I used Satscan to determine where clusters of disease are located. The software includes in its output the relative risk of someone living in an area developing a disease versus someone not in the area developing the disease. Out of curiosity I used relative risk as the dependent variable in a regression model (ZINB) and it was correlated to several of my independent variables.
Now I get to try to explain what this means. Is it even meaningful?