This was a secondary analysis of the Preterm SAMBA study, a multicenter cohort study performed in 5 different centers in Brazil. From July 2015 to March 2018, 1200 healthy nulliparous pregnant women were enroled and received follow-up during prenatal care, including only singleton pregnancies, without any fetal malformations or previous chronic maternal disease . Ethical approval for the study was obtained from relevant institutional review boards and competent authorities of each center where the study was conducted. More detailed information on the study design and methods used in this study have already been previously published .
Participants and procedures
Criteria for participant enrollment were nulliparous women with singleton pregnancies between 19 and 21 weeks of gestation. Exclusion criteria included: previous history of chronic hypertension, use of medication, fetal malformations, diabetes mellitus, nephropathy, autoimmune diseases (systemic erythematous lupus or antiphospholipid syndrome), sickle cell disease, uterine malformations, previous cervical surgery, previous cerclage, history of 3 or more abortions, HIV infection, chronic use of corticosteroids or aspirin or calcium above 1 g/day or fish oil above 2.7 g/day or vitamin C above 1000 mg/day or vitamin E above 400 UI/day or heparin. This criteria is in accordance with another cohort study published previously .
At least three routine hospital visits were scheduled. Systolic and diastolic blood pressure of the women were measured, according to standard clinical procedure on the 3 occasions: at 19–21 weeks, 27–29 weeks and 37–39 weeks of gestation, using a manual sphyngomanometer, calibrated according standard procedures and using the same model in all participating centres. In these 3 occasions information about proteinuria was obtained based on a regular urinalysis performed in the routine prenatal care (in the first trimester and in the third trimester).
During the first visit, maternal characteristics and medical history were recorded. In addition, blood and hair samples were collected and stored appropriately in a biobank for subsequent analysis by metabolomics technology. Gestational age was estimated from the date of the last menstrual period and confirmed by an early ultrasonography performed before 20 weeks. For each scheduled visit, blood pressure was measured 3 times. Women were allowed to rest for 15 min before the first blood pressure measurement was performed. Between blood pressure measurements, the investigator waited for at least 2 min. During the examination, participants remained in a sitting position, with their right arm supported at the level of their heart. An adult blood pressure cuff was used, selecting the proper size for each participant. Pressure reading at phase V of Korotkoff sounds corresponded to diastolic pressure. Mean arterial blood pressure was obtained by the equation (2DBP + SBP)/3. The mean blood pressure at each gestational age for the three measurements was obtained by the average of three mean blood pressure measurements [BPm = (BP1m + BP2m + BP3m)/3]. We also calculated the difference in mean blood pressure with measurements at 19–21 weeks and 27–29 weeks and measurements at 19–21 weeks and 37–39 weeks. Calculation was made in two steps: first, the difference was determined for each woman; and second, the mean difference was calculated.
Preeclampsia was the main outcome of this analysis. It was defined as the onset of hypertension (systolic blood pressure of 140 mmHg or more and/or diastolic blood pressure of 90 mmHg or more) after 20 weeks of gestation, measured on at least two different occasions, in conjunction with proteinuria (≥300 mg/day or at least 1 g/L [1+] on dipstick testing or spot urine protein/creatinine ≥30 mg/mmol [0.3 mg/mg]) or any signs of organ dysfunction . Systemic complications were defined as: hematological complications (thrombocytopenia, disseminated intravascular coagulation or hemolysis); hepatic dysfunction (elevated transaminases); neurological dysfunction (examples include eclampsia, altered mental status, blindness, stroke or more commonly hyperreflexia when accompanied by clonus, severe headache when accompanied by hyperreflexia, persistent visual scotomata); renal dysfunction (creatinine ≥1.2 mg/dL) .
After delivery, each woman was classified as having a normal pregnancy (control group) or preeclampsia (case group). Cases were categorized into early-onset preeclampsia (women who developed preeclampsia before 34 weeks of gestation) and late-onset preeclampsia (women who developed preeclampsia after 34 + 1 weeks of gestation)  .
Initially, the three groups were compared regarding sociodemographic characteristics of women using a Chi-square design-based test. Mean arterial blood pressure was then compared among the three groups (early-onset preeclampsia, late-onset preeclampsia and normotensive) using Student’s t-test. The mean difference in MBP measured at 27 and 37 weeks was estimated and compared to values at 20 weeks. Finally, we checked to see whether mean arterial blood pressure had any predictive power at three time periods (20, 27 or 37 weeks of gestation) by comparing the area under the receiver-operating characteristic curves (AUROC). Analyses were performed using SPSS and Stata software.