>Replicated association [...]

"association" undersells it a bit, because the data is better than the typical cohort study, which has issues like "what if people who got the vaccine are also richer and care about their health more?". There's quasi-randomization going on. From the more in depth article that's linked:

>Research is also revealing unexpected interventions that help to keep ageing minds sharp. One of the most promising derives from an analysis by Pascal Geldsetzer of Stanford University and his team of a natural experiment in Wales. In 2013 the British region started offering people aged 70-79 free vaccinations through the public-health system. This change resembled an RCT, in that a large number of people were separated almost at random into two groups: those who had already turned 80 in the weeks before the programme started, and so were not eligible to be jabbed; and those who turned 80 in the weeks after, roughly half of whom were duly vaccinated.