Random-slope, random intercept linear mixed-effects modeling showed that malaria infection at Visit 1 (13–23 weeks) was associated with changes in inflammatory and angiogenic mediators across pregnancy.

Fixed effects included a spline of gestational age, treatment group, maternal age, gravidity, socioeconomic status, educational status, BMI, and hemoglobin at Visit 1. Malaria status at Visit 1 (13–23 weeks) was added to the model, and an interaction term for gestational age and malaria status was added to a third model; then the models were compared (χ2, p value). A restricted cubic spline of gestational age was used as both main effect and in interaction terms. Extended tables including estimates for all fixed effects and parameters of model fit (AIC, BIC) are available in the supporting information (S4–S6 Tables).