PERFORMANCE REVIEW
Semaglutide (GLP-1 Receptor Agonist) — Outcome Documentation Audit
Period Under Review: 2018–2026
Issuing Division: Secondary Endpoint Documentation, Unlabeled Outcomes Division
Classification: Field Note — Post-Authorization Evidence Accumulation
The label is complete.
That is the first thing the Bureau wishes to establish. The official prescribing information for Wegovy — semaglutide, manufactured by Novo Nordisk — is a thorough document. It lists nausea, diarrhea, vomiting, constipation, stomach pain, headache, fatigue, dizziness, bloating, and heartburn. It lists thyroid tumors, pancreatitis, gallbladder problems, hypoglycemia, kidney problems, increased heart rate, suicidal ideation, and aspiration risk. It runs several pages.
It does not list the marriages.
The Bureau is not suggesting that something was omitted. The Bureau is noting that the label's scope is physiological, its endpoints were weight loss and cardiovascular risk, its trials were designed accordingly, and anything that occurred outside those endpoints is, in the formal language of pharmacovigilance, not an adverse event. It is not classified at all. It occupies a category the label does not have a name for.
The Bureau is proposing a name: Undocumented Secondary Outcome, Social Architecture Class.
I. Primary Outcome — Weight Reduction
Rating: ACHIEVED
The drug performs its approved function. In clinical trials, patients taking semaglutide lost on average 15–17% of body weight. GLP-1 prescriptions have more than tripled since 2020. The RAND Corporation documented in August 2025 that approximately 15 million Americans are currently taking GLP-1 drugs and approximately one in eight Americans has tried them. Pharmacy records show over 19 million patients.
The primary endpoint was weight. The drug reached the primary endpoint. This section requires no further analysis.
II. Secondary Outcome — Marital Stability
Rating: NOT APPLICABLE (ENDPOINT NOT REGISTERED)
In 2018, Bruze et al. published a study in JAMA Surgery following two large Swedish cohorts — approximately 31,000 gastric bypass patients and their matched controls — for up to ten years. The study was not investigating divorce. The study was investigating quality of life, relationship status, and psychosocial outcomes following bariatric surgery as a proxy for rapid weight loss.
It found that married patients who underwent bariatric surgery had an adjusted hazard ratio of 1.41 for divorce versus controls. In the primary cohort, the ten-year cumulative incidence of divorce was 17.1% for surgery patients versus 11.6% for controls. A separate US multicenter study — the LABS-2 prospective cohort, published in 2022 — found that 13% of married bariatric patients experienced separation or divorce within five years.
The mechanism, per Per-Arne Svensson of the Sahlgrenska Academy at the University of Gothenburg and lead researcher on the primary study, is not pharmacological. The drug does not act on the marriage. The drug acts on the patient. The patient, in a condition of improved self-esteem, expanded social options, and lifestyle transformation, then acts on the marriage. Svensson describes the result as "couples drifting apart" and notes that enhanced confidence may "empower individuals to leave unhealthy relationships."
The marriage was not harmed by the drug. The marriage was load-bearing on the previous configuration. The configuration changed.
BUREAU NOTE: The Bureau's records show that in the bariatric surgery cohorts, approximately 13–17% of married patients divorced within five to ten years of treatment. The Bureau's records also show that the general-population divorce rate over the same period was approximately 8–11%. The Bureau's records do not show a corresponding entry in the drug label. The Bureau's records show only that the label's scope is physiological. Whether "scope" and "record" are the same word is a question the Bureau has forwarded to the appropriate lexical authority. No response has been received.
III. Scale Assessment — Trial Population vs. Current Deployment
Rating: PHASE 2 ENROLLMENT SUBSTANTIALLY EXCEEDS PHASE 1 PARAMETERS
The Bruze et al. study and the LABS-2 cohort were conducted on populations numbering in the tens of thousands. Bariatric surgery is performed on approximately 250,000 to 280,000 Americans per year.
GLP-1 drugs are currently being taken by 15 million Americans.
The marriage-dissolution signal was detected at a scale of tens of thousands. It is now running through a population roughly fifty times larger. The original studies were the pilot. GLP-1 adoption is the rollout. The Bureau notes that no adjustment to the trial design has been recommended, as the trial was not a trial and the rollout is not a study.
IV. Mechanism — Constraint Removal, Relational Applications
Rating: CONSISTENT WITH APPROVED FUNCTION
The drug removes a metabolic constraint. This is what the drug is designed to do. The drug removes the constraint. The patient's circumstances reorganize around the new configuration. In some cases, the patient's body is not the only thing that was load-bearing on the old weight.
Svensson told Fox News in April 2026 that weight loss "may increase a sense of autonomy or confidence, potentially empowering individuals to leave unhealthy relationships." He described a staged timeline: substantial weight loss in year one, followed by relationship dynamic changes, followed eventually by separation or divorce.
The drug produced a transformed patient. The transformed patient produced a transformed life. The label describes the first transformation. The second transformation is in the peer-reviewed literature. The connection between them is in the newspapers. The gap between the newspapers and the label is what the Bureau is auditing.
BUREAU NOTE: In April 2026, coverage of the Swedish study's extrapolation to GLP-1 users ran across Fox News, the Daily Beast, Yahoo, LAD Bible, AOL, and smaller outlets. Across all of them, the framing was consistent: relationship advice. What should couples do? How do you Ozempic-proof your marriage? The question of why a documented, peer-reviewed, cross-nationally replicated outcome does not appear in the drug's official documentation was not asked. The Bureau does not interpret this as journalistic failure. The Bureau interprets it as accurate jurisdictional assignment. The label belongs to pharmacovigilance. The marriage belongs to the lifestyle section. The gap between the two sections belongs to no one.
V. Novo Nordisk Comment
Rating: NOT FILED
No comment from Novo Nordisk on the marriage dissolution data has been located in the public record.
The Bureau notes this without characterizing it as silence, evasion, or indifference. The Bureau notes it as a data field that is empty. The field was not created by the manufacturer. The field was created by the research. The manufacturer has no obligation to populate a field that did not exist in its original documentation architecture.
The label is complete.
VI. Bureau Determination
The Bureau of Secondary Endpoint Documentation, Unlabeled Outcomes Division, finds as follows:
The drug was tested against weight loss. It passed.
The drug was not tested against marriage stability. The bariatric surgery cohorts ran a proxy test, retrospectively, without design or registration, in a population now known to be approximately fifty times smaller than the current GLP-1 user base. That test found an elevated divorce rate. The finding is peer-reviewed, cross-nationally replicated, and statistically significant. It is not on the label.
The Bureau does not conclude from this that the label is wrong. The Bureau concludes that the label was asked a specific question and answered it correctly. A second question was never asked. The answer to the second question has been accumulating in the literature for eight years.
The Bureau is marking the secondary outcome field: Not Applicable — Endpoint Not Registered.
The Bureau acknowledges that "not applicable" and "not found" are different entries. The Bureau acknowledges that the current field contains neither.
BUREAU NOTE: The Bureau wishes to clarify, for any reader who has reached this section expecting a recommendation: the Bureau does not make recommendations about marriages. The Bureau does not make recommendations about drugs. The Bureau documents gaps between what a system was designed to measure and what it apparently measured anyway. The Bureau files the gap. What happens inside the gap is outside the Bureau's jurisdiction, which the Bureau considers one of its better-designed features.
Bureau of Secondary Endpoint Documentation, Unlabeled Outcomes Division — a sub-bureau of the Bureau of Public Agreement, established at the intersection of the approved indication and everything adjacent to it.