Incretin hypersecretion in gestational diabetes mellitus

Objective

The aim of our study was to investigate the response of incretins and glucagon during an oral glucose tolerance test in pregnancy in a large cohort of well phenotyped women with normal glucose tolerance and GDM using precise pre-analystics and specific immunoassays.

Subjects

We analyzed data from an ongoing study aiming to characterize metabolic alterations during pregnancy (PREG), a cohort study recruiting women undergoing oral glucose tolerance tests for screening of GDM (NCT04270578). Pregnant women were examined between gestational week 24+0 and 31+6 with a 2h-OGTT with 75 g glucose. GDM was diagnosed using the IADPSG criteria ⁹ and patients were subsequently treated according to national guidelines, which was not part of the study. The detailed study protocol has been described elsewhere ¹⁰. Informed written consent was obtained from every study participant. The study protocol has been approved by the local ethics boards and the study is conducted in accordance with the declaration of Helsinki. Patients were not involved in the design of the study. Meetings to connect study participants and researchers are held annually.

OGTT, laboratory analyses and anthropometric assessment

After an overnight fasting period of 12h, all study participants underwent 5-point OGTT with 75 g of glucose. Venous blood was collected at fasting and after 30, 60, 90 and 120 minutes. Plasma glucose and nonesterified fatty acids were measured from sodium-fluoride plasma in an ADVIA chemistry XPT autoanalyzer (Siemens Healthcare Diagnostics) at all timepoints. Serum insulin and C-peptide were analyzed using ADVIA Centaur XPT immunoassay system (Siemens AG) at all timepoints. For the measurement of total GIP, total GLP-1, glicentin and glucagon EDTA-plasma from timepoint 0, 30 and 120 minutes was stabilized with 300 ng/mL of the protease inhibitor aprotinin (Sigma, Merck, Germany) und subsequently processed at 4°C and kept frozen at −80°C until batch measurement. Incretins and glucagon were measured with commercially available ELISA assays (Mercodia, Sweden) according to the manufacturer’s instructions. Height and weight were measured at the day of OGTT. Pregestational weight, parity, gestational age at birth and birthweight were obtained from maternal medical logs. Birth outcome data were available from 135 participants. Insulin sensitivity was calculated using the NEFA-ISI index ¹¹. Insulin secretion was assessed with and with ΔC-peptide_0-30.

Statistical analysis

All statistical analyses were performed with R Version 3.6.1. In linear regression models skewed variables were log-transformed to approximate normal distribution. Group differences were tested with t-tests for normal distributed variables, Kruskal-Wallis-test for non-normal distributed variables and Chi²-test for categorical variables. The provided β coefficients are standardized estimates of the linear regression model terms. A p-value <0.05 was considered statistically significant.

Glucose and hormones during OGTT

Insulin secretion and sensitivity were lower in GDM (Figure1 A, B). Areas under the curve (AUCs) of glucose and C-peptide were higher in GDM (Figure 1 C, D). While fasting levels of GLP-1 and GIP were similar, the levels at 30min were higher in GDM (Table 2). The AUCs of GLP-1 and GIP were also higher in GDM (all P<0.05, Figure 1 E, F, Table 2). Fasting glicentin, glucagon, and their post-load kinetics were not different between groups (p ≥ 0.1, Figure 1 G, H, Table 2).

• Download figure
• Open in new tab

Figure 1.

Insulin secretion (A) and insulin sensitivity (B) in women with and without GDM and time course of glucose (C), C-peptide (D), GLP-1 (E), GIP (F), glicentin (G) and glucagon (H) during OGTT. Data are presented as mean ± standard error. Blue represent NGT, red GDM. Group differences between GDM and NGT for AUCs of glucose, insulin and C-peptide were tested with Kruskal-Wallis-test. Differences between groups for insulin sensitivity, incretins and glucagon were tested with multivariate linear regression adjusted for age, BMI and gestational week and for insulin secretion additionally adjusted for insulin sensitivity.

GLP-1 associates with insulin secretion only in women with GDM

We hypothesized that higher incretin levels in GDM represent a compensatory effect to boost insulin secretion. Therefore, we analyzed the associations of GLP-1 and glucose with insulin secretion (Figure S1). In NGT, increase in glucose between 0-30min associated with insulin secretion (Figure S1 A, P<0.0001). No such association was present for GLP-1 (panel B, p=0.43).

In contrast, women with GDM showed no association of glucose (Figure S1 C, p=0.6) but a positive association of GLP-1 with insulin secretion (panel D, p=0.0004). This remained significant after adjustment for glucose change and basal insulin. Glicentin was similarly associated with insulin secretion in GDM and NGT (Figure S2). For GIP there was no association with insulin secretion (Figure S3).

Higher GLP-1 associates with lower birth weight

Birth weight negatively associated with AUC GLP-1 (p=0.0182, β=-76.94), 30min GLP-1 (p=0.0248, β=-72.6) and trend for negative association with 30min GIP (p=0.0814, β=-57.45, all models adjusted for gestational age at birth, fetal sex and pregestational BMI). Since GLP-1 and GIP concentrations were higher in GDM, the models were additionally adjusted for GDM. Associations of birth weight with AUC GLP-1 and 30min GLP-1 remained significant (p=0.0296, β=-72.43 and p=0.0381, β=-68.49, respectively).

Main findings

In this study, we find significantly higher postprandial GLP-1 and GIP levels in pregnant women with gestational diabetes compared to women with normal glucose tolerance, independent from age, body mass index and gestational age. The postprandial GLP-1 increase associated with insulin secretion in women with GDM but not in women with NGT, where insulin secretion was glucose driven. Offspring of women with higher GLP-1 levels also had lower birth weights.

Interpretation

GLP-1 and GIP have previously been investigated in women with and without GDM in smaller studies with different results ^4,6,7,12. One study used a different stimulus (mixed meal test) ⁶ and large BMI differences ^6,12 between NGT and GDM subjects might further confound analyses, because BMI negatively associates with GLP-1 and GIP ¹³. Similar BMI between groups allowed analyzing impact of GDM on incretin responses, without adiposity as confounder.

Intestinal L-cells also secrete glicentin, a potential biomarker of L-cell secretion ¹⁴. Lack of difference in glicentin between NGT and GDM argues against an unselective L-cell hypersecretion in GDM.

In contrast to findings on glucagon outside of pregnancy ¹⁵, we and others ⁶ did not detect links to diabetes. This argues against a major contribution of glucagon in the pathogenesis of GDM, while it might have some value in the prognosis of subsequent insulin requirement for the treatment of GDM ¹⁶.

Of note, women with GDM had lower insulin secretion despite higher GLP-1 concentrations. Our correlational analyses indicate an important contribution of GLP-1 to insulin secretion, especially in GDM. However, this GLP-1 stimulus is still not sufficient to control hyperglycemia. The failure of adequate insulin secretion despite elevated GLP-1 indicates incretin resistance in GDM. The more pronounced GLP-1 response in GDM could be counteracting incretin resistance.

One potential contributor to incretin resistance is genetic background. Polymorphisms in TCF7L2 associate with elevated GDM risk ¹⁷ and reduced GLP-1-stimulated insulin secretion ¹⁸. Another possible contributor might be dipeptidyl peptidase IV (DPP-4) activity, the enzyme degrading GLP-1 and GIP. Liu et al. ¹⁹ reported no differences of DPP-4 in maternal serum and the unaltered fasting GLP-1 levels in GDM in our study argue against largely different DPP-4 activity. Furthermore, insulin resistance and hyperglycemia which are present in GDM are associated with incretin resistance ¹⁸.

In our study neonates of women with higher GLP-1 levels during pregnancy had lower birth weight, independent of glycemia. This points towards a protection of GLP-1 against excessive fetal growth. Our findings are in line with a report of negative association between fasting maternal GLP-1 and fetal abdominal circumference and birth weight ²⁰. Higher GLP-1, also via decelerated gastric emptying, might reduce postprandial glucose which is associated with fetal overgrowth. If this hypothesis holds true, GLP-1 receptor agonists or DPP-4 inhibitors might have benefits in GDM. To our knowledge, currently no studies test GLP-1R agonists in GDM. One trial reported higher insulin secretion after 16 weeks treatment with the DPP-4 inhibitor sitagliptin in GDM ²¹. However, neither incretins nor birth outcomes were reported.

Strengths and limitations

The strength of our study is the large number of subjects of our extensively phenotyped cohort of pregnant women. The sample preparation and measurement of incretins was done according to standard operating procedures and with state-of-the-art measuring method. One weakness is the relatively low number of patients with GDM.