In a study where researchers followed 56,000 pregnancies, 53,271 infants were born alive and survived the neonatal period, 193 babies died of SIDS. The crude (unadjusted) risk of SIDS was found to be 40% higher (RR=1.4) in infants of Black mothers compared to non-Black mothers. Black mothers had low income and education. The risk of SIDS was inversely related to maternal age: compared to children born to mothers aged 35-41 years, the risk of SIDS was more than twice as high (RR=2.3) among infants born to mothers aged 18-19. However, this analysis did not consider the possibility that the data were confounded by parity (number of births the mother has had). A separate study was conducted to investigate the effect of maternal smoking during pregnancy on SIDS. Previous studies had shown that smoking during pregnancy considerably increases the risk of pre-term delivery and low birthweight. Pre-term delivery and low birthweight are also known risk factors for SIDS. Should birth-weight and gestation (pre- versus full-term delivery) be considered as potential confounders in the analysis?

1. Explain clearly how you arrived at your conclusion and share a directed acyclic graph (DAG) to support your reasoning.

2. Clearly describe how you would test for confounding.