Statistical analysis
Maternal TG concentrations and early-pregnancy BMI are presented as medians and quartile 1 and quartile 3 categorical variables as frequencies with proportions. Multilevel linear regression models were employed to assess associations between maternal baseline characteristics, early-pregnancy BMI, and maternal TG concentrations among the maternal early-pregnancy blood pressure groups. Logistic regression was used to estimate odds ratios with 95% CI for risk of preeclampsia in women with stage 1 and stage 2 hypertension, and elevated blood pressure compared with normal blood pressure for women with or without risk factors.
Multivariable analyses were adjusted for potential confounders. Maternal age (≤ 24, 25-29, 30-34, or ≥ 35 years), parity (nulliparous, parous), education (≤ 12, 13-17, or ≥ 18 years), birthplace of the mother (residents or immigrants) and pre-pregnancy BMI (≤ 18.4, 18.5-24.9, or ≥ 25 kg/m2) were included as covariates and to assess the risk factors for the association between maternal blood pressure in early pregnancy and the risks for preeclampsia. For the analysis of the impact of maternal obesity and high TG concentrations on the association of maternal blood pressure and the risks of preeclampsia, women were stratified based on the quartile values of BMI and TG concentrations.
The absolute risk difference compared to normal blood pressure was calculated as the difference between the risk of preeclampsia among women with one or two risk factors. To reduce the bias caused by fewer samples, women were stratified into BMI ≥ 25 and < 25 kg/m2 subgroups; and TG ≥ 90th and < 90th subgroups. The confidence interval was determined using standard statistical packages. (20) A confidence interval that contained a zero meant that there was no significant difference between the event and the control in terms of risk.
All statistical analyses were performed using SPSS package ver. 16.0 (SPSS Inc., Chicago, IL, USA). A two-tailed P -value < 0.05 was used as the threshold for statistical significance.