Correlation studies are fine, but I do worry about national GDP as a confounding factor. Maybe poorer countries have lower life expectancies, (or whatever PYLL

Correlation studies are fine, but I do worry about national GDP as a confounding factor. Maybe poorer countries have lower life expectancies, (or whatever PYLL measures is higher), and work longer hours. I would like to see this graph broken down by quantiles of national GDP.