A study finds smokers have a relative risk (RR) of 15 for lung cancer, while sedentary adults have an RR of 1.3 for type 2 diabetes. Which factor could make diabetes prevention the higher public health priority despite the smaller RR?
AThe diabetes study used a larger sample size
BSedentary behavior is far more prevalent, giving it a much higher population attributable fraction
CAn RR below 2 is not statistically meaningful
DThe odds ratio for sedentary behavior corrects for confounding in ways RR cannot
Population attributable fraction (PAF) depends on both the strength of association (RR) and the prevalence of the exposure. If 60% of the population is sedentary, even an RR of 1.3 can account for a huge share of all diabetes cases. A lung cancer RR of 15 is dramatic, but if few people smoke, fewer total cases are attributable. This is why absolute and population-level measures must accompany relative risk when setting priorities.
Question 2 True / False
A relative risk of 8.0 generally indicates a greater public health burden than a relative risk of 1.4.
TTrue
FFalse
Answer: False
Relative risk captures how much more likely an outcome is in the exposed group compared to the unexposed — it does not account for how common the exposure is or how frequent the baseline outcome is. An RR of 8.0 on an exposure affecting 0.1% of the population may cause far fewer total cases than an RR of 1.4 applied to an exposure affecting 70% of the population. Public health impact requires examining attributable risk and population attributable fraction alongside relative measures.
Question 3 Short Answer
In a case-control study of a rare disease, why is the odds ratio (OR) preferred over the relative risk (RR)?
Think about your answer, then reveal below.
Model answer: Case-control studies select participants based on disease status, not exposure, so true population incidence rates cannot be directly calculated. The OR can be computed from the resulting 2×2 table and approximates the RR closely when the disease is rare.
Relative risk requires incidence data from defined exposed and unexposed populations followed over time — a cohort design. In a case-control study, researchers work backward from disease outcomes to past exposures, so the proportion of 'exposed' among cases and controls does not reflect true population incidence. The OR — calculated as (exposed cases × unexposed controls) ÷ (unexposed cases × exposed controls) — avoids this problem and is a valid approximation of RR under the rare disease assumption.