A nationwide randomised trial is now recruiting adults aged 65 or older in Denmark to test whether the recombinant shingles vaccine reduces the risk of dementia and of major cardiovascular events. According to its ClinicalTrials.gov record (NCT07485283), DAN-ZOSTER is an open-label phase 4 trial that plans to randomise approximately 162,000 people in a 1:1 ratio either to two intramuscular doses of Shingrix, given 2 to 6 months apart, or to no study vaccination. The trial has two dual-primary outcomes: major adverse cardiovascular events, defined as a composite of non-fatal myocardial infarction, non-fatal stroke or cardiovascular death, and incident dementia. Participants are identified and invited through national registries and the Danish digital mail system, and all follow-up is collected by linkage to nationwide health registries rather than clinic visits. The record lists estimated primary completion and study completion in April 2029. The trial is run from the Department of Cardiology at Copenhagen University Hospital Herlev and Gentofte, with Tor Biering-Sorensen as study chair, and carries EU clinical trial number 2025-524944-37-00.

The design matters because every dementia finding for zoster vaccines so far has come from observational or quasi-experimental data, which cannot on its own establish cause. This newspaper reported in June on a target trial emulation published in the Annals of Internal Medicine on 16 June 2026, in which Kaleen Hayes and colleagues followed more than 500,000 Medicare beneficiaries aged 66 or older admitted to skilled-nursing facilities and found roughly a 24 percent lower four-year risk of a dementia diagnosis among those who received the recombinant vaccine; that was an association drawn from claims and nursing-home records, not a randomised comparison. The most widely cited earlier work, a natural experiment published in Nature on 2 April 2025 using a date-of-birth eligibility cutoff in Wales, reported about a one-fifth lower probability of a new dementia diagnosis over seven years, but it studied the older live-attenuated vaccine, not the recombinant one now in use. DAN-ZOSTER is not alone: a GlaxoSmithKline-sponsored placebo-controlled, observer-blind trial in Finland (NCT07502560), posted on 31 March 2026, will randomise adults aged 76 or older 3:1 to recombinant zoster vaccine or placebo, with estimated primary completion in April 2030 and study completion in 2037.

Randomisation does not remove every difficulty. DAN-ZOSTER is open-label, the control group receives no placebo, and the trial's own participant information states that people allocated to the control group remain free to buy the vaccine outside the trial, which can blur the comparison over time. Enrolment depends on consent, so volunteers may differ from the wider Danish population aged 65 and over; people already diagnosed with dementia, those previously vaccinated against shingles and those on immunosuppressive treatment for chronic inflammatory rheumatic disease are excluded. Dementia is counted from registry diagnosis codes rather than from research-standard cognitive assessment, and the follow-up window to April 2029 is short relative to a disease process that develops over decades. Danish specialist coverage of the trial has noted that dementia results are not expected before 2029.

The recombinant vaccine is licensed in the European Union for prevention of shingles and post-herpetic neuralgia, not for prevention of dementia, and no regulator has approved it for that purpose. Nothing in the registration of DAN-ZOSTER changes what is currently known: the dementia signal remains unproven in randomised testing, and that is precisely why the trial is being run.