Hi,

I used binomial regression to calculate the risk ratio of outcomes (e.g. diabetes, renal failure, heart failure) comparing cases and controls.
I did this for patients in 4 age groups
And this gave me the following table:
18-29 years 30-39 years 40-49 50+ p value for the trend
Renal failure 5.60 2.30 3.50 2.20 ?
Diabetes 1.70 1.70 1.70 1.50 ?
Heart failure 85.40 29.50 19.90 8.20 ?
Now my questions is, how do I calculate the p value of the trend in these risk ratios across the age groups? Its obvious that the difference in prevalence of diabetes between cases and controls stay the same across age groups and decrease dramatically for heart failure. How do I show if that difference is statistically significant?

Thank you