In which study design does the case serve as its own control and is particularly suited when exposure can change over time and the disease has short duration?

• A. Case-control
• B. Cross-sectional
• C. Case-crossover
• D. Cohort

Answer: (C)

The key idea is a design where the person is compared to themselves, using exposure in a short window before an acute event as the case period and exposure in one or more other time periods for the same person as control periods. Because the comparison is within the same individual, fixed personal characteristics that could confound a between-person analysis cancel out. This makes the approach particularly powerful when the exposure can change over time and the disease is short-lived or episodic, since it examines whether a transient exposure precedes an acute outcome.

In this setup, you look at whether exposure occurred during the hazard window just before the event and compare it to exposure during selected control windows for the same person. This design naturally controls for stable confounders (like genetics, chronic health status, or long-standing behaviors) and helps address confounding by constant factors, while focusing on temporality.

Other designs don’t fit as well here. A case-control study compares cases to unrelated controls, reintroducing between-person confounding and making it harder to capture transient, time-varying exposures. A cross-sectional study measures exposure and disease at one point in time, so temporality and acute exposure effects are hard to establish. A cohort follows exposure over time in many individuals, which can be inefficient for studying brief, transient exposures leading to short-lived events, and still suffers from between-person confounding unless very large and carefully designed.