Metformin, a cornerstone medication for type 2 diabetes, primarily lowers blood glucose by suppressing hepatic glucose production and through various gastrointestinal actions. While the hepatic effects are well-established, a significant portion of metformin's glucose-lowering impact occurs within the gastrointestinal tract, regardless of systemic bioavailability. Enteral metformin administration is demonstrably more effective in reducing glucose levels than intravenous or intraportal routes. Moreover, delayed-release metformin formulations, which minimize systemic exposure, prove equally effective as immediate- or extended-release versions. Research has illuminated several mechanisms through which metformin modulates gastrointestinal function: incretin hormone GLP-1 stimulation, slowing gastric emptying, suppression of intestinal glucose absorption, inhibition of bile acid resorption, and alteration of the gut microbiota. Notably, GLP-1 plays a crucial role in mediating metformin's glucose-lowering action, as evidenced by studies showing that GLP-1 blockade significantly diminishes metformin's effectiveness. GLP-1's pleiotropic actions include insulin secretion stimulation, glucagon release suppression, gastric emptying slowing, and appetite inhibition. Given that strategies to lower postprandial glycemia through boosted GLP-1 secretion are most effective when administered pre-meal, the researchers hypothesized that the timing of metformin administration relative to food intake significantly impacts its efficacy. This is particularly relevant as the standard recommendation to take metformin with meals lacks robust evidence-based support. A previous pilot study hinted at improved glucose control, enhanced GLP-1 secretion, and slowed gastric emptying when metformin was taken 30 minutes prior to a meal. This current study aims to rigorously investigate the effects of varying metformin administration timing on glucose control, GLP-1 and insulin responses in individuals with type 2 diabetes already receiving metformin.

Existing research indicates metformin's ability to reduce postprandial glycemic excursions in type 2 diabetes patients by influencing gastrointestinal function and GLP-1 release. However, the impact of varying metformin administration timing on postprandial glucose metabolism remains poorly understood. Studies have shown that enteral administration of metformin is more effective in lowering glucose than intravenous or intraportal routes, highlighting the importance of gastrointestinal effects. Preclinical and clinical findings suggest metformin's multifaceted gastrointestinal impact includes stimulation of GLP-1, slowed gastric emptying, suppressed intestinal glucose absorption, and modulation of gut microbiota. Evidence strongly suggests that GLP-1 plays a significant role in metformin's glucose-lowering action, with studies demonstrating that GLP-1 signaling blockade significantly diminishes metformin's efficacy. Strategies aiming to reduce postprandial glycemia by enhancing GLP-1 secretion often prove more effective when administered prior to a meal, suggesting a potential impact of timing with metformin administration. However, current clinical practice recommends taking metformin with meals, primarily to minimize potential gastrointestinal side effects. A previously published pilot study conducted in a small group of patients (n = 5) has indicated that taking metformin 30 minutes prior to a meal may be more efficacious than taking it with the meal.

Sixteen participants with type 2 diabetes, effectively managed by metformin monotherapy (500–2000 mg daily for at least 3 months), participated in a double-blind, randomized, crossover study. Participants were instructed to maintain their usual metformin dose, adjusting only the time of administration. A standardized meal was consumed the evening prior to each study day, followed by an overnight fast. A nasoduodenal catheter was inserted to allow for precise intraduodenal administration of metformin and glucose. The study employed four distinct treatments in a randomized crossover design: (1) 1000 mg metformin at t = -60 min followed by saline at t = -30 min and 0 min; (2) 1000 mg metformin at t = -30 min followed by saline at t = -60 min and 0 min; (3) 1000 mg metformin at t = 0 min followed by saline at t = -60 min and -30 min; and (4) saline at t = -60, -30, and 0 min (control). At t = 0 min, a standardized intraduodenal glucose infusion (12.56 kJ/min for 60 min) commenced. Blood samples were collected at various time points (t = -60 min to t = 120 min) to measure plasma glucose, total GLP-1, and insulin levels. Nausea and appetite sensations were also assessed using visual analogue scales. The integrated AUC (iAUC) for plasma glucose, insulin, and total GLP-1 were calculated using the trapezoidal rule. Whole-body insulin sensitivity was estimated using the Matsuda index. One-factor repeated-measures ANOVA was used to compare fasting plasma glucose, insulin, GLP-1 levels, and the Matsuda index between the four study days. Two-factor repeated-measures ANOVA analyzed plasma glucose, GLP-1, and insulin levels with treatment and time as factors. Post hoc comparisons were conducted using Bonferroni correction. The study was powered to detect a 20% difference in glucose iAUC0–120min between metformin treatments and the control.

The study revealed a significant treatment-by-time interaction for metformin's effects on plasma glucose, GLP-1, and insulin levels (p < 0.05 for each). Plasma glucose levels were significantly lower when metformin was administered at t = -60 or -30 min compared to t = 0 min (p < 0.05 for each). Increased GLP-1 levels were observed only when metformin was given at t = -60 or -30 min (p < 0.05 for each). Metformin did not significantly impact insulin sensitivity but enhanced glucose-induced insulin secretion (p < 0.05), with comparable increases in plasma insulin across all three metformin administration days. Specifically, comparing the integrated area under the curve (iAUC) from 0 to 120 minutes post-infusion, the reduction in plasma glucose was significantly greater with metformin administration at -60 min and -30 min compared to metformin at 0 min and the control group. Similarly, the increase in GLP-1 iAUC was significantly greater for -60 min and -30 min metformin administration than for the 0 min metformin and the control. Fasting levels of glucose, GLP-1 and insulin were not significantly different between the four treatment groups.

The findings strongly suggest that administering metformin 30–60 min before an intraduodenal glucose load is substantially more effective in reducing glycemia and increasing GLP-1 release than administering it concurrently with glucose. The enhanced GLP-1 response observed with earlier metformin administration likely plays a crucial role in the improved glucose-lowering effect. The absence of gastrointestinal symptoms further supports the potential benefit of pre-meal metformin administration. The results challenge the current clinical recommendation of taking metformin with meals, implying that this practice might compromise its efficacy in reducing postprandial hyperglycemia. Improved postprandial glucose control is critical for preventing microvascular complications and reducing cardiovascular risks associated with type 2 diabetes. The mechanisms underlying the improved GLP-1 response with earlier metformin administration might involve metformin's influence on glucose absorption in the upper small intestine. By inhibiting glucose absorption proximally, increased glucose exposure to L cells more distally may stimulate higher GLP-1 release. While metformin augmented insulin secretion, other pathways may also contribute to the observed glucose-lowering effect, as the increase in GLP-1 was only evident with the earlier metformin administration.

This study demonstrates that in well-controlled type 2 diabetes patients on metformin, pre-meal metformin administration (30-60 minutes prior) is more effective in lowering postprandial glucose levels than concurrent administration. This enhanced efficacy is associated with a significant increase in GLP-1 levels. These findings challenge the current clinical practice and suggest that adjusting the timing of metformin administration may significantly improve glycemic control. Future research should investigate the effects of different metformin formulations and explore the optimal timing of metformin administration in real-world meal settings.

Several limitations warrant consideration. The study utilized a nasoduodenal catheter for metformin administration, a somewhat artificial method compared to oral administration. The use of a glucose-only solution, rather than a mixed nutrient load, might limit the generalizability of these findings. The relatively small sample size and unequal sex distribution may also limit the generalizability of the findings. Future research should expand upon these limitations with larger, more diverse cohorts and more physiological methods of metformin and nutrient administration.