Vitamin D and gestational diabetes mellitus: a systematic review based on data free of Hawthorne effect

BackgroundGestational diabetes mellitus (GDM) is an increasingly prevalent disorder, associated with low blood vitamin D level. ObjectivesTo evaluate the relationship between vitamin D and GDM. Search strategyEMBASE, MEDLINE, Cochrane Library and China Biology Medicine disc were searched up to May 2017. The references of previous studies were screened. Selection criteriaObservational studies on the relationship between vitamin D and GDM free from Hawthorne effect and randomised controlled trials of vitamin D supplementation during pregnancy for preventing or treating GDM were included. Data collection and analysisData and information of included articles were extracted by duplicate using piloted tables. Newcastle-Ottawa Scale and Cochrane Handbook were used for quality assessment. Random-effects models were used for meta-analyses. Heterogeneity tests, sensitivity analysis and analysis of publication bias were conducted. Main resultsEighty-seven observational studies and 25 randomised controlled trials involving 55859 and 2445 women, respectively, were included. Low blood vitamin D level during pregnancy was associated with a higher risk of GDM (OR 1.850, 95% CI 1.471-2.328). Blood vitamin D level for women with GDM were lower than in the control women. Blood vitamin D level was associated with fasting plasma glucose (FPG) and homeostasis model of assessment for insulin resistance index (HOMA-IR) (r=-0.100 and r=-0.351), whereas the correlation between blood vitamin D level and fasting insulin (FINS) might be concealed by publication bias. Vitamin D intervention during pregnancy could change the blood levels of vitamin D, FINS, FPG, HOMA-IR, glutathione, C-reactive protein and lipid. ConclusionsLow blood vitamin D level could increase the risk of GDM, and vitamin D supplementation during pregnancy could ameliorate the condition of GDM.