From the use of surrogate endpoints (based on biological plausibility and the results of preclinical studies) to the distinction between correlation and causation, confounding (whether the effect size is too large to be explained away by confounding), and the hypothesis-generating nature of the inferential process.
If that confounding variable can be taken care of, though, and the effect persists, the next obvious step is to determine what mechanism within the mother's body causes the discrepancy.