GLP-1 Receptor Agonists Drug-Drug Interaction Table

Why GLP-1 Receptor Agonists Create Drug Interactions

GLP-1 receptor agonists (GLP-1 RAs) bind to the GLP-1 receptor on gastric smooth muscle, pancreatic beta cells, and central vagal afferents. The result is dose-dependent slowing of gastric emptying, which serves as the primary pharmacokinetic mechanism behind most drug-drug interactions in this class [1]. This effect is most pronounced with short-acting agents such as exenatide twice daily and becomes attenuated (though not eliminated) with longer-acting formulations.

Gastric Emptying as the Core Mechanism

A scintigraphy study published in Diabetes Care measured gastric emptying half-time in patients receiving exenatide 10 mcg twice daily versus placebo. Exenatide prolonged gastric emptying half-time from 72 minutes to 120 minutes, a 67% increase [2]. For oral medications that depend on rapid duodenal absorption to reach therapeutic peak concentrations, this delay matters. Drugs with wide therapeutic indices (metformin, most antihypertensives) tolerate the change. Drugs with narrow windows (warfarin, digoxin, levothyroxine) require closer monitoring.

Short-Acting Versus Long-Acting Agents

Short-acting GLP-1 RAs (exenatide twice daily, oral semaglutide taken once daily in a fasting window) produce pulsatile gastric slowing timed to the injection or ingestion. Long-acting agents (semaglutide weekly, dulaglutide, tirzepatide) produce continuous but physiologically attenuated gastric slowing because of tachyphylaxis at the gastric GLP-1 receptor [3]. The 2022 ADA/EASD consensus report notes that "the magnitude of gastric emptying delay with once-weekly GLP-1 RAs is clinically smaller than with short-acting formulations, but it is not zero" [4]. Prescribers should still evaluate the co-medication list at every GLP-1 RA initiation.

Patient Factors That Amplify the Risk

Diabetic gastroparesis, opioid co-administration, and anticholinergic burden each slow gastric motility independently. Layering a GLP-1 RA onto any of these creates additive delay. A 2023 retrospective cohort analysis in JAMA Internal Medicine (N=16,922) found that patients on a GLP-1 RA plus an opioid had a 33% higher rate of bowel obstruction-related ED visits compared to patients on a GLP-1 RA alone [5]. Screening for baseline motility impairment before prescribing is a practical first step.

High-Priority Interactions: Insulin, Sulfonylureas, and Hypoglycemia

The single most common serious adverse event with GLP-1 RA combination therapy is hypoglycemia, and it almost always involves co-prescribed insulin or a sulfonylurea. The mechanism is straightforward: GLP-1 RAs enhance glucose-dependent insulin secretion, and stacking that on top of glucose-independent insulin release from exogenous insulin or sulfonylureas removes the glucose-dependent safety brake.

Insulin Dose Reduction at GLP-1 RA Initiation

In the SUSTAIN-5 trial (N=397), patients on semaglutide 1 mg plus basal insulin experienced a symptomatic hypoglycemia rate of 11.3%, versus 1.1% in the semaglutide-plus-metformin-only arm [6]. The Endocrine Society's 2023 clinical practice guideline recommends reducing basal insulin by 20% when adding a GLP-1 RA in patients with an HbA1c <8% [7]. For patients already at goal (HbA1c <7%), a 30% to 40% reduction is reasonable, with self-monitored glucose guiding further titration.

Sulfonylurea Dose Adjustment

The AWARD-2 trial (N=810) compared dulaglutide 1.5 mg to insulin glargine, both added to glimepiride plus metformin. Symptomatic hypoglycemia occurred in 8.6% of the dulaglutide arm [8]. Current labeling for all GLP-1 RAs includes a warning to consider reducing the sulfonylurea dose when initiating combination therapy. A practical approach: halve the sulfonylurea dose at GLP-1 RA initiation and retitrate upward only if postprandial glucose remains above target for two consecutive weeks.

Meglitinides

Repaglinide and nateglinide carry similar hypoglycemia-stacking risk. No large prospective trial specifically addresses GLP-1 RA-plus-meglitinide combinations, but the pharmacologic rationale is identical to sulfonylureas. Reduce the meglitinide dose by 50% or consider discontinuation if postprandial control improves substantially after GLP-1 RA titration [7].

Warfarin and Other Anticoagulants

Warfarin's narrow therapeutic index makes any absorption change clinically relevant. The FDA-approved labeling for exenatide notes that co-administration increased INR in some post-marketing reports, though a dedicated pharmacokinetic study showed no change in warfarin AUC [9]. The disconnect likely reflects individual variability in gastric emptying effects and dietary intake changes (reduced appetite leading to lower vitamin K consumption).

Monitoring Recommendations

Check INR within 2 to 4 weeks of starting or up-titrating any GLP-1 RA in warfarin-treated patients. Repeat after each dose escalation. The American College of Cardiology's 2024 expert consensus statement on anticoagulation management notes: "Patients initiating GLP-1 receptor agonists who are on warfarin should be treated as though a new interacting medication has been added, with INR rechecked at short intervals until stability is confirmed" [10].

Direct Oral Anticoagulants

Apixaban, rivarelbaan, edoxaban, and dabigatran depend on intestinal absorption but have wider therapeutic indices than warfarin. No prospective pharmacokinetic interaction studies exist for DOAC-plus-GLP-1-RA combinations. However, because DOACs have predictable pharmacokinetics and do not require routine monitoring, the clinical risk is lower. Delayed Tmax without meaningful AUC change (the typical GLP-1 RA absorption pattern) should not reduce DOAC efficacy. No dose adjustment is recommended at this time [10].

Oral Contraceptives

Semaglutide's prescribing information includes a dedicated pharmacokinetic substudy evaluating co-administration with a combined oral contraceptive (ethinyl estradiol 0.03 mg / levonorgestrel 0.15 mg). Semaglutide 1 mg reduced ethinyl estradiol Cmax by 12% and levonorgestrel Cmax by 12%, with AUC reductions of 1% and 4%, respectively [11]. These changes fall within accepted bioequivalence boundaries (80% to 125%).

Clinical Guidance

The 12% Cmax reduction is unlikely to cause contraceptive failure. The ACOG Practice Bulletin on hormonal contraception does not list GLP-1 RAs as interacting drugs [12]. Patients switching from a combined oral contraceptive to a progestin-only pill or non-oral method for unrelated reasons do not need to accelerate that switch because of GLP-1 RA co-administration. Counsel patients that GI side effects (nausea, vomiting) during GLP-1 RA titration could impair oral contraceptive absorption through vomiting-related missed doses, and recommend a backup method during the first 4 weeks of titration if vomiting occurs.

Levothyroxine and Thyroid Hormone Replacement

Levothyroxine depends on fasting gastric pH and rapid duodenal absorption for consistent bioavailability. GLP-1 RA-mediated gastric slowing delays levothyroxine transit into the duodenum, potentially reducing peak absorption. A small crossover study (N=28) found that exenatide co-administration reduced levothyroxine Cmax by 16% without significantly changing AUC [13].

Practical Prescribing

Instruct patients to take levothyroxine 30 to 60 minutes before food, as they normally would, and to separate the levothyroxine dose from any short-acting GLP-1 RA injection by at least 1 hour. For once-weekly injectable GLP-1 RAs, separation timing is less critical because the gastric effect is continuous and attenuated. Recheck TSH 6 to 8 weeks after GLP-1 RA initiation or dose escalation. Adjust levothyroxine as guided by TSH results [13].

Metformin, SGLT2 Inhibitors, and DPP-4 Inhibitors

These three oral antidiabetic classes are the most common co-prescriptions with GLP-1 RAs in type 2 diabetes management.

Metformin

Dulaglutide reduced metformin Cmax by 9% and delayed Tmax by 1 hour but did not change metformin AUC [14]. No dose adjustment is needed. Metformin's extended-release formulation, which already has a prolonged absorption profile, is even less susceptible to gastric emptying changes.

SGLT2 Inhibitors

No pharmacokinetic interaction studies have identified clinically meaningful changes when empagliflozin, dapagliflozin, or canagliflozin are combined with GLP-1 RAs. The DURATION-8 trial (N=695) evaluated exenatide once weekly plus dapagliflozin and found additive HbA1c reduction (1.5% versus 1.0% for either monotherapy) without increased hypoglycemia [15]. The combination is considered complementary, not competitive.

DPP-4 Inhibitors

DPP-4 inhibitors (sitagliptin, saxagliptin, linagliptin) work on the same incretin axis as GLP-1 RAs. Co-prescribing is pharmacologically redundant. The ADA's 2024 Standards of Care states: "There is no additional glycemic benefit to combining a DPP-4 inhibitor with a GLP-1 receptor agonist, and the combination is not recommended" [16]. Discontinue the DPP-4 inhibitor when initiating a GLP-1 RA.

Statins, ACE Inhibitors, and ARBs

Statins

Exenatide delayed atorvastatin Cmax by 38% and Tmax by approximately 4 hours. AUC decreased by only 2%, meaning total drug exposure was preserved [17]. Because statin efficacy depends on sustained hepatic HMG-CoA reductase inhibition (an AUC-driven effect, not a Cmax-driven effect), the interaction is not clinically meaningful. No statin dose adjustment is warranted with any GLP-1 RA. Rosuvastatin and simvastatin show similar patterns.

ACE Inhibitors and ARBs

Lisinopril Cmax was reduced by 15% and delayed by 2.5 hours when co-administered with liraglutide [18]. AUC was reduced by 3%. Blood pressure control was unaffected in the LEADER trial (N=9,340), where 82% of patients were on renin-angiotensin system inhibitors and no signal of hypertension worsening appeared [19]. No dose adjustment needed.

Digoxin, Narrow-Index Oral Drugs, and Practical Timing

Digoxin has a narrow therapeutic index (0.5 to 0.9 ng/mL for heart failure, per 2022 AHA/ACC/HFSA guidelines) [20]. Exenatide delayed digoxin Tmax by 2.5 hours and reduced Cmax by 17%, though AUC decreased by only 3.5% [17].

Monitoring Protocol

Check a trough digoxin level 5 to 7 days after GLP-1 RA initiation and again after each dose escalation. Because digoxin toxicity can present as nausea and vomiting, symptoms that overlap with GLP-1 RA side effects, maintain a low threshold for checking levels in any patient reporting new or worsening GI symptoms after a GLP-1 RA dose increase.

Other Narrow-Index Drugs

Phenytoin, carbamazepine, and cyclosporine have narrow therapeutic indices and depend on predictable oral absorption. No dedicated pharmacokinetic studies exist for these drugs combined with GLP-1 RAs. The prudent approach: check trough levels of any narrow-index oral medication 1 to 2 weeks after GLP-1 RA initiation and after each titration step.

Oral Semaglutide: A Special Absorption Case

Oral semaglutide (Rybelsus) uses the absorption enhancer SNAC (sodium N-[8-(2-hydroxybenzoyl)amino]caprylate) to support gastric absorption. SNAC requires a fasting stomach and a specific pH environment. Co-administered oral medications taken within 30 minutes of oral semaglutide can interfere with SNAC-mediated absorption of semaglutide itself, not just the other way around [21].

Dosing Window

The prescribing information specifies: take oral semaglutide with no more than 4 ounces of plain water, at least 30 minutes before the first food, beverage, or other oral medication of the day [21]. Levothyroxine and proton pump inhibitors, both of which also require fasting administration, create a scheduling conflict. The recommended sequence: oral semaglutide first, wait 30 minutes, then levothyroxine or PPI, then wait another 30 minutes before eating. Alternatively, switch to injectable semaglutide to simplify the medication schedule.

Proton Pump Inhibitors and H2 Blockers

Omeprazole 40 mg co-administered with oral semaglutide showed no significant change in omeprazole AUC or Cmax [21]. For injectable GLP-1 RAs, PPI interactions are not expected because the mechanism involves gastric transit, not gastric pH. H2 receptor antagonists (famotidine, ranitidine) have no documented interaction with any GLP-1 RA. No dose adjustments are necessary for either class.

Alcohol and Recreational Substances

Alcohol slows gastric emptying independently. Adding a GLP-1 RA to regular alcohol consumption can produce pronounced nausea, delayed alcohol absorption with unpredictable intoxication kinetics, and worsened GI side effects. No controlled pharmacokinetic study exists, but the STEP-1 trial (N=1,961) excluded heavy alcohol users, so safety data in this population is limited [22]. Counsel patients that alcohol tolerance may change after GLP-1 RA initiation and that they should titrate alcohol consumption cautiously during the first 8 to 12 weeks.

Quick-Reference DDI Summary Table

| Co-administered Drug | GLP-1 RA Effect | Clinical Action | |---|---|---| | Insulin (basal) | Additive hypoglycemia | Reduce insulin 20% to 40% at initiation | | Sulfonylurea | Additive hypoglycemia | Halve the dose at initiation | | Warfarin | Possible INR increase | Check INR at 2 to 4 weeks; recheck after titration | | Oral contraceptives | Cmax reduced ~12% | No dose change; counsel on vomiting-related missed doses | | Levothyroxine | Cmax reduced ~16% | Recheck TSH at 6 to 8 weeks | | Digoxin | Tmax delayed 2.5 hours, Cmax reduced 17% | Check trough level at 5 to 7 days | | Metformin | Cmax reduced 9%, AUC unchanged | No adjustment | | Statins | Cmax reduced up to 38%, AUC unchanged | No adjustment | | ACE inhibitors/ARBs | Cmax reduced ~15%, AUC unchanged | No adjustment | | DPP-4 inhibitors | Pharmacologic redundancy | Discontinue DPP-4i | | SGLT2 inhibitors | Additive efficacy, no PK interaction | No adjustment | | DOACs | No data; theoretical Tmax delay | No adjustment; routine DOAC monitoring not standard |