Table 3

β-Coefficients for final ordinary least squares regression model including meteorological parameters, confounders, pollution indices, and the significant interaction term*.

Predictor

Coefficient

P

Constant

460.30

<0.001

Heart disease mortality

-0.19781

<0.001

Average annual unemployment rate

- 1.5025

<0.001

Acres of land used to grow crops

-0.00001040

0.188

Shortwave radiation

- 33.10

0.011

Heating degree days

- 0.008341

<0.001

Average annual snowfall

0.00789

0.636

Average annual rainfall

-0.11708

<0.001

Heating degree days squared

0.00000157

<0.001

Shortwave radiation squared

0.9851

0.021

Interaction term (Crop)(Snow)

0.00000046

<0.001

*This model explained approximately 18.5% of the variation in the county level average annual incidence rate of prostate cancer (R² = 0.185) and had the lowest AIC in the GWR analysis.

St-Hilaire et al. International Journal of Health Geographics 2010 9:19   doi:10.1186/1476-072X-9-19