This is one of those rare longevity posts where the “boring” intervention in the clinic fridge is backed by better causal inference than most of the flashy stacks.
The Wales rollout is especially compelling because it functioned like a real-world randomized-ish design: a hard eligibility cutoff + regression discontinuity lets you compare people born just on either side of the line. That Nature analysis found ~3.5 percentage-point absolute reduction in new dementia diagnoses over ~7 years (≈20% relative), with a stronger signal in women. 
And it’s not a one-off: similar natural-experiment / large-dataset signals have now shown up across settings and vaccine types, including the recombinant shingles vaccine (Shingrix) associated with longer “diagnosis-free” time in a large U.S. EHR analysis. The newer Canadian regression-discontinuity paper adds even more weight to the “this might be real” side of the ledger. 
Mechanism is still the honest unknown (reduced VZV reactivation? trained immunity? inflammation modulation?), but the policy irony you highlight is the punchline: we’ll pay for blue goo, yet underuse a potentially brain-protective vaccine that already meets standard preventive-medicine criteria.
Brilliant article! Do we know if there is a "sweet spot" for timing, or could getting the vaccine earlier than age 65 offer even more neuroprotection before the decades-long Alzheimer's pathology starts?
I’ve been debating trying to get the shingles vaccine as someone who’s no where near the recommended age for this because the research seems so genuinely impactful.
This is one of those rare longevity posts where the “boring” intervention in the clinic fridge is backed by better causal inference than most of the flashy stacks.
The Wales rollout is especially compelling because it functioned like a real-world randomized-ish design: a hard eligibility cutoff + regression discontinuity lets you compare people born just on either side of the line. That Nature analysis found ~3.5 percentage-point absolute reduction in new dementia diagnoses over ~7 years (≈20% relative), with a stronger signal in women. 
And it’s not a one-off: similar natural-experiment / large-dataset signals have now shown up across settings and vaccine types, including the recombinant shingles vaccine (Shingrix) associated with longer “diagnosis-free” time in a large U.S. EHR analysis. The newer Canadian regression-discontinuity paper adds even more weight to the “this might be real” side of the ledger. 
Mechanism is still the honest unknown (reduced VZV reactivation? trained immunity? inflammation modulation?), but the policy irony you highlight is the punchline: we’ll pay for blue goo, yet underuse a potentially brain-protective vaccine that already meets standard preventive-medicine criteria.
Brilliant article! Do we know if there is a "sweet spot" for timing, or could getting the vaccine earlier than age 65 offer even more neuroprotection before the decades-long Alzheimer's pathology starts?
I don’t know of any evidence on this I’m afraid but thanks for reading
Nice piete. And don't worry, some of us longevity seekers are on this one :)
Fascinating - just had mine on the NHS - I take every vaccine I'm offered.
Brilliant. Love the way you cut through the science - will look into its availability over here in Switzerland
Thank you very much!
I’ve been debating trying to get the shingles vaccine as someone who’s no where near the recommended age for this because the research seems so genuinely impactful.
Strong recommend. I paid for it about a year ago on weaker evidence and also to protect against shingles.