This article is light on a lot of details, I'd be cautious about believing this in any significant way. Taxi driving in London is known to have a heavy cognitive load because of the training and skill required. Where I live taxi and rideshare services now rely exclusively on software for route finding. But even in London the causality may run the opposite way - without exceptional hippocampal function you may not be able to become a taxi driver. So the activity is not necessarily neuroprotective, but the selection process may weed out vulnerable people.

Secondly the odds ratio is not impressive here. The non-taxi groups get Alzheimer's at a rate of 1.66%. the taxi groups get it at 1%. So this not a stochastic difference, just a decrease in the rate. I'm on vacation so I'll leave it to someone else to look up the study numbers and punch them into a calculator.

Also of interest is the occupational definition of how long someone needs to be a "taxi driver" to be classified that way. If someone has declining cognitive performance they may be streamed into another occupation before retirement or diagnosis. And so on.