A meta-analysis shows a dose-response relation: each additional 10 g/day of fiber is associated with a 10% lower risk of coronary heart disease. Which inference is most appropriate?

• A. The dose-response is irrelevant because fiber only affects intestinal cancers and cannot plausibly affect the cardiovascular system under any mechanistic pathway.
• B. This finding proves causality because any dose-response relationship observed in meta-analyses can only result from true biological effects and never from residual confounding.
• C. A 10% lower risk per 10 g must be due entirely to measurement error in the prospective cohorts and therefore is not informative for dietary guidance.
• D. There is consistent evidence of an association with a graded relation, which strengthens plausibility, but causation still requires consideration of confounding and trial evidence for confirmation.

Answer: (D)