Metformin slows intestinal glucose absorption in type 2 diabetes, irrespective of the timing of its administration

1. INTRODUCTION

Metformin remains the first‐line pharmacotherapy for type 2 diabetes (T2D). In addition to suppressing hepatic glucose production and lowering fasting hyperglycaemia, metformin augments the secretion of the incretin hormone glucagon‐like peptide‐1 (GLP‐1) and reduces the glycaemic excursion after meals. ¹ , ² , ³ Although patients have been routinely advised to take metformin with meals, recent evidence has shown that the lowering of postprandial glycaemia by metformin is significantly enhanced when administered before meals. ⁴ This was linked, at least in part, to augmented GLP‐1 secretion, ⁴ although the underlying mechanisms remain uncertain. Previous studies from both our own group ⁵ , ⁶ and others ⁷ have demonstrated that the stimulation of GLP‐1 release by metformin occurs primarily during the postprandial phase, with minimal, if any, effect on fasting GLP‐1 levels. These observations suggest that the enhanced GLP‐1 response to meals following metformin treatment is mediated indirectly, probably in part driven by the inhibition of glucose absorption in the upper small intestine ⁸ resulting in enhanced glucose interaction with more distal small intestinal regions where GLP‐1‐secreting L‐cells are much more abundant. ⁹ To delineate the potential relevance of small intestinal glucose absorption to the effects of metformin on postprandial glucose handling, we have now retrospectively analysed a marker of intestinal glucose absorption in our recently reported study, where metformin was administered by intraduodenal infusion at different time intervals (i.e., 0, 30 or 60 min) before an intraduodenal glucose load in individuals with metformin‐treated T2D. ⁴

2. MATERIALS AND METHODS

2.1. Participants

As reported previously, ⁴ 16 participants with T2D managed by stable metformin monotherapy (2 female and 14 male, aged 69.9 $\pm$ 1.9 years, BMI 28.7 $\pm$ 1.0 kg/m², HbA1c 6.6 $\pm$ 0.1% / 48.2 $\pm$ 1.6 mmol/mol, known duration of diabetes 10.4 $\pm$ 2.6 years, daily metformin dose 843.8 $\pm$ 139.2 mg), were studied in a double‐blinded, randomised, placebo‐controlled, crossover design after providing written informed consent. The study was approved by the Central Adelaide Local Health Network Human Research Ethics Committee (ID: 2021/HRE00130) and prospectively registered (ACTRN12621000878875).

2.2. Protocol

On each study day, a silicone catheter was inserted into the duodenum via an anaesthetised nostril after an overnight fast and withholding the preceding dose of metformin. Participants received 1 g metformin dissolved in 50 mL 0.9% saline via the nasoduodenal catheter at t = −60, −30 or 0 min and saline at the other timepoints (or at all timepoints for control), followed by an intraduodenal infusion of a 180 mL solution containing 45 g glucose and 5 g 3‐O‐methylglucose (3‐OMG, a non‐metabolised glucose analogue as a marker of glucose absorption ¹⁰ ) from t = 0 to 60 min. ‘Arterialised’ venous blood was sampled every 30 minutes between t = −60 and 120 min for measurements of plasma glucose, total GLP‐1 and insulin concentrations, which were reported recently. ⁴ In the current study, serum samples at t = 60, 90 and 120 min were utilised for measurement of 3‐OMG concentrations by liquid chromatography and mass spectrometry. ¹⁰

2.3. Statistical analysis

Serum 3‐OMG concentrations were compared using two‐factor repeated‐measures analysis of variance (ANOVA) with treatment and time as factors and post hoc comparisons adjusted by Bonferroni's correction. The analyses were performed using Prism 10.4 software (GraphPad, USA). A p value <0.05 was considered statistically significant.

3. RESULTS

After intraduodenal glucose administration, serum 3‐OMG increased with peak levels at t = 90 min, followed by a modest decline at t = 120 min (time effect: p < 0.001 for all study days). There was a significant treatment effect (p = 0.012) as well as a treatment by time interaction (p = 0.008) for metformin on serum 3‐OMG concentrations, which were lower than control at t = 90 min on all metformin days (p < 0.05), and at t = 120 min when metformin was administered 30 min before the glucose infusion (p = 0.006). Furthermore, 3‐OMG concentrations at t = 120 min were lower when metformin was administered 30 min, compared to 60 min, before intraduodenal glucose (p = 0.014) but did not differ between any other treatments (Figure 1).

FIGURE 1.

The impact of the timing of metformin administration on intestinal glucose absorption. Effect of metformin (1 g) administered at −60, −30 or 0 min, or 0.9% saline administered at all timepoints (control) on serum 3‐O‐methyl‐glucose (3‐OMG) concentrations following an intraduodenal infusion of a glucose solution containing 45 g glucose and 5 g 3‐OMG between t = 0 min and t = 60 min in participants with type 2 diabetes receiving metformin monotherapy (n = 16). Statistical significance for differences between the four study days was determined by repeated‐measures ANOVA. Symbols denote p < 0.05 for * metformin (−60 min) versus control, # metformin (−30) versus control, † metformin (0) versus control, and θ metformin (−30) versus metformin (−60 min). Data are means ± SEM.

4. DISCUSSION

This study shows a modest effect of metformin to slow the absorption of a small intestinal glucose load in individuals with T2D managed by metformin monotherapy. This observation extends the findings in people with diet‐controlled T2D, ⁸ indicative of a prompt effect of metformin to modulate the activity of intestinal glucose transporters. However, in contrast to both the lowering of blood glucose and stimulation of GLP‐1 secretion, ⁴ the magnitude of inhibition in intestinal glucose absorption by metformin was not affected by the timing of its administration. A relationship between the rise in GLP‐1 secretion and the inhibition in small intestinal glucose absorption was evident in a previous study in people with diet‐controlled T2D. ⁸ However, the dissociation between inhibition of intestinal glucose absorption with augmentation of GLP‐1 secretion and lowering of blood glucose shown in the current analysis does argue against a major role for inhibition of intestinal glucose absorption in facilitating postprandial glucose disposal by metformin. Of note, metformin also inhibits the resorption of intestinal bile acids, which may augment GLP‐1 secretion via their interactions with L‐cells located more densely in the distal gut regions. ¹¹ , ¹² However, the extent to which this pathway may contribute to glucose‐lowering by metformin remains to be determined.

Several limitations of our study should be noted. First, our participants were predominantly male. While this was unintentional, a sex disparity in the rate of glucose absorption cannot be excluded. Second, in the interest of establishing ‘proof‐of‐concept’, metformin and glucose were administered intraduodenally. This design, while ‘unphysiological’, bypasses potential major confounders, such as variations in gastric emptying between individuals. The translational relevance of our findings is supported by comparable actions of metformin when administered orally. ⁸ Third, the effect of a single dose of immediate‐release metformin 1 g was utilised given established efficacy in attenuating hyperglycaemia and enhancing GLP‐1 secretion, ⁹ , ¹³ but this inevitably limited conclusions on metformin effects at other doses and their longer‐term impact on the gut microbiota. Fourth, due to the relatively brief observational window, serum 3‐OMG concentrations had not returned to baseline, such that the overall effect of metformin on total glucose absorption could not be determined.

In conclusion, our findings suggest that metformin reduces the rate of intestinal glucose absorption in a manner unrelated to the timing of its administration and that the enhanced postprandial glucose‐lowering and GLP‐1 release associated with pre‐meal metformin administration are unrelated to the rate of glucose absorption in T2D.

AUTHOR CONTRIBUTIONS

MJB contributed to the conception of this study, data collection and interpretation, statistical analysis and drafting of the manuscript. CX, WH, MB and JG supported data collection and interpretation, and review of the manuscript. CC coordinated the analysis for 3‐OMG concentrations, contributed to data interpretation and review of the manuscript. MH, KLJ and CK contributed to the study design, data interpretation and review of the manuscript. TW contributed to the conception and design of the study, data collection and interpretation, statistical analysis and manuscript writing. TW is the guarantor of this work. All authors approved the final version of the manuscript.

FUNDING INFORMATION

MJB is supported by both a University of Adelaide Research Training Scholarship and a Royal Adelaide Hospital Dawes Scholarship. Both CX and WH are supported by an Australian Diabetes Society Skip‐Martin Early Career Fellowship. CKR is supported by the Michell Bequest Foundation. TW has received funding support from Diabetes Australia on projects relating to metformin (IDs: Y17G‐WUTO, Y19G‐WUTO).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

PEER REVIEW

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.16570.

Borg MJ, Xie C, Chen C, et al. Metformin slows intestinal glucose absorption in type 2 diabetes, irrespective of the timing of its administration. Diabetes Obes Metab. 2025;27(9):5370‐5372. doi: 10.1111/dom.16570

DATA AVAILABILITY STATEMENT

The full datasets are available from the corresponding author upon reasonable request.