveal how other, non-viral causes work.
My doctor advised me to get the Shingrix vaccine with a 97% effective rate instead of Zoster with a 52% effective rate. But if Zoster also prevents dementia and Shingrix doesn't, what's the expectation-maximizing choice?
12 comments
[ 3.4 ms ] story [ 41.8 ms ] threadThe explicit date removes the vast amount of confounding factors the usually occur in large studies. They are to compare the outcomes for people born in the same place in the week preceding Sep 2 '33, and the week after, and see an 8.5% difference in rates of dementia.
> Using our regression discontinuity approach, we find that being eligible for the zoster vaccine caused a 1.3 (95% CI: 0.2 – 2.7; p=0.022) percentage point absolute, and 8.5% relative, reduction in the probability of a new dementia diagnosis over our seven-year follow-up period[0]
1.3% absolutely reduction seems like a big deal.
[0] https://www.medrxiv.org/content/10.1101/2023.05.23.23290253v...