In uncomplicated pregnancies, birthweight is inversely associated with adult non-communicable disease (NCD) risk. One proposed mechanism is maternal malnutrition during pregnancy. Another explanation is that shared genes link birthweight with NCDs. Both hypotheses are supported, but evolutionary perspectives address only the environmental pathway. We propose that genetic and environmental associations of birthweight with NCD risk reflect coordinated regulatory systems between mother and foetus, that evolved to reduce risks of obstructed labour. First, the foetus must tailor its growth to maternal metabolic signals, as it cannot predict the size of the birth canal from its own genome. Second, we predict that maternal alleles that promote placental nutrient supply have been selected to constrain foetal growth and gestation length when fetally expressed. Conversely, maternal alleles that increase birth canal size have been selected to promote foetal growth and gestation when fetally expressed. Evidence supports these hypotheses. These regulatory mechanisms may have undergone powerful selection as hominin neonates evolved larger size and encephalisation, since every mother is at risk of gestating a baby excessively for her pelvis. Our perspective can explain the inverse association of birthweight with NCD risk across most of the birthweight range: any constraint of birthweight, through plastic or genetic mechanisms, may reduce the capacity for homeostasis and increase NCD susceptibility. However, maternal obesity and diabetes can overwhelm this coordination system, challenging vaginal delivery while increasing offspring NCD risk. We argue that selection on viable vaginal delivery played an over-arching role in shaping the association of birthweight with NCD risk. Birthweight robustly predicts the risk of adult diseases such as hypertension, type 2 diabetes and cardiovascular disease. Potential underlying mechanisms include maternal malnutrition in pregnancy, or common alleles underlying both birthweight variability and adult disease risk. Since heavier babies have both better short-term survival and lower adult disease risk, why are most babies born with lower birthweights than would maximize these benefits? We argue that all mothers are at risk of gestating a baby that is too large to pass through the birth canal, resulting in obstructed labour. This problem increased during the evolution of the genus Homo, as neonates evolved larger head size and weight. Natural selection therefore favoured regulatory mechanisms, involving both genetic and physiological pathways, that constrain fetal growth and gestation length to match the dimensions of the maternal pelvis. These mechanisms appear to work better for maternal height, correlated with pelvic dimension, than for high maternal adiposity which was rare until recent millennia. We argue that selection on viable vaginal delivery played an over-arching role in shaping the association of birthweight with NCD risk. Public health interventions to increase birthweight through nutritional supplementation have modest impact, as they may trigger the defence mechanisms we describe.
