How do I adjust for confounding variables?

Hello everybody,

I wish to look at the correlation between blood pressure and IQ score, sample size is relatively small, N=58.
I have so far performed a Spearman's correlation, but I wish to adjust for confounding variables.
What is the best statistical method to use? Should I use something else than Spearman's?

I'm using STATA.


TS Contributor
Are these cross-sectional data?

As far as I know, blood pressure and IQ score both are interval scaled,
therefore a linear regression analysis would be possible, where you can
include additional variables (not too many, though, if you want to
prevent overfitting).

With kind regards

Thank you for the quick response. Data on blood pressure is retrospectively obtained as they were measured 15 years ago, while data on IQ was obtained recently.
If I wish to adjust for confounding variables, would a multivariate linear regression then be suitable?

Kind regards


Less is more. Stay pure. Stay poor.
Just a comment, a confounder is a common cause of both variables. So did the confounder occur prior to 15 years ago?
for instance, in this study, a potential confounder could be age at time of diagnosis as blood pressure varies with age. Additionally, IQ have previously been shown to be more impaired in individuals who are diagnosed early in life.