Short sleep duration and hyperglycemia in pregnancy: Aggregate and individual patient data meta-analysis

Abstract

Short sleep duration has been linked to maternal hyperglycemia. Systematic review and meta-analysis were performed to evaluate the relationship between sleep duration and hyperglycemia in pregnancy or gestational diabetes (GDM). MEDLINE and Scopus were searched until July 2017. Studies that assessed sleep duration and had objective measurements of hyperglycemia during pregnancy were eligible. Aggregate data were available from eight studies, n=17,308 (seven with self-reported and one with objectively measured sleep duration). Meta-analysis was applied for pooling aggregate data using a random-effects model. Identified authors provided individual patient data (IPD) from four studies with objectively measured sleep duration, n=287. A one-stage approach with a hierarchical mixed-effect logit model was applied to pool IPD across studies. Aggregate data analysis revealed that women with short sleep duration (<6-7h) were more likely to have GDM than women without short sleep duration, odds ratio 1.70 (95% CI: 1.24, 2.33). IPD analysis demonstrated that, compared to sleeping >6.25h, women who slept ≤6.25h had higher 1-h glucose levels after 50-goral glucose tolerance testing by 0.65mmol/L (0.18, 1.13) and an increased risk of GDM, adjusted odds ratio 2.84 (1.25, 6.44). In conclusion, short sleep duration in pregnancy, both self-reported and objectively measured, is associated with hyperglycemia and an increased GDM risk.