Between and within subject regression

[Ray Mason]

Let’s assume I want to see the change in blood pressure(bp) after exercise sessions in two groups [between subject variable] after adjusting for age and gender.
The data are in long format.
bp=outcome variable (continuous), id=identifying variable, health=grouping variable(healthy vs unhealthy), age=independent variable(continuous) and gender=independent variable(male vs female).
What would be the command for running a regression? Linear mixed effect model?

Thanks in advance.

[Carlo Lazzaro]

Ray:
welcome to this forum.
If you have one wave of data only (that is, the difference in BP) as it is often the case with the so-called mirror studies (in which the participant acts as her/his control), you may want to consider an OLS:

regress diff_BP i.group c.age##c.age

The -age- interaction aims at exploring possibe non-linear relationship with the regressand.

[Ray Mason]

Thanks for the reply.
What if I want to compare the pre and post-session blood pressure, not the difference? and don't we need to specify the within and between subject variables?
What about the linear mixed effect model?

xtmixed c.outcome group independent ||group:time_variable||participant_id:, cov(uns)

[Carlo Lazzaro]

Ray:
thanks for clarifying.
Basically, you seem to have a panel dataset with two waves of data.
I think that you can go -xtreg,fe- and -xtreg,re- and see which one fits your dataset better via -hausman- (if you use default standard errors).
-mixed- can be another option, provided that it is a cousin of -xtreg, mle- (that is a random effect estimator).
In addition, the dependent variable does not support -fvvarlist- notation (that is, you shoul code -outcome- instead of -c.outcome-).