Metabolic effects of GLP-1 receptor agonists in rheumatology patients and inflammatory outcomes: a single-centre retrospective cohort Authors: Miriam Retuerto-Guerrero, Ignacio González Fernández, Mirian Alejo Ramos, Ismael González Fernández, Andrea Vélez Blanco, Anibal Martin Sanz, Octavio Miguel Rivero-Lezcano, Elvira Díez-Álvarez Published: 2026 ABSTRACT AbstractBackgroundObesity and type 2 diabetes are common in rheumatology, often exacerbated by glucocorticoids. Real-world data on glucagon-like peptide-1 receptor agonists (GLP-1RAs) in this setting remain limited.ObjectivesTo evaluate the metabolic effectiveness and safety of GLP-1RAs in rheumatology outpatients, including inflammatory marker and subgroup analyses.MethodsSingle-centre retrospective study (October 2024–October 2025) including 119 patients (from 169 screened) treated with GLP-1RAs. Paired analyses were performed at 3, 6 and 12 months. Exploratory comparisons included participants with chronic inflammatory arthritis (n=46) and those with non-inflammatory disease (n=43).ResultsPatients had high baseline cardiometabolic burden (mean body mass index (BMI) 34.9±9.9 kg/m²; diabetes 96.6%). GLP-1RA initiation was associated with sustained weight loss: −4.0 kg, −4.5 kg and −7.7 kg at 3, 6 and 12 months (all p<0.001; −8.2% at 12 months), with 56.5% improving ≥1 BMI category. Haemoglobin A1c decreased by −0.97%, −0.69% and −0.91% (all p<0.001), alongside significant reductions in fasting glucose (up to −24.8 mg/dL; p<0.001). Lipid improvements included reductions in low-density lipoprotein-cholesterol (−12.2 mg/dL; p=0.006) and triglycerides (−21.7 mg/dL; p=0.007). At 12 months, erythrocyte sedimentation rate decreased significantly (p=0.023), while CRP showed a non-significant decline (p=0.076); both were significantly reduced in the inflammatory subgroup. No significant differences were observed between inflammatory and non-inflammatory groups in adjusted metabolic trajectories. Adverse events were mainly gastrointestinal (13.4%); pancreatitis (0.8%) and cholecystitis (1.7%) were rare. Discontinuation occurred in 6.7%. One death (mesenteric ischaemia) had unclear causality.ConclusionsGLP-1RAs were associated with meaningful metabolic improvements in patients with rheumatic diseases, with a favourable safety profile. Potential anti-inflammatory effects require confirmation in prospective studies.Keywords: Cardiovascular Diseases, Inflammation, Arthritis WHAT IS ALREADY KNOWN ON THIS TOPIC Obesity is common in rheumatology and worsens clinical and cardiometabolic outcomes. Obesity is common in rheumatology and worsens clinical and cardiometabolic outcomes. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) improve weight and glycaemic control, but rheumatology-specific real-world data remain limited. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) improve weight and glycaemic control, but rheumatology-specific real-world data remain limited. WHAT THIS STUDY ADDS In this cohort, GLP-1RAs were associated with sustained weight loss, better glycaemic control and improved lipid profile. In this cohort, GLP-1RAs were associated with sustained weight loss, better glycaemic control and improved lipid profile. We also observed lower inflammatory markers in chronic inflammatory arthritis, with no excess of flares or debut of new rheumatic disease. We also observed lower inflammatory markers in chronic inflammatory arthritis, with no excess of flares or debut of new rheumatic disease. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY These findings support GLP-1RAs as a useful cardiometabolic option in rheumatology. These findings support GLP-1RAs as a useful cardiometabolic option in rheumatology. Prospective studies should clarify whether they have a clinically relevant anti-inflammatory effect. Prospective studies should clarify whether they have a clinically relevant anti-inflammatory effect. Introduction Obesity is a major comorbidity in rheumatology because it contributes to cardiometabolic risk and to higher disease activity, pain, disability and poorer therapeutic response. Overall, around 70% of adults with rheumatic diseases are overweight or obese, and excess adiposity is consistently associated with worse clinical outcomes and lower quality of life.1 In rheumatoid arthritis (RA), overweight and obesity have been linked to higher inflammatory activity, while in psoriatic arthritis (PsA) they are associated with greater disease severity and lower probability of achieving therapeutic targets; conversely, weight loss has been associated with clinical improvement.2 3 This burden is often compounded by glucocorticoid (GC) treatment. Prolonged systemic GC use is associated with clinically relevant weight gain, with a dose-dependent effect that may also be observed with long-term low-dose prednisolone.4 5 Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have shown robust efficacy in obesity and type 2 diabetes mellitus (T2DM), with consistent benefits on body weight, glycaemic control and cardiovascular risk.6 7 In addition, emerging evidence suggests that GLP-1RAs may modulate low-grade inflammation, mainly through modest reductions in C reactive protein (CRP) and tumour necrosis factor-alpha in patients with T2DM.8 These observations have generated growing interest in rheumatology, with recent editorials proposing GLP-1RAs’ relevance in rheumatic diseases due to their dual metabolic/anti-inflammatory potential.9 10 The aim of this study was to describe the use of GLP-1RAs in rheumatology outpatients and to evaluate their metabolic effectiveness, inflammatory signal and safety in routine clinical practice, including a comparative subanalysis between patients with chronic inflammatory arthritis and those with non-inflammatory/mechanical disease. Methods We conducted a single-centre retrospective real-world cohort study at Complejo Asistencial Universitario de León between October 2024 and October 2025. Among 11 788 adult patients followed in rheumatology outpatient clinics during the study period, 169 were identified as receiving GLP-1RA therapy through cross-referencing institutional pharmacy dispensing records and electronic clinical registries. Of these, 119 met the predefined criteria for effectiveness analyses, requiring ≥3 months of exposure, ≥2 weight or laboratory assessments during the first year and ≥60% adherence, estimated by proportion of days covered. Patients not meeting these criteria were excluded from longitudinal effectiveness analyses because of insufficient treatment exposure and/or lack of paired follow-up data for robust outcome assessment. Because case ascertainment relied on institutional pharmacy dispensing records and electronic clinical documentation, some patients obtaining GLP-1RAs privately outside this pathway may not have been captured. A comparative subanalysis was performed between 46 patients with chronic inflammatory arthritis and 43 patients with non-inflammatory/mechanical disease. The chronic inflammatory arthritis subgroup included RA (n=16), PsA (n=15), seronegative chronic arthritis (n=10), inflammatory bowel disease-associated arthritis (n=3) and spondyloarthritis (n=2). The non-inflammatory/mechanical group comprised osteoarthritis-related disorders (n=27), rotator cuff pathology (n=2), trochanteric pain syndrome (n=2), fibromyalgia (n=2), osteoporosis (n=2), Paget disease (n=1), Dupuytren disease (n=1), resolved statin-associated myalgia (n=1), non-Sjögren sicca syndrome (n=2), Raynaud phenomenon (n=1) and isolated antinuclear antibody positivity without other autoimmune manifestations (n=2). Patients with any concomitant immune-mediated inflammatory disease were excluded from the comparator group. Continuous variables are presented as mean±SD or median (IQR), and categorical variables as n (%). Intrapatient changes were analysed using paired t-tests or Wilcoxon signed-rank tests. Between-group comparisons used linear regression models adjusted for baseline value, age, sex and treatment indication, and are presented as adjusted mean differences with 95% CIs. Inflammatory marker analyses excluded measurements obtained during episodes of acute infection. All tests were two-sided, with p<0.05 considered significant. Analyses were performed using IBM SPSS Statistics V.23. Results Among the 11 788 adult patients followed in rheumatology outpatient clinics during the study period, 169 (1.4%) were identified as receiving GLP-1RA therapy and 119 met the predefined criteria for effectiveness analyses. Of the 119 included patients (61.3% female; mean age 62.2±11.0 years), 64.7% received GLP-1RAs for T2DM and 35.3% for obesity. Semaglutide was the most frequently prescribed agent (85.7%), followed by dulaglutide (12.6%), with liraglutide and tirzepatide accounting for 0.8% each. Mean weight was 95.6±19.0 kg, mean body mass index (BMI) was 34.9±9.9 kg/m²; T2DM was present in 96.6% of patients, with a mean HbA1c of 7.67±1.39%. Hypertension and dyslipidaemia were present in 74.8% and 75.6% of patients, respectively. The cohort comprised a heterogeneous spectrum of rheumatic diseases (figure 1). Distribution of rheumatological diagnoses among 119 patients receiving glucagon-like peptide-1 receptor agonists in rheumatology practice. Bars indicate patient numbers; percentages relative to total cohort. GCA, giant cell arteritis. Median follow-up was 19 months (IQR 11–35). GLP-1RA initiation was associated with early and sustained reductions in weight and BMI (figure 2A,B). By 12 months, weight loss reached −7.7 kg (−8.2%), with over half of patients improving by at least one WHO category. Significant parallel improvements were observed in HbA1c, fasting glucose and lipid profiles (figure 2C–F, detailed statistics in figure 2 legend). At 12 months, C reactive protein (CRP) decreased from 7.11±6.4 to 4.31±5.3 mg/L (p=0.076; n=46), while erythrocyte sedimentation rate (ESR) declined from 17.8±16.9 to 11.9±11.0 mm/hour (p=0.023; n=46). Longitudinal metabolic outcomes post-GLP-1RA initiation. (A) Weight, (B) BMI, (C) HbA1c, (D) fasting glucose, (E) LDL-cholesterol and (F) triglycerides. Data are mean±SD (shaded area); n=pairs per time point. GLP-1RA initiation was associated with sustained weight loss (median paired changes: −4.0 kg, −4.5 kg and −7.7 kg at 3, 6 and 12 months; all p<0.001; relative reductions: −4.1%, −4.9% and −8.2%). Of patients with serial BMI data, 56.5% improved by ≥1 WHO category. Parallel metabolic improvements included HbA1c reductions (−0.97%, −0.69% and −0.91% at 3, 6 and 12 months; all p<0.001) and decreased fasting glucose (−14.4 mg/dL at 3 m, −20.5 mg/dL at 6 m and −24.8 mg/dL at 12 m; p<0.001). LDL-C decreased by 12.2 mg/dL at 3 months (p=0.006) and triglycerides by 21.7 mg/dL at 6 months (p=0.007). BMI, body mass index; GLP-1RA, glucagon-like peptide-1 receptor agonist; HbA1c, haemoglobin A1c; LDL, low-density lipoprotein. Adverse events were mainly gastrointestinal (13.4%). Pancreatitis occurred in one patient and cholecystitis in two patients. Discontinuation due to adverse events occurred in 6.7% after a median exposure of 4.5 months. One man aged 74 years with multiple cardiovascular risk factors died from mesenteric ischaemia during follow-up. In the chronic inflammatory arthritis subgroup, median paired weight changes were −3.6 kg (IQR −8.1 to −1.7; n=14 pairs; p=0.0053) at 3 months, −6.2 kg (IQR −12.4 to −2.7; n=18 pairs; p=0.0016) at 6 months and −7.0 kg (IQR −12.9 to −2.8; n=18 pairs; p<0.001) at 12 months. HbA1c mean paired changes were −1.48%, −0.73% and −0.98%. At 12 months, CRP decreased from 5.1 mg/L (IQR 2.5–7.7) to 1.75 mg/L (IQR 0.8–4.0; p=0.0056; n=18 pairs), and ESR from 12.5 mm/hour (IQR 4.0–32.8) to 5.0 mm/hour (IQR 3.0–19.0; p=0.0178; n=18 pairs). Among patients with GC data, median dose remained 5.0 mg/day at 3 and 6 months, falling to 3.75 mg/day at 12 months; four patients discontinued GCs. Three mild-to-moderate flares occurred, all controlled with GCs without biologic or Janus kinase (JAK) inhibitor changes. Compared with the non-inflammatory group, the inflammatory subgroup had more frequent obesity indication (table 1). Adverse events were similar. Adjusted analyses showed no significant between-group differences in weight, HbA1c or fasting glucose trajectories, although weight loss was numerically greater in the non-inflammatory group at all time points. Table 1Baseline characteristics, rheumatological treatment and safety according to inflammatory versus non-inflammatory disease statusVariableInflammatoryN=46Non-inflammatoryN=43P valueFemale, n (%)30 (65.2%)29 (67.4%)1.000Age at GLP-1RA initiation, years61.0±12.262.1±10.30.834Indication: obesity, n (%)21 (45.7%)10 (23.3%)0.044Semaglutide, n (%)40 (87.0%)36 (83.7%)0.768Dulaglutide, n (%)5 (10.9%)6 (14.0%)0.753Liraglutide, n (%)1 (2.2%)01.000Tirzepatide, n (%)01 (2.3%)0.483Baseline weight, kg91.4±21.4 (n=40)95.3±23.6 (n=40)0.102Baseline BMI, kg/m235.3±6.7 (n=40)35.9±7.7 (n=39)0.294WHO BMI: overweight6 (15.0%)3 (7.7%)0.481WHO BMI: obesity grade I18 (45.0%)16 (41.0%)0.821WHO BMI: obesity grade II8 (20.0%)9 (23.1%)0.790WHO BMI: obesity grade III8 (20.0%)11 (28.2%)0.439Type 2 diabetes, n (%)44 (95.7%)43 (100.0%)0.495Hypertension, n (%)31 (67.4%)34 (79.1%)0.241Dyslipidaemia, n (%)36 (78.3%)32 (74.4%)0.804Cardiovascular disease, n (%)8 (17.4%)11 (25.6%)0.440Chronic kidney disease, n (%)7 (15.2%)7 (16.3%)1.000Hepatic steatosis, n (%)14 (30.4%)15 (34.9%)0.821Use of csDMARDs, n (%)26 (56.5%)–– Hydroxychloroquine3 (6.5%)–– Methotrexate19 (41.3%)–– Leflunomide5 (10.9%)––Use of bDMARDs, n (%)14 (30.4%)–– Abatacept1 (2.2%)–– Adalimumab4 (8.7%)–– Bimekizumab2 (4.3%)–– Etanercept1 (2.2%)–– Rituximab4 (8.7%)–– Secukinumab2 (4.3%)––Use of tsDMARDs, n (%)2 (4.3%)–– Upadacitinib1 (2.2%)–– Tofacitinib1 (2.2%)––Baseline corticosteroid use, n (%)15 (32.6%)2 (4.7%)<0.001Treatment duration, months18.5 (10.0–34.5)20.0 (12.0–38.5)0.535Adverse event5 (10.9%)5 (11.6%)1.000Serious adverse event001.000Discontinuation due to AE, n (%)4 (8.7%)2 (4.7%)0.678AE, adverse event; bDMARDs, biologic disease-modifying antirheumatic drugs; BMI, body mass index; csDMARDs, conventional synthetic disease-modifying antirheumatic drugs; GLP-1RA, glucagon-like peptide-1 receptor agonist; tsDMARDs, targeted synthetic disease-modifying antirheumatic drugs. Baseline characteristics, rheumatological treatment and safety according to inflammatory versus non-inflammatory disease status AE, adverse event; bDMARDs, biologic disease-modifying antirheumatic drugs; BMI, body mass index; csDMARDs, conventional synthetic disease-modifying antirheumatic drugs; GLP-1RA, glucagon-like peptide-1 receptor agonist; tsDMARDs, targeted synthetic disease-modifying antirheumatic drugs. Discussion In this single-centre real-world rheumatology cohort, GLP-1RA therapy coincided with early and sustained weight loss of clinical magnitude, alongside improvements in glycaemic control and lipid parameters. These metabolic responses are broadly consistent with the wider obesity and T2DM literature and appear particularly relevant in a population with high baseline cardiometabolic burden and frequent GC exposure.4–8 From a rheumatological perspective, obesity amplifies cardiovascular risk and worsens inflammatory activity, pain and disability across multiple conditions.1 In RA and PsA specifically, higher body weight correlates with increased disease activity and reduced therapeutic response, while weight loss associates with clinical improvement.2 3 The magnitude of weight loss (−8.2% at 12 months) was attenuated compared with randomised trials of semaglutide (Semaglutide Treatment Effect in People with Obesity [STEP-1]: −14.9%) and tirzepatide (SURMOUNT-1, Tirzepatide Once Weekly for the Treatment of Obesity: −20.9%)11 but aligns closely with real-world reports, such as Rodriguez et al.12 who observed reductions of −5.8% with semaglutide and −11.7% with tirzepatide in routine practice. This attenuation is plausible given our cohort’s diagnostic heterogeneity, high comorbidity burden, frequent GC use, mixed treatment indications (obesity vs diabetes) and less standardised titration than in trial settings. Inflammatory marker analyses revealed numerically reduced CRP and significantly lower ESR at 12 months. In the chronic inflammatory arthritis subgroup, both markers declined significantly, consistent with emerging meta-analytic evidence of GLP-1RA effects on low-grade inflammation primarily in T2DM populations.8 These observations are hypothesis-generating but limited by small numbers, the absence of standardised disease activity measures, possible GC adjustments and intercurrent inflammatory confounders. They neither establish causality nor rule out weight loss-mediated effects on acute-phase reactants. Accordingly, the observed reductions in ESR and CRP should not be interpreted as direct evidence of improved rheumatic disease control. The clinical literature on GLP-1RAs in rheumatology remains focused on symptomatic/structural rather than biomarker end points. In fibromyalgia, TriNetX data suggested reduced symptom burden and opioid use13; STEP-9 demonstrated weight, pain and function benefits in knee osteoarthritis.14 Our findings address a specific gap regarding real-world metabolic responses in rheumatology practice and the potential influence of chronic inflammation, although definitive conclusions require larger studies. Adjusted between-group comparisons showed no statistically significant differences in weight, HbA1c or glucose trajectories between patients with chronic inflammatory arthritis (n=46) and non-inflammatory controls (n=43). Although numerical trends favoured the non-inflammatory group, the wide CIs suggest limited statistical power and warrant cautious interpretation. Baseline GC use (32.6% vs 4.7%) and obesity indication (45.7% vs 23.3%) differed substantially, although regression adjustment mitigated measured confounding. These analyses should be interpreted as hypothesis-generating given sample size constraints, diagnostic heterogeneity within inflammatory arthritis and unmeasured factors including disease activity and physical function. Preliminary phase IIIb TOGETHER-PsA (Efficacy and Safety of Ixekizumab or Ixekizumab Concomitantly Administered With Tirzepatide in Adult Participants With Active Psoriatic Arthritis and Obesity or Overweight) data support the plausibility of an immunometabolic strategy in PsA, where ixekizumab plus tirzepatide achieved superior American College of Rheumatology 50% improvement criteria and ≥10% weight loss versus ixekizumab monotherapy.15 Although top-line results require peer-reviewed confirmation, they align conceptually with our metabolic findings and suggest potential synergy with background disease-modifying antirheumatic drug therapy. Safety findings were dominated by expected gastrointestinal events, consistent with class pharmacology. The mesenteric ischaemia event should be interpreted cautiously. A causal relationship with GLP-1RA therapy cannot be established in our cohort, particularly given the patient’s multiple comorbidities and cardiovascular risk factors. Moreover, the most robust evidence currently available, a recent systematic review and meta-analysis including 55 randomised controlled trials and 106 395 participants, concluded that GLP-1RAs probably have little or no effect on the risk of gastrointestinal ischaemia compared with placebo, with increased risk observed only for cholelithiasis and gastro-oesophageal reflux disease.16 Absence of immune-mediated signals (lupus-like, vasculitis)17–19 is reassuring but severely underpowered (n=119, 19 months median follow-up cannot exclude events <1%). This observation aligns with the NYU retrospective lupus cohort,20 where GLP-1RA use coincided with significant BMI reduction without increased flare rates beyond expected background levels. However, direct comparability is limited by key population differences: the lupus cohort received GLP-1RAs almost exclusively for weight loss, was younger and had a lower cardiometabolic burden and GC exposure. During follow-up, three mild-to-moderate flares occurred in the inflammatory subgroup, all successfully managed with GCs without requiring biologic or JAK inhibitor modification. This study adds real-world evidence on the use of GLP-1RAs in rheumatology practice, showing meaningful metabolic benefit together with a possible inflammatory signal. Interpretation is limited by the retrospective single-centre design, the modest sample size, diagnostic heterogeneity and the absence of standardised disease activity assessment. GLP-1RAs may therefore represent a useful cardiometabolic option in selected patients with rheumatic diseases, although prospective comparative studies are needed to better define their immunometabolic role and their independent effects on inflammatory disease activity. Conclusions GLP-1RA therapy was associated with clinically meaningful weight loss and glycaemic improvement in this real-world rheumatology cohort. Exploratory analyses suggested potential reductions in inflammatory markers, although these findings require confirmation. No major safety signals emerged, but detection of rare events was limited. Prospective studies are needed to clarify the immunometabolic role of GLP-1RAs in rheumatology. Acknowledgements The authors would like to thank the patients who participated in this study. Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Patient consent for publication: Not applicable. Provenance and peer review: Not commissioned; externally peer reviewed. Ethics approval: This study was approved by the Ethics Committee of Complejo Asistencial Universitario de León (reference: 26078). As this was a retrospective study using anonymised data, the requirement for informed consent was waived. Patient and public involvement statement: Patients or the public were not involved in the design, conduct, reporting, or dissemination plans of this study. References