I would like to use the stata power command, to estimate the smallest effect size I would be able to detect, given the sample size available.

I am assessing the effect of smoking status on bone density (a continuous outcome) using a multivariate linear regression (adjusted for age, sex and a couple of other covariates). I've found no association between smoking status and bone density, and so would like to quantify at what point the effect size would be too small to detect at a range of power and alpha values.

I am a little confused by the many variations of the power command in stata.If anyone could point me in the right directions or had any examples of using stata in this way that would be really helpful.