Which statement correctly defines a time-varying covariate?

• A. A time-varying covariate is one that changes across time.
• B. A time-varying covariate is fixed at baseline.
• C. Time-varying covariate effects are independent of time.
• D. Time-varying covariates cannot be modeled in Cox models.

Answer: (A)

The main idea is that a time-varying covariate is one whose value changes during the follow-up period. This matters because the hazard at any moment can depend on the covariate’s current value, not just what it was at the start of the study. In survival analysis, you handle this by using time-dependent covariates in models like the Cox model, updating the covariate as it changes and often splitting the follow-up time into intervals where the covariate is constant. That’s why the statement that a time-varying covariate is one that changes across time is the correct definition.

The other statements don’t define time-varying covariates accurately. Being fixed at baseline describes a time-fixed covariate, not time-varying. Saying that time-varying covariate effects are independent of time misstates the concept, since the effect of a covariate can itself vary over time. And claiming that time-varying covariates cannot be modeled in Cox models is incorrect because extended Cox models explicitly handle time-dependent covariates.