You might handle it with a longitudinal SEM or causal net since you have time separating effects (parental intelligence comes before wealth which comes before nutrition/prenatal-care which comes before childrens' intelligence); but for that specific case, nutrition & prenatal care are already largely ruled out as causally relevant since they fall under 'shared environment', which for IQ in the West is very low.