Fig. 4
Association between severity of nausea and vomiting of pregnancy symptoms and birth outcomes. Legends: Linear regressions were used for continuous outcomes (birth weight, grams; birth weight, SD; placental weight, grams), and logistic regression was used for a categorical outcome (SGA risk, odds ratio). All analyses were adjusted for maternal age, height, pre-pregnancy body mass index, household income, education, smoking status, and infant sex. Multiple imputation was used to impute the following missing values: weight at 7–14 weeks (n = 10,840; 11.9% of the sample), measurement timing at 7–14 weeks (n = 9,752; 10.7%), weight at 20–28 weeks (n = 9,189; 10.1%), measurement timing at 20–28 weeks (n = 9,048; 9.9%), weight at delivery (n = 1,801; 2.0%), and placental weight (n = 3,562; 3.9%). Weight change in the 1st trimester was calculated from pre-pregnancy weight and weight at the 1st trimester visit (at 7–14 weeks). Weight gain up to the 2nd trimester was calculated from pre-pregnancy weight and weight at the 2nd trimester visit (at 20–28 weeks). SD, standard deviation; SGA, small for gestational age