Fig. 1

Overview of the three analytical approaches used to investigate the effect of maternal body mass index on adverse pregnancy and perinatal outcomes. A brief description of each approach is presented in the context of exploring the effect of maternal BMI on APPOs’ risk. Given each approach has different strengths and limitations, findings that agree across approaches are likely to be more credible. The description of each approach is simplified for illustration purposes. An extensive description of assumptions and sources of bias for each approach has been reported previously (e.g. [17,18,19,20,21]). The box around the confounders in the multivariable regression reflects the assumption of the method that all confounders were accurately adjusted for in the analyses. BMI, body mass index; APPOs, adverse pregnancy and perinatal outcomes