This study had numerous key personal risk factors and these were used in the analyses to reduce biases from confounding. The study must still be considered as ecological, as comparisons are made between regions. One can consider these adjustments as a refined method of standardization, within the context of logistic regression.