BMJ fact-check says UK semaglutide death headlines rest on suspected-reaction reports

The BMJ has examined UK news coverage that tied GLP-1 drugs to hundreds of deaths and thousands of adverse events, and explained that the totals come from a system that records suspected reactions rather than confirmed harms ^[1].

The fact-check responds to coverage led by a story in The Times on 5 August 2026, which reported that the drugs had been linked to hundreds of deaths and thousands of adverse events in the UK ^[1]. The BMJ's piece, dated 22 August 2026, walked through what the underlying numbers do and do not show ^[1].

The figures come from the Yellow Card scheme, the UK system for collecting reports of suspected side effects. A report in that scheme is a suspicion logged by someone, not a finding that a drug caused the problem. The BMJ said the headline totals should not be read as confirmed harms ^[1].

In the totals the BMJ reviewed, semaglutide — sold as Ozempic, Wegovy and Rybelsus — accounted for 3,454 serious reports in women and 1,216 in men ^[1]. Fatal outcomes were nearly even between the sexes, at 28 reports in women and 29 in men ^[1]. The deaths were clustered in the 60 to 69 age band ^[1].

The fact-check does not conclude that GLP-1 drugs caused any of the deaths it describes. Whether the drugs played a role in those cases has not been established ^[1]. The BMJ also did not report a UK regulator conclusion that the medicines caused the reported deaths.

A genetics pilot has stopped enrolling

Genomics England and the UK's Medicines and Healthcare products Regulatory Agency have closed recruitment to a pilot called the Yellow Card Biobank ^[1]. The project is testing whether a person's genetics can predict side effects from GLP-1 medicines ^[1]. Analysis of the collected data is under way, and no results or publication date have been disclosed ^[1].

Why it matters for patients

Large report counts travel faster than the caveats attached to them. In the UK scheme the BMJ describes, a report means someone suspected a link between a drug and a health problem. It does not mean the drug caused that problem ^[1]. The same caution applies to any tally of reported reactions.

Two patterns in the numbers stand out, and neither is explained by the fact-check. Serious reports were far more common in women than in men — 3,454 versus 1,216 — but fatal reports were almost identical, at 28 and 29 ^[1]. The BMJ did not say why serious reports skew female while fatal reports do not, or whether reporting behavior, prescribing patterns or biology drives the difference.

Deaths clustered among people aged 60 to 69 ^[1]. The fact-check does not say whether age by itself raises risk, or whether other health conditions common in that age group were involved. That question has not been answered in the piece.

The fact-check covers UK reports only. Comparable US reporting figures and any US regulatory review of the same question were not part of it ^[1]. It also does not include guidance on symptoms to watch for or when to seek care.

What happens next

Analysis of the Yellow Card Biobank pilot is under way after recruitment closed ^[1]. No date has been disclosed for those results. It also has not been reported whether the MHRA plans to change labels, warnings or prescribing guidance in response to the Yellow Card totals it holds ^[1]. Until then, the central question raised by the UK headlines — whether GLP-1 drugs caused any of the reported deaths — remains unanswered in the material the BMJ reviewed ^[1].