Gestational diabetes mellitus and adverse pregnancy outcomes: systematic review and meta-analysis
Wenrui Ye,Cong Luo,Jing Huang,Chenglong Li,Zhixiong Liu,Fangkun Liu
DOI: https://doi.org/10.1136/bmj-2021-067946
2022-05-25
BMJ
Abstract:Abstract Objective To investigate the association between gestational diabetes mellitus and adverse outcomes of pregnancy after adjustment for at least minimal confounding factors. Design Systematic review and meta-analysis. Data sources Web of Science, PubMed, Medline, and Cochrane Database of Systematic Reviews, from 1 January 1990 to 1 November 2021. Review methods Cohort studies and control arms of trials reporting complications of pregnancy in women with gestational diabetes mellitus were eligible for inclusion. Based on the use of insulin, studies were divided into three subgroups: no insulin use (patients never used insulin during the course of the disease), insulin use (different proportions of patients were treated with insulin), and insulin use not reported. Subgroup analyses were performed based on the status of the country (developed or developing), quality of the study, diagnostic criteria, and screening method. Meta-regression models were applied based on the proportion of patients who had received insulin. Results 156 studies with 7 506 061 pregnancies were included, and 50 (32.1%) showed a low or medium risk of bias. In studies with no insulin use, when adjusted for confounders, women with gestational diabetes mellitus had increased odds of caesarean section (odds ratio 1.16, 95% confidence interval 1.03 to 1.32), preterm delivery (1.51, 1.26 to 1.80), low one minute Apgar score (1.43, 1.01 to 2.03), macrosomia (1.70, 1.23 to 2.36), and infant born large for gestational age (1.57, 1.25 to 1.97). In studies with insulin use, when adjusted for confounders, the odds of having an infant large for gestational age (odds ratio 1.61, 1.09 to 2.37), or with respiratory distress syndrome (1.57, 1.19 to 2.08) or neonatal jaundice (1.28, 1.02 to 1.62), or requiring admission to the neonatal intensive care unit (2.29, 1.59 to 3.31), were higher in women with gestational diabetes mellitus than in those without diabetes. No clear evidence was found for differences in the odds of instrumental delivery, shoulder dystocia, postpartum haemorrhage, stillbirth, neonatal death, low five minute Apgar score, low birth weight, and small for gestational age between women with and without gestational diabetes mellitus after adjusting for confounders. Country status, adjustment for body mass index, and screening methods significantly contributed to heterogeneity between studies for several adverse outcomes of pregnancy. Conclusions When adjusted for confounders, gestational diabetes mellitus was significantly associated with pregnancy complications. The findings contribute to a more comprehensive understanding of the adverse outcomes of pregnancy related to gestational diabetes mellitus. Future primary studies should routinely consider adjusting for a more complete set of prognostic factors. Review registration PROSPERO CRD42021265837.
medicine, general & internal