Semaglutide Induces Changes in Gastric Electrical Activity in Patients With Overweight and Obesity: A Pilot Study

Abstract

Background/Aims

Semaglutide is associated with gastroduodenal symptoms, such as nausea and vomiting. This pilot study used body surface gastric mapping (BSGM) to measure the effect of semaglutide on gastric function and associated symptoms.

Methods

Patients without gastrointestinal symptoms (n = 8) underwent BSGM at baseline and on 1 mg semaglutide, as per standard subcutaneous dosing. Spectral metrics included Principal Gastric Frequency, Gastric Alimetry Rhythm Index (GA-RI), body mass index-adjusted amplitude, fed:fasted amplitude ratio, and meal response ratio. Symptoms were assessed using validated questionnaires. Data were compared using paired t tests.

Results

Body mass index-adjusted amplitude showed a significant decrease on the drug (P = 0.04). Five patients (63%) developed spectral abnormalities on the drug, with 2 displaying a low or undetectable GA-RI (< 0.25). The patient assessment of upper gastrointestinal disorders-symptom severity index fullness/early satiation subscale significantly increased on the drug (P = 0.005).

Conclusions

Semaglutide appears to alter gastric electrical activity on BSGM and increase early satiation, offering potential biomarkers for detecting drug effects. Further studies are needed.

Introduction

Semaglutide (brand name: Ozempic, Wegovy) is a long-acting glucagon-like peptide-1 (GLP-1) receptor agonist (GLP-1RA) indicated for glycemic control in type 2 diabetes and weight loss in body mass index (BMI) > 30 or > 27 with a weight-related comorbidity, presenting numerous benefits including in cardiometabolic risk factors.¹ However, GLP-1RAs have been linked to the rise in gastroduodenal side effects, such as nausea and vomiting, mirroring symptoms of gastroparesis, a condition characterized by delayed gastric emptying.^2,3 Up to 50% of patients on semaglutide report gastroduodenal side effects, with the most frequent being nausea.^1-3 While (American style) alterations in gastric function, such as delayed gastric emptying, have been noted with semaglutide use,⁴ the exact mechanism by which semaglutide may cause gastroduodenal symptoms is unknown.

Gastric Alimetry is a novel non-invasive diagnostic tool to assess gastric myoelectrical activity that incorporates body surface gastric mapping (BSGM) and validated patient-reported symptom tracking to identify biomarkers of gastric dysfunction.^5-7 The system utilizes a 64-electrode stretchable array placed over the epigastrium, a wearable Reader and an associated Gastric Alimetry app to measure gastric myoelectrical activity from the skin surface along with monitoring real-time patient symptoms. This observational pilot study aims to measure the effect of semaglutide on gastric function using BSGM and to assess the association with symptoms.

Materials and Methods

Ethics approval was obtained from the Western Sydney University Human Research Ethics Committee (H15157). All patients provided informed consent. Data used for analysis will be made available on reasonable request, conditional on ethical approvals. Clinical trial registration is NCT06401746.

Patients aged ≥ 18 years, did not present with gastroduodenal symptoms and prescribed semaglutide as part of their standard care (for weight loss or diabetes) were recruited from a tertiary center. Exclusion criteria included pregnancy, breastfeeding, previously diagnosed gastroduodenal disease or functional disorder, or inability to undergo BSGM testing due to history of adhesive allergies or damaged epigastric skin. Adhesive allergies or damaged epigastric skin are a contraindication according to the Instructions for Use,⁸ as this could lead to skin irritation, and does not introduce selection bias.

All subjects underwent BSGM testing using the Gastric Alimetry System (Alimetry Ltd, Auckland, New Zealand), as per standard protocol.⁵ The standard test consists of array placement, 30-minute fasting recording, 10-minute standard 480 kcal meal, and 4-hour post-prandial recording. The meal consisted of a nutrient drink (230 kcal, 230 mL; Ensure, Abbott Nutrition, Chicago, IL, USA) and an oatmeal energy bar (Clif Bar Nutrition Bar Chocolate Chip; 250 kcal, 5 g fat, 45 g carbohydrate, 10 g protein, 7 g fiber; Clif Bar & Company, Emeryville, CA, USA). Symptoms were reported via the validated Gastric Alimetry app and a total symptom burden score was calculated (out of 70), as previously described.⁶ Testing was performed at baseline and while on drug. BSGM metrics were defined and compared to normative reference intervals:⁷ Principal Gastric Frequency (PGF), Gastric Alimetry Rhythm Index (GA-RI), BMI-adjusted amplitude, and Fed:Fasted Amplitude Ratio (ffAR). A low GA-RI < 0.25 is indicative of low rhythm stability (gastric dysrhythmia).^5,7,9 A low PGF < 2.65 cpm or high PGF > 3.35 cpm may be indicative of impaired pacemaker activity.^7,9 A low BMI-adjusted amplitude < 22 µV or high BMI-adjusted amplitude > 70 µV may be indicative of impaired gastric muscle contractions.^7,9 A ffAR > 1.08 is defined as a meal response.^7,9 The meal response ratio (MRR) is a novel metric that was calculated as the ratio of the average amplitude in the first 2 hours postprandially to that of the last 2 hours, with a delayed meal response defined as MRR < 1.¹⁰

Subjects were prescribed semaglutide as part of their clinical care and started the drug following standard dosing: 0.25 mg subcutaneously weekly for 4 weeks, 0.5 mg weekly for 4 weeks and 1 mg weekly maintenance dose thereafter. BSGM was repeated when the 1 mg dose was reached or at the maximum tolerated dose.

Symptoms were evaluated using validated standard questionnaires: Rome IV criteria, which is used to diagnose gastroduodenal disorders of gut–brain interaction based on specific symptom patterns;¹¹ Patient assessment of upper gastrointestinal disorders-symptom severity index (PAGI-SYM), which is a self-reported survey to assess upper gastroduodenal symptom severity in the last 2 weeks;¹² and gastroparesis cardinal symptom index (GCSI), which is a self-reported survey to assess the severity of gastroparesis symptoms over the last 2 weeks.¹³

Data are presented as median with interquartile range (IQR) or mean ± standard deviation. Analyses were conducted on data collected at both time points (baseline and on drug). Absent PGF values were excluded from analysis. Absent GA-RI values were included and treated as ‘0.’ Metric and symptom data were treated as continuous variables and compared using paired t tests. A P-value of < 0.05 was considered significant. Confidence intervals were calculated at the 95% level.

Results

Nine patients were recruited: 6 females (66%), median age 55 years (IQR, 52-60), median BMI 45 kg/m² (IQR, 42.8-48.5). Eight patients completed the study, with one drop-out due to unavailability of the drug. Of the remaining eight patients, one patient met the Rome IV criteria for Functional Dyspepsia prior to testing. All patients consumed a median 100% (IQR, 97.5-100.0) of the meal at each time point and reached the maximum dose of 1 mg semaglutide. The time to titration was a median 35 days (IQR, 29.0-39.0) between 0.25 mg and 0.5 mg, and a median 32 days (IQR, 28.0-57.0) between 0.5 mg and 1 mg. The days between BSGM tests were median 92 (IQR, 71.0-187.0). BMI was similar between baseline and on drug (median 46.4 kg/m² [IQR, 44.4-49.6] vs median 46.6 kg/m² [IQR, 43.2-49.4]; P = 0.274). Body weight displayed a decreasing trend between baseline and on drug (median 140 kg [IQR, 128.5-152.3] vs 137 kg [IQR, 128.0-153.5]; P = 0.194).

While on semaglutide, PGF remained similar between baseline and on drug (mean 3.00 ± 0.10 to mean 3.12 ± 0.22; mean difference 95% CI, –0.31-0.12; P > 0.05; Fig. 1A). There was a general trend towards decreased GA-RI (mean 0.57 ± 0.24 at baseline to mean 0.44 ± 0.24 on drug; mean difference 95% CI, –0.04-0.29; P = 0.125; Fig. 1B), ffAR (mean 2.08 ± 1.06 at baseline to mean 1.64 ± 0.58 on drug; mean difference 95% CI, –0.06-0.94; P = 0.076; Fig. 1C), and MRR (mean 1.42 ± 0.89 at baseline to mean 1.03 ± 0.33 on drug; mean difference 95% CI, –0.40-1.17; P = 0.284; Fig. 1D). BMI-adjusted amplitude showed a significant decrease after being on the drug (mean 37.05 ± 8.14 µV at baseline to mean 32.94 ± 5.42 µV on drug; mean difference 95% CI, 0.26-7.96; P = 0.040; Fig. 1E). Five patients (63%) developed spectral abnormalities on drug: 2 patients (25%) with a low or absent GA-RI (defined as GA-RI < 0.25; indicative of low rhythm stability), 1 patient (13%) with a high frequency (defined as PGF > 3.35 cpm), and 2 patients (25%) with a delayed meal response (MRR < 1). The overall BSGM metrics for the patient cohort are shown in Table and representative case examples are shown in Figure 2.

Figure 1.

Change in overall body surface gastric mapping (BSGM) metrics (Principal Gastric Frequency [PGF], Gastric Alimetry Rhythm Index [GA-RI], Fed:Fasted Amplitude Ratio [ffAR], meal response ratio [MRR], and body mass index-adjusted amplitude) between baseline and on drug (semaglutide at 1 mg). Data are presented as median with interquartile range. Dotted black lines indicate the mean value. Dotted red lines show the established abnormal reference intervals: high PGF > 3.35 cpm; low GA-RI < 0.25; low ffAR < 1.08; delayed meal response: MRR < 1.

Table.

Overall Body Surface Gastric Mapping Metrics at Baseline and on Drug for the Patient Cohort Compared to the Established Normative Reference Intervals

Metric	Reference interval	Baseline	On drug
Principal Gastric Frequency (cpm), mean (SD)	2.65-3.35	3.00 (0.10)	3.12 (0.22)
BMI-Adjusted Amplitude (μV), mean (SD)	22-70	37.05 (8.14)	32.94 (5.42)
Gastric Alimetry Rhythm Index, mean (SD)	≥ 0.25	0.57 (0.24)	0.44 (0.24)
Fed:Fasted Amplitude Ratio, mean (SD)	≥ 1.08	2.08 (1.06)	1.64 (0.58)

Figure 2.

Representative spectrograms and symptom profiles of 2 patients at baseline (left) and on drug (semaglutide at 1 mg; right). (A) 55-year-old male with body mass index (BMI) 52.7, showing the reduction in BMI-adjusted amplitude. (B) 64-year-old male with BMI 44.9, showing the appearance of dysrhythmia (low Gastric Alimetry Rhythm Index [GA-RI]) with the increase in excessive fullness.

The overall gastroduodenal symptom scores were low at baseline (total symptom burden: mean 6.25 ± 5.44; PAGI-SYM: mean 0.70 ± 0.68; GCSI: mean 0.63 ± 0.54; Fig. 3A-C), with a trend toward higher GCSI symptom burden when on the drug (GCSI: mean 0.85 ± 0.33; mean difference 95% CI, –0.63-0.19; P = 0.244; Fig. 3C). Notably, the PAGI-SYM fullness/early satiation subscale increased from mean 0.81 ± 0.76 to mean 1.72 ± 0.54 (mean difference 95% CI, –1.44-–0.37; P = 0.005; Fig. 3D). All other symptom subscales were similar between baseline and on drug (P > 0.050).

Figure 3.

Change in symptom scores (total symptom burden, Patient assessment of upper gastrointestinal disorders-symptom severity index [PAGI-SYM], Gastroparesis Cardinal Symptom Index [GCSI], and PAGI-SYM fullness/early satiation subscale) between baseline and on drug (semaglutide at 1 mg). Data are presented as median with interquartile range. Dotted black lines indicate the mean value.

Discussion

This novel pilot data showed that semaglutide at 1 mg in a clinical use scenario significantly reduced the BMI-adjusted amplitude on BSGM and increased the fullness/early satiation symptom burden. This decreased gastric amplitude has also been observed in different studies using the GLP-1RAs liraglutide in healthy volunteers¹⁴ and exendin-4 in ferrets.¹⁵ The amplitude-suppressing effects of GLP-1RA drugs offer a potential new biomarker for monitoring GLP-1RA-related effects on the stomach.

Changes in gastric electrical activity due to semaglutide were reliably detected by BSGM, but the mechanism of GLP-1RAs on the stomach remain poorly understood. While BMI-adjusted amplitude remained within the normal reference interval,⁷ a decreased gastric electrical amplitude was noted in 7/8 (87.5%) patients along with the development of specific spectral abnormalities in 5/8 (63%) patients on drug. This may be related to the reduced stomach contractility with GLP-1RA use observed on antroduodenal manometry.¹⁶ Furthermore, the increase in early satiation seen in the patient cohort is consistent with the known GLP-1RA mechanism whereby semaglutide interacts with the GLP-1RA receptors in the hypothalamus, reducing hunger sensations and increasing feelings of satiation.⁴ GLP-1RAs appear to impair gastric motility through a combination of reduced gastric muscle contractions, delayed gastric emptying, increased pyloric tone, and vagal nerve influences, leading to sensations of fullness and reduced appetite.^4,16,17 While correlation analysis between changes in gastric activity and symptoms or physiological variables (such as weight loss) was not performed due to a small sample size, future, large-scale, dose-response studies could explore whether such changes in gastric function are associated with the onset of symptoms or physiological changes.

It is worth noting that BSGM identified gastric dysrhythmias in some patients on the drug. In 2 cases, GA-RI changed from normal (≥ 0.25) to low or undetectable on the drug. A low GA-RI is indicative of myoelectrical dysfunction, which is likely associated with poor coordination of gastric slow waves.⁵ Another 2 patients developed a delayed meal response (MRR < 1), which has been associated with delayed gastric emptying, and post-prandial distress in patients with functional dyspepsia, and nausea and vomiting disorders.^10,18 Further studies with a larger cohort are required to corroborate these findings.

Some limitations of this study should be noted. Firstly, there was a small sample size due to a drug shortage of semaglutide because of a high demand in Australia during the study period. This may affect the generalizability and statistical power of these pilot results and therefore, larger long-term prospective studies are needed to confirm these initial findings. Secondly, given the patient demographics for GLP-1RA use, all patients included were BMI ≥ 35, with the reference intervals derived from a normal control population of BMI < 35.⁷ A high BMI may attenuate amplitude due to abdominal adiposity and potentially reduce the signal-to-noise ratio, which can increase the risk of overestimating low rhythm stability (GA-RI) due to lower signal quality.⁷ However, this was addressed through the use of the BMI-adjusted amplitude metric and paired analysis in this study. This paired design utilized each patient as their own control, significantly reducing the risk of bias from high BMI and allowed for the observed true drug-induced change in amplitude. Moreover, all results were still interpretable per the Gastric Alimetry Instructions For Use and interpretation guidelines.^8,9 A further limitation is the absence of parallel data from other test modalities such as gastric emptying scintigraphy or conventional electrogastrography. In a different study, it was found that liraglutide significantly delayed gastric emptying, which was independent of the reduction in BMI-adjusted amplitude.¹⁴ While BSGM specifically assesses gastric myoelectrical activity, future, larger studies should incorporate gastric emptying testing to provide a comprehensive electrical and mechanical assessment of semaglutide's effects. Finally, the absence of a control group (eg, patients on alternative weight-loss medications or placebo) limits our ability to definitively attribute the observed changes solely to semaglutide alone. Future, larger studies are warranted to include non-GLP-1RA controls to better isolate drug-specific effects.

BSGM identified changes in gastric motor function induced by semaglutide use with concurrent fullness/early satiation symptoms. Further larger studies are now required to confirm these findings, and to explore the long-term effects, dose-response and impact of stopping GLP-1RA on gastric function and symptoms in comparison to a control group.