Study design and population
Our study was conducted in the ongoing Rotterdam Periconceptional cohort (Predict study), a tertiary hospital based, prospective observational study that was set up in 2009 to investigate maternal and paternal periconceptional health and the impact on reproduction, pregnancy and neonatal outcomes. (15, 16) At study entry, pregnant couples fill out an extensive online questionnaire including a study-specific 196-item food frequency questionnaire (FFQ) for information about parental dietary patterns. Data collected from November 2014 up until December 2020 was available for analysis. During this period, a total of 2637 female participants were included in the Predict study, of which 1043 did not submit an FFQ. We calculated the outcome events per predictor variable (EPV) to estimate adequacy of the sample size. (17)