The gestational weight gain and perinatal outcomes among underweight women with twin pregnancies: Propensity score matched analysis from a three-year retrospective cohort

Objective No recommendations are available for gestational weight gain (GWG) in underweight women with twin pregnancies. We aimed to evaluate whether underweight women with twin pregnancies should gain more weight than normal-weight women in order to optimize perinatal outcomes. Study design This retrospective cohort study compared the GWG and perinatal outcomes among normal-weight and underweight women who gave birth to viable twins between 2015 and 2018 at the Maternal and Child Health Hospital in Foshan, China. Gestational weight gain (GWG) was categorized as adequate or inadequate GWG, based on the US Institute of Medicine 2009 guidelines for normal-weight women (≥ 0.46 kg/week). The outcomes of interest included spontaneous preterm birth (sPTB) <37 weeks, <35 and <32 weeks, small for gestational age (SGA), gestational hypertensive disorder (GHD), gestational diabetes mellitus (GDM), birth weight discordance (BDW) ≥20%, neonatal intensive unit (NICU) admission and neonatal respiratory distress syndrome (NRDS). Propensity score matching (PSM, in a 1:1 ratio) was utilized to minimize the effects of confounders on the differences in the two cohorts. Multivariable logistic models were also used to verify the results from PSM analysis. Results There were 475 normal-weight and 111 underweight women included in the analysis. Our results suggested that the incidence of adequate GWG was comparable between underweight and normal-weight women (37.5% vs. 45.1%, P = 0.141). The prevalence of GDM was significantly lower among underweight women (9.9%) than among normal-weight women (20.4%) (P = 0.010). There was no evidence of differences in other perinatal outcomes between the two groups. 102 underweight women and 102 normal-weight women were included in PSM analyses. There was a lower incidence of GDM in underweight women than in normal-weight women, but the difference was not significant (9.8% vs. 18.6%, P = 0.071). No evidence of any differences in the other outcomes, including sPTB, GHD, BWD≥20%, SGA, NICU admission and NRDS, was found between the underweight and normal-weight women. Multivariable logistic regression models yielded similar results. Conclusions For Chinese twin pregnant women with twin pregnancies, our data does provide evidence to suggest underweight women need to gain more weight than normal-weight women to optimize perinatal outcomes. Future studies with larger number of underweight women with twin gestations are warranted to establish an optimal range of GWG.