Which piece of evidence most strongly supports the policy that 'no amount of alcohol is safe during pregnancy'?

A single case report showing no adverse outcomes after heavy drinking in one pregnancy, proving that population-wide recommendations are unnecessary.

Longitudinal research demonstrating that alcohol exposure during the third trimester only causes transient, fully reversible learning delays that resolve by early childhood in all cases.

The inability of the early embryo/fetus to effectively metabolize alcohol combined with data showing a range of dose-independent outcomes (FASD spectrum) and the absence of a known safe threshold for critical developmental periods.

Epidemiologic studies that conclusively demonstrate moderate drinking improves fetal neurodevelopmental outcomes compared with complete abstinence in randomized controlled trials.