Age-adjusted P value

#1
Hi,

I have a table of baseline characteristics of a cohort in which I have compared the differences between those with kidney disease and those without, using a combination of Chi-square test, t-test, or Kruskall-Wallis test as appropriate. How do I age-adjust the P values that I get from these tests?

Thanks so much for your help!

Marie
 

Karabiner

TS Contributor
#2
You cannot age-adjust p values. You can include age into your predictive model, which will become
a bit more complicated, though, with 2 predictors (group & age) rather than with one. For example,
you can use a linear regression with group (dummy coded as 0 / 1) and age as predictors, instaed
of a t-test.

With kind regards

Karabiner
 

hlsmith

Not a robit
#3
You could use linear regression and control for it in the model. How did you find your control group? Watch out for health worker bias in controls or survival bias in the cases.
 

ondansetron

TS Contributor
#4
Hi,

I have a table of baseline characteristics of a cohort in which I have compared the differences between those with kidney disease and those without, using a combination of Chi-square test, t-test, or Kruskall-Wallis test as appropriate. How do I age-adjust the P values that I get from these tests?

Thanks so much for your help!

Marie
It is generally inappropriate to employ p-values when looking at baseline characteristics.