Associations of maternal early-pregnancy blood glucose and insulin concentrations with DNA methylation in newborns

BACKGROUND

Intrauterine exposure to a disturbed maternal glucose metabolism is associated with adverse offspring outcomes. DNA methylation is a potential mechanism underlying these associations. We examined whether maternal early-pregnancy glucose and insulin concentrations are associated with newborn DNA methylation. In a population-based prospective cohort study among 935 pregnant women, maternal plasma concentrations of non-fasting glucose and insulin were measured at a median of 13.1 weeks of gestation (95% range 9.4-17.4). DNA methylation was measured using the Infinium HumanMethylation450 BeadChip (Ilumina). We analyzed associations of maternal early-pregnancy glucose and insulin concentrations with single-CpG DNA methylation using robust linear regression models. Differentially methylated regions were analyzed using the dmrff package in R. We stratified the analyses on normal weight versus overweight or obese women. We also performed a look-up of CpGs and differently methylated regions from previous studies to be associated with maternal gestational diabetes, hyperglycemia or hyperinsulinemia, or with type 2 diabetes in adults.

RESULTS

Maternal early-pregnancy glucose and insulin concentrations were not associated with DNA methylation at single CpGs nor with differentially methylated regions in the total group. In analyses stratified on maternal BMI, maternal early-pregnancy glucose concentrations were associated with DNA methylation at one CpG (cg03617420, XKR6) among normal weight women and at another (cg12081946, IL17D) among overweight or obese women. No stratum-specific associations were found for maternal early-pregnancy insulin concentrations. The two CpGs were not associated with birth weight or childhood glycemic measures (p values > 0.1). Maternal early-pregnancy insulin concentrations were associated with one CpG known to be related to adult type 2 diabetes. Enrichment among nominally significant findings in our maternal early-pregnancy glucose concentrations was found for CpGs identified in a previous study on adult type 2 diabetes.

CONCLUSIONS

Maternal early-pregnancy glucose concentrations, but not insulin concentrations, were associated with DNA methylation at one CpG each in the subgroups of normal weight and of overweight or obese women. No associations were present in the full group. The role of these CpGs in mechanisms underlying offspring health outcomes needs further study. Future studies should replicate our results in larger samples with early-pregnancy information on maternal fasting glucose metabolism.

Retinal blood flow is increased in type 1 diabetes mellitus patients with advanced stages of retinopathy

BACKGROUND

Diabetic retinopathy (DRP) is a common microvascular complication seen in patients with type 1 diabetes mellitus (T1DM). The effects of T1DM and concomitant (proliferative) DRP on retinal blood flow are currently unclear. Therefore, we measured retinal vascular blood flow in T1DM patients with and without DRP and non-diabetic controls. We further assessed the acute effects of panretinal photocoagulation on retinal microvascular bloodflow in eight patients with diabetes.

METHODS

Thirty-three T1DM patients with proliferative DRP, previously treated with panretinal photocoagulation (pDRP), 11 T1DM patients with untreated non-proliferative retinopathy (npDRP) and 32 T1DM patients without DRP (nDRP) were compared with 44 non-diabetic gender-matched controls. Using scanning laser Doppler flowmetry (HRF, Heidelberg) blood flow in the retinal microvasculature was measured temporal and nasal of the optic disc and averaged into one flow value per eye. The right eye was used as a default for further analyses. Eight patients with novel proliferative retinopathy (4 T1DM and 4 with type 2 diabetes) were measured before and several months after photocoagulation. Between-group differences in retinal blood flow were assessed using ANOVA corrected for multiple comparisons (Bonferroni).

RESULTS

Retinal blood flow was higher in the treated pDRP compared with the nDRP group and controls (all P Bonferroni < 0.01). Furthermore, there was a positive linear trend for blood flow with lowest blood flow in the control group and highest in the pDRP group (P-for-trend < 0.01). In the eight patients with novel proliferative retinopathy, blood flow did not significantly change before and after panretinal photocoagulation (P > 0.05). Using regression analysis, no variables were found as predictors of retinal blood flow.

CONCLUSIONS

In comparison with controls and nDRP patients, retinal blood flow significantly increased in the pDRP group, which previously underwent photocoagulation treatment, but not in the npDRP patients. These changes may be a consequence of a failing vascular autoregulation in advanced diabetic retinopathy.