What is a confounding variable?

What is a confounding variable?

• A. An exposure that causes the outcome directly.
• B. A measurement error in the outcome variable.
• C. A variable that only affects the exposure but not the outcome.
• D. A variable associated with both exposure and outcome that distorts the observed association.

Answer: (D)

Confounding happens when a variable is related to both the exposure and the outcome in a way that makes the observed association between exposure and outcome look different from the true relationship. It isn’t part of the causal path from exposure to outcome, but it can distort the link you measure. For example, in studying whether regular exercise reduces heart disease risk, age can be a confounder because older people may exercise less and also have a higher risk of heart disease; if you don’t account for age, you might wrongly attribute the higher disease risk to low exercise rather than to age. A variable that only affects the exposure but not the outcome wouldn’t distort the exposure–outcome link, and a direct causal factor of the outcome isn’t confounding. To limit confounding, researchers can randomize participants, stratify analyses by the confounder, or adjust for it statistically.

Confounding happens when a variable is related to both the exposure and the outcome in a way that makes the observed association between exposure and outcome look different from the true relationship. It isn’t part of the causal path from exposure to outcome, but it can distort the link you measure. For example, in studying whether regular exercise reduces heart disease risk, age can be a confounder because older people may exercise less and also have a higher risk of heart disease; if you don’t account for age, you might wrongly attribute the higher disease risk to low exercise rather than to age. A variable that only affects the exposure but not the outcome wouldn’t distort the exposure–outcome link, and a direct causal factor of the outcome isn’t confounding. To limit confounding, researchers can randomize participants, stratify analyses by the confounder, or adjust for it statistically.