Early continuous glucose monitoring-derived glycemic patterns are associated with subsequent insulin resistance and gestational diabetes mellitus development during pregnancy

Abstract

Background: Gestational diabetes mellitus (GDM) and insulin resistance (IR) increase the risk of adverse pregnancy outcomes. We aimed to examine the relationship of interstitial glucose assessed by continuous glucose monitoring (CGM) at early gestation, and the subsequent development of IR and GDM, and to determine 24-h interstitial glucose centile distributions in women with normal (non-IR and non-GDM) and suboptimal glycemic status (IR and/or GDM).

Methods: CGM measurements were taken for 3-10 days at 18-24 weeks' gestation, followed by fasting serum insulin and oral glucose tolerance testing at 24-28 weeks' gestation. IR and GDM were determined by the updated Homeostasis Model Assessment of IR score of ≥ 1.22 and 2013 World Health Organization criteria, respectively. Risks of IR and GDM were estimated using modified Poisson models, and hourly interstitial glucose centiles determined using Generalized Additive Models for Location, Scale and Shape.

Results: This prospective cohort study involved 167 pregnant women in Singapore, with a mean age of 31.7 years, body mass index of 22.9 kg/m², and gestation of 20.3 weeks. 25% of women exhibited IR and 18% developed GDM. After confounders adjustment, women with suboptimal glycemic control, indicated by higher mean daily glucose (risk ratio 1.42; 95% confidence interval 1.16, 1.73), glucose management indicator (1.08; 1.03, 1.12), and J-index (1.04; 1.02, 1.06), as well as those with greater glycemic variability, indicated by higher standard deviation (1.69; 1.37, 2.09), coefficient of variation (1.03; 1.00, 1.06), and mean amplitude of glycemic excursions (1.4; 1.14, 1.35) derived from CGM in early gestation were associated with higher risks of developing IR in later gestation. These associations were similarly observed for the development of GDM. Centile curves showed that, compared to those with normal glycemic status, women with suboptimal glycemic status had higher glucose levels, with greater fluctuations throughout 24 h.

Conclusions: In pregnant women who subsequently developed IR and GDM, interstitial glucose levels assessed by CGM were elevated and varied greatly. This supports the potential use of CGM to screen for glycemic changes early in pregnancy.

Keywords: Continuous glucose monitoring; Gestational diabetes mellitus; Glycemic control/variability; Insulin resistance.

Fig. 1

The predicted 24-h interstitial glucose levels for women by (a) IR (GMR 1.12; 95% CI 1.05, 1.19) and (b) GDM status (1.11; 1.03, 1.20) based on the GEE analysis. Red represents the women with IR or GDM and blue represents the women with non-IR or non-GDM. The circle markers and capped vertical lines represent the predicted mean daily glucose levels and the respective 95% CI based on the exponentiated log-transformed hourly median glucose values. Models were adjusted for age, ethnicity, years of education, parity, history of GDM or family history of diabetes, pre-pregnancy body mass index, irregular meal, physical activity, and an interaction term between glycemic status and time. CI confidence intervals, GDM gestational diabetes mellitus based on 2013 World Health Organization criteria, GEE generalized estimating equations, GMR geometrical mean ratio, IR insulin resistance based on HOMA2-IR of at least 1.22

Fig. 2

The centile curves of 24-h interstitial glucose levels from Box-Cox transformation among pregnant women with (a) non-IR and non-GDM, and (b) IR and/or GDM. A line across the 95th centile for pregnant women without IR and GDM corresponds to the 85th centile among those with IR and/or GDM, which suggests that those above the 95th centile are three times more likely to have IR and/or GDM. GDM, gestational diabetes mellitus based on 2013 World Health Organization criteria; IR, insulin resistance based on HOMA2-IR of at least 1.22