If we want to adjust by age in a case-control study, we could do a matching 1:N by age (conditional logistic regression) or just consider age as another covariate in the model, without a matching (non-conditional logistic regression).
What are the advantages of both procedures?