Lead Author(s): Jeff Martin, MD

Definition of Restriction

Definition of Restriction: Restrict enrollment to only those subjects who have a specific value/range of the confounding variable

For example, when age is confounder, like it often is because age is related to many disease states,

Restriction to Reduce Confounding

In the following DAG, you see that restriction:
  1. Eliminates any possible association between the exposure and the confounder
  2. Eliminates any possible association between the disease and confounder

There is no variability in the values of the confounding variable.

Example Restriction to Reduce Confounding - Down Syndrome

No variability in the Confounder precludes associaton with Exposure and Disease


In studies looking at the association between birth order and Down Syndrome, One way to deal with this would be to conduct a study where you restricted enrollment to women of a certain age or narrow age range.

Example Restriction to Reduce Confounding - HHV

Research Question: Is there an association between sexual behavior and acquisition of HHV-8 infection?
Issue: Is association confounded by injection drug use?

Restriction is particularly useful in situations Consider, for example, the research question of whether there is an association between sexual behavior and the acquisition of HHV-8 infection. Cannon found that the study sample was restricted to persons denying injection use.

Advantages and Disadvantages of Restriction


(1) Conceptually straightforward - (2) Handles difficult to quantitate variables - (3) Can be used in the analysis phase or the design phase - Disadvantages:

(1) May limit number of eligible subjects - (2) Inefficient to screen subjects, then not enroll - (3) Residual confounding may persist if restriction categories not sufficiently narrow - (4) Limits generalizability - The first order of business is internal validity. Generalizability can follow later.

(5) Not possible to evaluate the relationship of interest at different levels of the restricted variable - Bottom Line: Not used as much as it should be


Cannon, M. J., Dollard, S. C., Smith, D. K., Klein, R. S., Schuman, P., Rich, J. D., et al. (2001). Blood-borne and sexual transmission of human herpesvirus 8 in women with or at risk for human immunodeficiency virus infection. N Engl J Med, 344(9), 637-643.