O-10 - EPIC egg-IHD inverse association became nonsignificant after excluding first 4 years

Link to original

Why this is evidence

Per-member (H-2, H-3, H-4, H-9, H-10, H-12):

  1. H-2: disconfirmed - the protective signal does not survive a reverse-causation lag.
  2. H-3: mildly favoured relative to H-2 (removes the counter-evidence) but nothing positive.
  3. H-4: this is close to a signature prediction - H-4 attributes the inverse IHD signal to reverse causation, and the lag analysis is what would show that.
  4. H-9: predicted (no association).
  5. H-10: predicted (no material effect).
  6. H-12: unengaged. The main discrimination is H-4 up, H-2 down.