We investigated the clinical and laboratory characteristics of pregnancies with systemic lupus erythematosus (SLE) and identified lupus flare predictors during pregnancy. Additionally, we examined lupus activity and pregnancy outcomes in SLE patients who continued, discontinued or underwent no hydroxychloroquine (HCQ) treatment during pregnancy. We retrospectively analyzed 179 pregnancies in 128 SLE patients at Seoul St. Mary’s Hospital, Korea, between 1998 and 2012 and then assessed the clinical profiles and maternal and fetal outcomes. Overall, 90.5% of pregnancies resulted in a successful delivery and were divided into two groups: those who experienced lupus flares (80 pregnancies, 44.7%) and those who did not (99 pregnancies, 55.3%). Increased preeclampsia, preterm births, low birth weight, intrauterine growth restriction (IUGR), and low 1-minute Apgar scores occurred in pregnancies with lupus flares compared to pregnancies in quiescent disease. Lupus flares were predicted by HCQ discontinuation, a history of lupus nephritis, high pre-pregnancy serum uric acid and low C4 levels. Our study indicates that achieving pre-pregnancy remission and continuing HCQ treatment during pregnancy are important for preventing lupus flares.

This study was undertaken to investigate the pregnancy outcomes in patients with systemic lupus erythematosus (SLE) and the appropriate timing of pregnancy. We performed a retrospective evaluation of 183 pregnancies with SLE at Catholic University Medical Center during the 13-year period from 1998 to 2010. Pregnancy outcomes were compared according to SLE characteristics. The predictive value of the different cut-off points of the stable period before conception on adverse pregnancy outcomes was calculated by ROC (Receiver operating characteristics) curve analysis. In multivariate analysis, the presence of antiphospholipid antibodies (aPLs) increased the risk of pregnancy loss (p<0.0001) and premature birth (p=0.0040). Active disease at conception increased the risk of premature birth (p< 0.0001) and complications (IUGR, PIH, or both) (p= 0.0078). The other predictor of complications was found to be lupus flare (p=0.0252). At a cut-off level of stable period of 4 months before conception, sensitivity and specificity were 70.8% and 53.2%, 71.4% and 61.5%, and 63.6 % and 59.8 %, respectively on reducing pregnancy loss, premature birth, and complications. Pregnancies with aPLs, active disease at conception and SLE flares are at a higher risk of adverse outcomes. It is essential that disease activity remains stable at least 4 months before conception, for favorable pregnancy outcomes.

This study developed a machine learning algorithm to predict gestational diabetes mellitus (GDM) using retrospective data from 34,387 pregnancies in multi-centers of South Korea. Variables were collected at baseline, E0 (until 10 weeks’ gestation), E1 (11–13 weeks’ gestation) and M1 (14–24 weeks’ gestation). The data set was randomly divided into training and test sets (7:3 ratio) to compare the performances of light gradient boosting machine (LGBM) and extreme gradient boosting (XGBoost) algorithms, with a full set of variables (original). A prediction model with the whole cohort achieved area under the receiver operating characteristics curve (AUC) and area under the precision-recall curve (AUPR) values of 0.711 and 0.246 at baseline, 0.720 and 0.256 at E0, 0.721 and 0.262 at E1, and 0.804 and 0.442 at M1, respectively. Then comparison of three models with different variable sets were performed: [a] variables from clinical guidelines; [b] selected variables from Shapley additive explanations (SHAP) values; and [c] Boruta algorithms. Based on model [c] with the least variables and similar or better performance than the other models, simple questionnaires were developed. The combined use of maternal factors and laboratory data could effectively predict individual risk of GDM using a machine learning model.

The World Health Organization (WHO) international body mass index (BMI) cut-off points defining pre-pregnancy BMI categories in the Institute of Medicine (IOM) guidelines are not directly applicable to Asians. We aimed to define the optimal gestational weight gain (GWG) for the Korean population based on Asia-specific BMI categories. Data from 2702 live singleton deliveries in three tertiary centers between 2010 and 2011 were analyzed retrospectively. A multivariable logistic regression analysis was conducted to determine the lowest aggregated risk of composite perinatal outcomes based on Asia-specific BMI categories. The perinatal outcomes included gestational hypertensive disorder, emergency cesarean section, and fetal size for gestational age. In each BMI category, the GWG value corresponding to the lowest aggregated risk was defined as the optimal GWG. Among the study population, 440 (16.3%) were underweight (BMI < 18.5), 1459 (54.0%) were normal weight (18.5 ≤ BMI < 23), 392 (14.5%) were overweight (23 ≤ BMI < 25) and 411 (15.2%) were obese (BMI ≥ 25). The optimal GWG by Asia-specific BMI category was 20.8 kg (range, 16.7 to 24.7) for underweight, 16.6 kg (11.5 to 21.5) for normal weight, 13.1 kg (8.0 to 17.7) for overweight, and 14.4 kg (7.5 to 21.9) for obese. Considerably higher and wider optimal GWG ranges than recommended by IOM are found in our study in order to avoid adverse perinatal outcomes. Revised IOM recommendations for GWG could be considered for Korean women according to Asian BMI categories. Further prospective studies are needed in order to determine the optimal GWG for the Korean population.