This is the problem with all cohort studies. Since there aren't strong controls, you can't determine causality. Fundamentally, this is why we know get consiste

This is the problem with all cohort studies. Since there aren't strong controls, you can't determine causality. Fundamentally, this is why we know get consistently conflicting advice about nutrition: we take cohort studies as law instead of using them to form hypotheses that are tested with well controlled experiments.