A systematic review examining respiratory adverse events temporally associated with glucagon-like peptide-1 receptor agonist therapies has delineated a complex pulmonary safety profile across clinical trials, observational cohorts, and pharmacovigilance repositories. Published in Cardiovascular Diabetology – Endocrinology Reports by lead author Aarushi Ahuja and colleagues, the evidence synthesis aggregates findings from nineteen eligible investigations encompassing 1,124,537 clinical participants alongside 498,892 spontaneous adverse event reports. The authors structured their project as a signal-detection and evidence-mapping exercise rather than a causal-inference analysis, reflecting the predominance of observational and spontaneous reporting data. Rather than demonstrating a uniformly hazardous pulmonary impact, the findings reveal that common upper respiratory tract conditions occur at frequencies indistinguishable from comparator arms, whereas rare life-threatening pulmonary complications appear primarily within case reports and pharmacovigilance entries. Consequently, the review establishes an evidence map to guide clinical vigilance while delineating the need for standardised prospective respiratory research.
The clinical context underpinning this investigation reflects the rapid global expansion of glucagon-like peptide-1 receptor agonists, initially approved for glycaemic control in type 2 diabetes mellitus and subsequently authorised for chronic weight management and cardiovascular risk reduction. Major medications including semaglutide, liraglutide, dulaglutide, and the dual agonist tirzepatide are currently prescribed to more than fifty million individuals worldwide. Although gastrointestinal side effects have been extensively documented throughout regulatory reviews and clinical literature, the pulmonary pharmacodynamics of these agents have remained incompletely defined. Glucagon-like peptide-1 receptors are biologically expressed across pulmonary epithelial cells, airway smooth muscle, alveolar macrophages, and pulmonary vasculature. This anatomical distribution establishes a plausible biological substrate for diverse respiratory pharmacodynamic responses, ranging from potential anti-inflammatory influences to bronchospastic reactions. Furthermore, the well-established pharmacological class effect of delayed gastric emptying introduces a predictable physical mechanism capable of increasing pulmonary aspiration risks, particularly in patients undergoing procedural sedation or general anaesthesia.
To systematically capture and appraise reported pulmonary adverse events, the investigators conducted their systematic review in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, prospectively registering the protocol with PROSPERO under registration CRD420261305875. The comprehensive literature search spanned PubMed, Embase, and Scopus from database inception through January 31, 2026, without language restrictions. Search queries combined Medical Subject Headings and free-text terms covering specific drug entities, diverse respiratory diagnostic outcomes, and adverse pharmacovigilance terminology. Supplementary investigations examined clinical trial registries, grey literature databases, medRxiv preprints, conference proceedings, and major regulatory portals including FAERS and EudraVigilance. The initial retrieval yielded 1,271 records, from which 483 duplicates were removed. Independent screening of 788 titles and abstracts eliminated 742 records, leaving 46 publications for full-text eligibility appraisal. Applying pre-specified exclusion criteria eliminated 27 papers, yielding 19 final studies with robust inter-rater agreement scores of 0.91 at screening and 0.88 at full-text assessment.
Methodological quality and risk of bias were systematically appraised using validated assessment instruments tailored to each underlying research design. Randomised controlled trials were evaluated using the Cochrane Risk of Bias 2 tool, retrospective cohort investigations were appraised using the Newcastle-Ottawa Scale, and case reports and pharmacovigilance studies were evaluated using Joanna Briggs Institute checklists. Across the four included randomised trials, overall risk of bias was rated low, although three trials exhibited domain-specific concerns regarding open-label outcome ascertainment for subjective symptoms like cough and dyspnoea. Among five retrospective cohort studies, three were classified as moderate risk, one exhibited high risk due to lacking a comparator group, and Henney’s large cohort received a moderate risk score of six out of nine. All three pharmacovigilance analyses received high risk ratings because spontaneous reporting systems lack baseline population denominators and cannot establish causality. Overall, nine studies demonstrated low risk, four moderate risk, and six high risk.
Upper respiratory tract infections and nasopharyngeal symptoms emerged as the most frequently documented adverse pulmonary manifestations, identified in eight of twelve studies explicitly monitoring upper airway outcomes. In randomised controlled trials, these occurrences took place at rates statistically comparable between active treatment groups and control arms. In the phase three SURPASS J-COMBO trial evaluating tirzepatide in 442 patients, nasopharyngitis occurred in seventeen percent of participants, showing no statistically significant deviation from control frequencies. Similarly, a phase two trial of mazdutide encompassing 180 participants documented symmetric distribution of upper respiratory complaints across active dosing cohorts and placebo controls. A randomised evaluation of a dulaglutide biosimilar involving 440 patients revealed equivalent respiratory infection rates across arms, while a retrospective cohort by Foer examining 1,642 patients with comorbid respiratory disease showed no excess exacerbation risk compared to active comparators, reporting a hazard ratio of 0.72. The reviewers established moderate certainty evidence for upper respiratory tract symptoms under the GRADE framework.
In contrast to benign upper airway symptoms, pharmacovigilance disproportionality analyses identified distinct statistical reporting signals for dyspnoea and bronchospasm that warrant careful scientific contextualisation. Dyspnoea represented the most frequently flagged serious pulmonary signal, appearing across ten of thirteen studies addressing the complaint. In the largest pharmacovigilance evaluation, Cazzola and colleagues analysed 251,145 adverse event reports from the FAERS database, identifying disproportionate reporting signals for asthma-like events with a reporting odds ratio of 1.63 and a 95% confidence interval of 1.41 to 1.88. Exenatide exhibited the most pronounced disproportionality signal for dyspnoea, generating a reporting odds ratio of 2.14 alongside thirty-two recorded respiratory-related fatalities. In a separate FAERS analysis by Zhang comprising 20,350 tirzepatide reports, dyspnoea produced a non-significant reporting odds ratio of 1.18. Crucially, the authors emphasise that disproportionality statistics represent exploratory signal-detection metrics rather than true incidence figures, relative risks, or causal proof. Spontaneous reporting databases are vulnerable to reporting bias, lack denominator data, and cannot determine individual risk.
The systematic review highlighted an important pharmacological distinction governing pulmonary hypersensitivity reactions, separating exendin-4-based agents from synthetic human glucagon-like peptide-1 analogues. Exendin-4-derived molecules such as exenatide and lixisenatide originate from a peptide isolated from the Gila monster, Heloderma suspectum, sharing only fifty percent amino acid sequence homology with endogenous human glucagon-like peptide-1. This structural divergence confers elevated immunogenic potential, facilitating anti-drug antibody formation and immunoglobulin E-mediated hypersensitivity reactions. Conversely, human peptide analogues such as liraglutide, semaglutide, and dulaglutide share ninety percent or greater sequence homology with native human sequences, presenting a lower immunogenic burden. Documented clinical occurrences supported this structural pattern through rare case reports. Published accounts detailed acute angioedema requiring mechanical ventilation after liraglutide, anaphylaxis triggered by lixisenatide or dulaglutide, and exenatide-induced pulmonary infiltrates with eosinophilia. Furthermore, biopsy-confirmed acute eosinophilic pneumonia resolved following exenatide discontinuation and corticosteroid therapy. Derived entirely from isolated case reports, evidence certainty was categorised as very low under GRADE.
Aspiration pneumonitis and aspiration pneumonia represented another critical domain of clinical concern, driven by the established physiological mechanism of delayed gastric emptying. Glucagon-like peptide-1 receptor agonists impair gastric motility through receptor pathways, resulting in substantial intragastric fluid and solid retention. For long-acting subcutaneous weekly formulations like semaglutide, which possesses an elimination half-life of approximately seven days, standard preoperative fasting intervals frequently fail to achieve adequate gastric emptying before anaesthesia. A published case series by Avraham documented two surgical patients who developed severe aspiration pneumonitis following elective procedures under general anaesthesia despite adhering to standard fasting protocols, with nasogastric aspiration confirming massive residual volumes. In a retrospective surgical cohort of 138,980 patients analysed by Rashid, 2,944 exposed individuals exhibited an adjusted hazard ratio of 1.28 for postoperative aspiration pneumonia, representing a biologically consistent trend. Clinical guidance from the American Society of Anesthesiologists recommends withholding weekly agents for a complete seven-day dosing interval before elective surgery, alongside point-of-care gastric ultrasound.
Directly challenging any simplistic characterisation of glucagon-like peptide-1 receptor agonists as uniformly harmful to respiratory health, the review evaluated a paradoxical large-scale protective signal regarding pneumonia incidence. In an extensive retrospective electronic health record cohort conducted by Henney using the TriNetX platform, investigators evaluated 331,863 matched patients with type 2 diabetes mellitus. Patients receiving glucagon-like peptide-1 receptor agonists demonstrated an adjusted hazard ratio of 0.60 with a 95% confidence interval of 0.58 to 0.62 for incident pneumonia compared to dipeptidyl peptidase-4 inhibitor users. This forty percent relative risk reduction remained consistent across multiple sensitivity and demographic subgroups. Proposed biological mechanisms explaining this protective association included enhanced respiratory mechanics following weight loss, decreased reflux microaspiration, direct anti-inflammatory receptor stimulation in alveolar macrophages, and altered mucosal immunity. Nevertheless, the authors urged caution, categorising certainty as low under GRADE. Dipeptidyl peptidase-4 inhibitors are active comparators, and retrospective observational designs remain inherently susceptible to residual confounding and prescribing bias.
Life-threatening acute respiratory failure and critical care admissions formed the most severe outcomes catalogued within the systematic review, though overall incidence could not be formally established. Acute respiratory failure was described across six studies, with catastrophic acute respiratory distress syndrome documented in two reports, including a fatal case involving a thirty-three-year-old male requiring veno-venous extracorporeal membrane oxygenation described by Rehan. Another case report documented spontaneous pneumomediastinum linked to severe vomiting barotrauma secondary to drug-induced gastroparesis. Secondary outcomes revealed that respiratory hospitalisation was noted in eleven of fourteen reporting studies, intensive care unit admission occurred in three studies, and mechanical ventilation was documented in three studies. Overall, evidence certainty was rated moderate for upper respiratory infections, low for pneumonia reduction, and very low for severe adverse events. Moving forward, the authors underscored that expanding global prescription of glucagon-like peptide-1 receptor agonists necessitates dedicated prospective trials with pre-specified pulmonary endpoints, standardised perioperative ultrasound registries, and denominator-based pharmacovigilance studies to transition from signal detection to causal understanding.
Subject of Research: Respiratory safety signals and pulmonary adverse events associated with GLP-1 receptor agonists
Article Title: Pulmonary adverse events associated with GLP-1 receptor agonists: a systematic review of respiratory safety signals
Article References: Ahuja, A., Prasad, S., Manoharan, S., Sen, A., Rana, R., Mody, M., Paras, F., Archita, A., Shaik, S., Mehta, K., Manoj, R., Mishra, P., Mehta, S., & Kapoor, P. (2026). Pulmonary adverse events associated with GLP-1 receptor agonists: a systematic review of respiratory safety signals. Cardiovascular Diabetology – Endocrinology Reports, 12(1), Article 43. https://doi.org/10.1186/s40842-026-00311-6
Image Credits: AI Generated
DOI: 10.1186/s40842-026-00311-6
Keywords: GLP-1 receptor agonists, pulmonary adverse events, respiratory safety, pharmacovigilance, systematic review, aspiration risk, exenatide, pneumonia risk, Pulmonary, adverse, events, associated
