• Bryan Kestenbaum


In observational studies, associations between potential risk factors and disease may or may not represent causal relationships. Confounding is a type of bias that occurs when characteristics other than the exposure of interest distort the observed association of the exposure with disease. A confounding characteristic is defined as a factor that is associated with the exposure, associated with the outcome, and not suspected to reside on the causal pathway of association. Strategies to control for confounding include restriction, stratification plus adjustment, matching, and regression. The presence of confounding is suspected when the size of the association of interest changes meaningfully after adjustment by one of these methods.

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Bryan Kestenbaum
    • 1
  1. 1.Division of Nephrology, Department of MedicineUniversity of WashingtonSeattleUSA

Personalised recommendations