Further, as pointed out in the review by David Rush : associations are not easily interpreted, and a higher weight of the mother is not necessarily linked to a reduced perinatal mortality. Biological associations are more frequently U-shaped or S-shaped than linear.