Cardiovascular and mortality outcomes with GLP-1 receptor agonists vs other glucose-lowering drugs in individuals with NAFLD and type 2 diabetes: a large population-based matched cohort study

Aims/hypothesis We aimed to determine whether the use of glucagon-like peptide-1 receptor agonists (GLP-1RA) in individuals with non-alcoholic fatty liver disease (NAFLD) and type 2 diabetes mellitus decreases the risk of new-onset adverse cardiovascular events (CVEs) and mortality rate compared with other glucose-lowering drugs in a real setting at a population level. Methods We conducted a population-based propensity-matched retrospective cohort study using TriNetX. The cohort comprised patients over 20 years old who were newly treated with glucose-lowering drugs between 1 January 2013 and 31 December 2021, and followed until 30 September 2022. New users of GLP-1RAs were matched based on age, demographics, comorbidities and medication use by using 1:1 propensity matching with other glucose-lowering drugs. The primary outcome was the new onset of adverse CVEs, including heart failure, composite incidence of major adverse cardiovascular events (MACE; defined as unstable angina, myocardial infarction, or coronary artery procedures or surgeries) and composite cerebrovascular events (defined as the first occurrence of stroke, transient ischaemic attack, cerebral infarction, carotid intervention or surgery), and the secondary outcome was all-cause mortality. Cox proportional hazards models were used to estimate HRs. Results The study involved 2,835,398 patients with both NAFLD and type 2 diabetes. When compared with the sodium–glucose cotransporter 2 (SGLT2) inhibitors group, the GLP-1RAs group showed no evidence of a difference in terms of new-onset heart failure (HR 0.97; 95% CI 0.93, 1.01), MACE (HR 0.95; 95% CI 0.90, 1.01) and cerebrovascular events (HR 0.99; 95% CI 0.94, 1.03). Furthermore, the two groups had no evidence of a difference in mortality rate (HR 1.06; 95% CI 0.97, 1.15). Similar results were observed across sensitivity analyses. Compared with other second- or third-line glucose-lowering medications, the GLP-1RAs demonstrated a lower rate of adverse CVEs, including heart failure (HR 0.88; 95% CI 0.85, 0.92), MACE (HR 0.89; 95% CI 0.85, 0.94), cerebrovascular events (HR 0.93; 95% CI 0.89, 0.96) and all-cause mortality rate (HR 0.70; 95% CI 0.66, 0.75). Conclusions/interpretation In individuals with NAFLD and type 2 diabetes, GLP-1RAs are associated with lower incidences of adverse CVEs and all-cause mortality compared with metformin or other second- and third-line glucose-lowering medications. However, there was no significant difference in adverse CVEs or all-cause mortality when compared with those taking SGLT2 inhibitors. Graphical Abstract Supplementary Information The online version contains peer-reviewed but unedited supplementary material available at 10.1007/s00125-023-06057-5.


ESM Fig 6:
Propensity score density graph for the users of glucagon-like peptide-1 receptor agonists versus the second-or third-line glucose-lowering medication users among patients with nonalcoholic fatty liver disease and type 2 diabetes before and after propensity score matching.

ESM Fig 7:
Cardiovascular outcomes and all-cause mortality between the new users of GLP-1RAs vs. second-or third-line glucose-lowering drugs in patients with nonalcoholic fatty liver disease and type 2 diabetes mellitus.

ESM Table 1:
Laboratory findings of new GLP-1A users vs. matched new SGLT-2i users in patients with NAFLD and type 2 diabetes.

ESM Table 2:
Baseline characteristics of new GLP-1A users vs matched new metformin users in patients with NAFLD and type 2 diabetes.3: Laboratory findings of new GLP-1A users vs matched new metformin users in patients with NAFLD and type 2 diabetes.4: Baseline characteristics of new GLP-1RA users vs. matched new Second-orthird-line glucose-lowering users in patients with NAFLD and type 2 diabetes.5: Laboratory findings of new GLP-1RA users vs. matched new Second -or third-line glucose-lowering drug users in patients with NAFLD and type 2 diabetes.7: Sensitivity analysis to assess the cardiovascular outcomes and all causes of mortality between the new users GLP-1RA vs. Metformin in patients with NAFLD and type 2 diabetes after excluding index events within 2 years after the index date.

ESM Table
ESM Table 8: Sensitivity analysis to assess the cardiovascular outcomes and all causes of mortality between the new users GLP-1RA vs. Second-or third-line glucose-lowering medications in patients with NAFLD and type 2 diabetes after excluding index events within 2 years after the index date.

Data Source
We used TriNetX (Cambridge, MA, USA), a global federated health research network providing real-time access to electronic health records (EHRs).TriNetX platform de-identifies and aggregates EHR data from 66 healthcare organizations (HCOs), most of which are large academic medical institutions with both inpatient and outpatient facilities at multiple locations across 50 states in the United States.The Real-time access to health insurance portability and accountability act-deidentified, compliant, and longitudinal clinical data to member HCOs is provided cloud-based.The deidentified clinical data, such as diagnoses, procedures, medications, laboratory values, and genomic information, are continuously aggregated directly from the EHR of the participating HCOs.
Participating HCOs include outpatient, inpatient, and specialty care services and provide care to a diverse patient population from different ethnicity, age groups, geographical region, and income levels.Both the patients and HCOs, as data sources, stay anonymous.
As a federated network, TriNetX data have been granted a waiver from the Western institutional review board since only aggregated counts and statistical summaries of de-identified information without any protected health information were received from participating HCOs.In addition, no study-specific activities are performed in retrospective analyses.

Standardizing the terminology and data quality check:
The TriNetX software verifies the basic formatting to confirm that data are appropriately characterized.Patient counts were rounded up to the nearest 10 in our analysis to safeguard protected health information.TriNetX has production capabilities that have been tested that map data extensively from each of these structures to the standard model within TriNetX and can extract details of interest from the narrative content of clinical documents using natural language processing.The

Selection of Patients:
The search was conducted following the criteria provided by TriNetX to identify potential patients.These codes included the ICD-9, 10 CM.We combined patients into a single cohort of NAFLD/NASH defined by a diagnosis of NAFLD based on ( 571.8, K76.0, and K75.81).

ESM Fig 5 :
Study Flow Chart of Patient Selection in the Study Cohort for New Users of Glucagon-Like Peptide-1 Receptor Agonist and New Users of second-or third-line glucoselowering drugs (Active-Comparator).
contributing EHR systems used United Medical Language System (UMLS) for coding.TriNetX maps the data to a standard and controlled set of clinical terminologies, for example, mapping disease terms from Systematized Nomenclature of Medicine-Clinical Terms (SNOMED CT) to International Classification of Diseases, and Clinical Modification (ICD-9 and 10 CM), drug terms from National Drug Codes (NDCs) to RxNorm.TriNetX enforces a list of required fields (e.g., patient identifier) and rejects those records where the required data is lacking.Referential integrity checking confirms that data spanning multiple database tables can be successfully joined together.TriNetX requires at least 1 non-demographic fact for a patient to be calculated in a given data set.Patient records with only demographic information are not included in data sets.As the data are refreshed, the TriNetX software monitors change in data volumes over time to ensure data validity.

Fig 1 :
Propensity score density graph for the users of glucagon-like peptide-1 receptor agonists (Purple line) versus sodium-glucose cotransporter-2 inhibitors (Green line) users among patients with nonalcoholic fatty liver disease and type 2 diabetes before and after propensity score matching.ESM Fig 2: Study Flow Chart of Patient Selection in the Study Cohort for New Users of Glucagon-Like Peptide-1 Receptor Agonist and New Users of Metformin (Active-Comparator).NAFLD, nonalcoholic fatty liver disease; T2DM, type 2 diabetes mellitus ;GLP-1RA, glucagon-like peptide-1 receptor agonists; eGFR, estimated glomerular filtration rate.Propensity score density graph for the users of glucagon-like peptide-1 receptor agonists (Purple line) versus metformin (Green line) users among patients with nonalcoholic fatty liver disease and type 2 diabetes before and after propensity score matching.ESM Fig 4: Cardiovascular outcomes and all-cause mortality between the new users of GLP-1RAs vs. metformin in patients with nonalcoholic fatty liver disease and type 2 diabetes mellitus.GLP-1RA, glucagon-like peptide-1 receptor agonists; Major adverse cardiac events as a composite endpoint was defined as the first occurrence of unstable angina, or myocardial infarction, and revascularization, including percutaneous coronary intervention or coronary artery bypass graft.A composite of cerebrovascular disease was defined as the first occurrence of stroke (ischemic or hemorrhagic stroke), cerebral infarction, transient ischemic attack, carotid intervention, or surgery.ESM Fig 5: Study Flow Chart of Patient Selection in the Study Cohort for New Users of Glucagon-Like Peptide-1 Receptor Agonist and New Users of second-or third-line glucose-lowering medication (Active-Comparator).NAFLD, nonalcoholic fatty liver disease; T2DM, type 2 diabetes mellitus ;GLP-1RA, glucagon-like peptide-1 receptor agonists; eGFR, estimated glomerular filtration rate.Adult Patients With Diagnosis of NAFLD and T2DM N = 4591936 Excluded: • <20 years of age • Patients who received insulin monotherapy • Female patients with a history of polycystic ovary syndrome or gestational diabetes • Patients with follow-up less than 1 year after the cohort Study period: January 1, 2013, and September 31, 2022.Patients With NAFLD and T2DM who had received GLP-1RA or New Users of Second-or Third Line Drugs N = 3691936 (During the study period) New Users of Second-or Third Line Drugs N = 2638687 Excluded: • Patients who had a diagnosis of cancer, organ transplantation, or dialysis • Patients with eGFR < 30 mL/min/1.73m 2 within 6 months before the cohort entry • Patients with a history of adverse cardiovascular events prior to cohort entry 1:1 Propensity Score Matching to Reduce the Risk of Bias Attributed to Confounding New Users of GLP-1RA N = 52166 New Users of Second-or Third Line Drugs N = 52166 New Users of GLP-1RAs N = 53249 ESM Fig 6: Propensity score density graph for the users of glucagon-like peptide-1 receptor agonists (Purple line) versus the second-or third line glucose-lowering medication (Green line) users among patients with nonalcoholic fatty liver disease and type 2 diabetes before and after propensity score matching.ESM Fig 7: Cardiovascular outcomes and all-cause mortality between the new users of GLP-1RAs vs. second-or third line antidiabetic drugs in patients with nonalcoholic fatty liver disease and type 2 diabetes mellitus.GLP-1RA, glucagon-like peptide-1 receptor agonists; cholangitis.Diagnosis codes based on the exclusion criteria are listed below.070, Viral Toxic liver disease with fibrosis and cirrhosis of the liver.

Table 2 : Baseline characteristics of new GLP-1A users vs matched new metformin users in patients with NAFLD and type 2 diabetes Variables Before propensity score match After propensity score match
Abbreviations: NAFLD, nonalcoholic fatty liver disease, GLP-1Ras, glucagon-like peptide-1 receptor agonists; SMD, standard mean difference; SD, standard deviation; BMI, body mass index; ACE, angiotensin-converting-enzyme.ESM

Table 5 : Laboratory findings of new GLP-1RA users vs matched new Second -or third-line drugs users in patients with NAFLD and type 2 diabetes Variables
filtration rate ; LDH, lactate dehydrogenase; CRP, ; CH, cholesterol; TG, triglycerides; LDL, low-density lipoprotein ; HDL, high-density lipoprotein.ESM

Table 6 : Sensitivity analysis to assess the cardiovascular outcomes and all cause of mortality between the new users of GLP-1RA vs. SGLT-2 inhibitors in patients with NAFLD and type 2 diabetes after excluding index events within 2 years after the index date
GLP-1RA, glucagon-like peptide-1 receptor agonists; SGLT-2 sodium-glucose cotransporter-2; HR, hazard ratio; MACE, Major adverse cardiovascular events; NAFLD, nonalcoholic fatty liver disease.