Adriëtte Oostvogels

40 Chapter 2 Descriptive analyses were performed with the SPSS package version 21.0 (SPPS Inc., Chicago IL). The statistical package R 2.15.3 with MLWIN plug-in was used to model BP and analyse the associations between prepregnancy weight status, early pregnancy lipid profile and BP course during pregnancy. Statistical significance was set at p<0.05. Results Study population The study population was divided into prepregnancy normal weight (<25 kg/m 2 , n=2500; 80.6%) and overweight (>25 kg/m 2 , n=600; 19.4%). Overweight women were less educated, more often multiparous, more often of non-Dutch ethnicity and multiparous, drank less alcohol during pregnancy, and had a more atherogenic lipid profile in early pregnancy (Table 1). Of the women with overweight, 118 had obesity (19.7%). Our subsample was somewhat older, leaner, higher educated, more often of Dutch origin, more often nulliparous and consumed more alcohol during pregnancy than women in the ABCD study who were not included in the present study because no early pregnancy lipid profile or ≤3 valid BP measurements were available (n=4061; Figure 1 and S1 Table). For each analysis, women were divided into tertiles based on their lipid levels. Cut-off levels for each lipid are presented in Table 2. Blood pressure course during pregnancy of women with normal weight For women with normal weight BP course was as follows: starting with an average SBP of 111.3 (SD=11.1) mmHg and an average DBP of 65.6 (SD=8.2) mmHg their BP decreased in the first period (10-16.5 weeks) by 0.2 (SBP) and 0.3 (DBP) mmHg/week. During the second (16.5-31.5 weeks), and BP increased by 0.2 (both SBP and DBP) mmHg/week, third (31.5-35.5 weeks) 0.5 (SBP) and 0.8 (DBP) mmHg/week and fourth period (35.5-42 weeks) 0.9 (SBP) and 1.0 (DBP) mmHg/week.(Figure 2).