[canjobear]

My conclusion is that olive oil consumption has no effect on mortality. Finding an effect in 1/2 of your data, and not in the other half, is a classic sign of p-hacking.