Plaintiffs' leadership told Judge Karen Marston in 2024 that more than 95% of the cases in MDL 3094 would allege gastroparesis. The public docket cannot say who those plaintiffs are. The short-form complaint collects residence, drug and injury, and the plaintiff fact sheets that carry age, sex and medical history are served on the defendants and never filed. The nearest public proxy is the FDA's adverse event database, and it answers a narrower question: who has been telling the FDA that a GLP-1 drug paralyzed their stomach.
The pull below queried the openFDA drug adverse event API on October 7, 2026, against data current through July 30, 2026. It matched any report naming semaglutide, tirzepatide, liraglutide or dulaglutide, by generic or brand name, with a reaction coded as impaired gastric emptying or gastroparesis.
FAERS codes nearly all of these reports to the first term. The database carries 5,672 reports of "impaired gastric emptying" on these drugs and a few dozen that say gastroparesis. A search on the clinical term alone returns almost nothing, and that coding gap is one reason the figure below has not circulated.
The reports track the docket, not the prescriptions
GLP-1 prescriptions have climbed since 2018. Stomach paralysis reports did not climb with them. From 2015 through 2022 the database received between 33 and 87 such reports a year. Jaclyn Bjorklund filed the first gastroparesis suit in August 2023, the Judicial Panel centralized the cases in February 2024, and the reporting curve bent at the same moments.
| Year | Reports |
|---|---|
| 2021 | 68 |
| 2022 | 87 |
| 2023 | 256 |
| 2024 | 992 |
| 2025 | 2,536 |
| 2026 through July 30 | 1,446 |
Reports received in 2024 or later total 4,974, which is 87% of the cohort. Add 2023 and the share rises to 92%. The 2026 figure covers seven months and is on pace to match 2025. Whatever the drugs do to the stomach, the database is measuring the litigation's intake machinery at least as much as it is measuring injury.
One report in five came from a lawyer
FAERS records who sent each report. Consumers sent 70% of the stomach paralysis reports, lawyers sent 21%, physicians 4%, other health professionals 4% and pharmacists 1%. The lawyer share is the number to hold onto. Across the 397,291 GLP-1 adverse event reports of every kind in the database, lawyers sent 1%. The ileus reports run 17% lawyer-filed and the intestinal obstruction reports 13%. The gastroparesis cohort is the most lawyer-driven of the three injuries, and it is the injury the MDL is built on.
The consumer share needs a second look. FAERS labels a report "consumer" when the patient or a family member sent it. Intake firms routinely tell signed clients to send their own report, and a report sent that way is counted as a consumer report, not a lawyer report. Some of the 70% is therefore litigation-driven too. The 21% lawyer share is the floor for reports the lawsuits generated. The true figure is higher, and the database cannot say by how much.
Who the patient is
Sex is recorded on 74% of the reports and age on 78%. Within the reports that carry them, the patient is a woman 73% of the time. The median age is 54. The fifties hold 34% of the cohort, the sixties 24%, the forties 22%, and only 8% of patients are 70 or older. The table sets the gastroparesis patients beside the two other MDL injuries and beside the full GLP-1 reporting base.
| Reaction | Reports | Female | Median age | Lawyer-filed |
|---|---|---|---|---|
| All GLP-1 reports | 397,291 | 68% | 59 | 1% |
| Impaired gastric emptying or gastroparesis | 5,699 | 73% | 54 | 21% |
| Ileus | 1,188 | 57% | 58 | 17% |
| Intestinal obstruction | 2,435 | 59% | 59 | 13% |
The gastroparesis patients are younger and more female than the ileus and obstruction patients, and younger than the GLP-1 reporting base as a whole. That pattern fits a cohort with a weight-management population mixed into the diabetic one.
The indication field says the diabetic population still dominates. Type 2 diabetes appears as the stated reason for a drug on 3,866 lines of these reports, and weight control on 1,057. The indication is recorded per drug rather than per patient, so those are mentions and not people, but a near four-to-one ratio does not come from a weight-loss crowd.
Ozempic carries the cohort, Wegovy barely registers
Ozempic is named on 3,436 reports, 60% of the cohort. Trulicity follows at 1,832, then Mounjaro at 1,484, Victoza at 526, Wegovy at 327 and Rybelsus at 307. The shares add to more than 100% because a report can name more than one drug. The overlap is the switchers, patients moved from one GLP-1 to another before the diagnosis.
Wegovy and Zepbound are the products with the weight-loss label and the television budget, and together they account for under 6% of the stomach paralysis reports. The patient in this database is a woman in her fifties taking Ozempic, Trulicity or Mounjaro for diabetes, more often than she is anyone else.
What the profile means in court
The defense will read the year table as stimulated reporting. That is the pharmacovigilance term for a spike driven by publicity and litigation rather than by a change in the drug's behavior. The reading has force. The reports began climbing with the lawsuits, lawyers sent a fifth of them, and FAERS cannot verify a diagnosis.
Judge Marston's August 15, 2025 Rule 702 ruling requires a gastric emptying study at diagnosis for any drug-induced gastroparesis claim. That ruling already reflects the court's concern that gastroparesis is being alleged without objective confirmation. The September 14 to 18 evidentiary hearing on warnings and general causation is where the reporting record will be argued, and this database is part of it.
Plaintiffs can read the same table the other way. Stimulated reporting is a known artifact of every major pharmaceutical mass tort, and it says nothing about whether the underlying injury is real. The pre-litigation baseline was not zero. The database recorded 33 to 87 stomach paralysis reports a year on these drugs for eight years before anyone sued. The drug class works by slowing gastric emptying.
What the profile means for inventory
For a firm screening GLP-1 intake, the FAERS profile is a benchmark for the files already signed. An inventory that is 73% women with a median age in the mid-fifties and Ozempic on most files looks like the reporting base. An inventory dominated by Wegovy weight-loss users in their thirties does not. A firm holding one should ask why its claimants differ from the population that reported the injury to the FDA.
The gastric emptying study requirement is the other filter. FAERS says nothing about which of these 5,699 patients had one. A file with the right demographic profile and no gastric emptying study at diagnosis is a file the court has already told the parties how to treat.
What the database cannot say
FAERS counts reports, and one patient can generate several. Manufacturers forward the reports they receive, the patient's lawyer may file, and the physician may file, each as a separate record. The counts here are deduplicated by openFDA's case identifier, which does not catch every repeat. Reporting is voluntary, so the base rate of injury cannot be computed from it. Age is missing from 22% of the cohort and sex from 26%. A report that names two drugs is counted under both.
None of those limits changes the shape of the year curve or the lawyer share, which are the two findings that matter. Both are reproducible from the public API by anyone with a free key.