[R-sig-Geo] areal data with point covariates

[James Rooney]

Dear List,

This is not strictly an R question so I hope it is allowed. I am a little unsure of what type of analysis I need to use next so hoping someone can point me in the right direction.

I have point data representing cases of disease. I also have population data attached to polygons. I have used this to calculate areal standardised incidence rates and applied the BYM model via Bayesian methods to draw a smoothed relative risks map.
Now I would like to check for correlation between this risk and spatial data from other sources - e.g. soil mineral measurements etc. Note that the cases and soil measurements are not colocated.
What should be my next step ? Do I need to krig the soil data and get mean values per polygon ? Or should I be taking an entirely different approach here ?
Again apologies if this is off-topic.

Many thanks,
James