Skip to content
Compare prices

Independent UK evidence check

GLP-1 safety headlines: what the evidence actually says

Can weight-loss injections cause blindness, pancreatitis or “stomach paralysis”? Some frightening headlines begin with a genuine safety signal, but leave out the scale, uncertainty or the fact that a finding applies to one medicine rather than the whole GLP-1 class.

Evidence checked: 20 August 2026 Reading time: about 12 minutes UK sources prioritised No sponsored conclusions

Our position: we will not dismiss a real safety concern to defend a medicine, and we will not turn a suspected report into a proven fact to win a click. This page compares the reporting with regulator conclusions, product information and peer-reviewed evidence. It is general information, not personal medical advice.

Reader controls

Choose how much detail you want

Showing the complete analysis, headline examples and source detail.

60-second evidence check

What are you concerned about?

Sudden vision loss: the short answer

What is knownNAION is a very rare recognised risk associated with semaglutide. The UK estimate is about one additional case per 10,000 people treated for one year.

Other risk factorsType 2 diabetes itself raises NAION risk. Smoking, high blood pressure and high cholesterol are also risk factors; they do not cancel the semaglutide finding.

What to doSudden sight loss or rapidly worsening vision needs urgent assessment through eye casualty or A&E.

The short version

The headlines are not all false. They are often incomplete.

What alarmist coverage gets wrong

A report is not automatically a confirmed side effect

Yellow Card and other adverse-event systems are designed to catch early warning signals. A report means somebody suspected a connection. It does not, by itself, prove the medicine caused the event or reveal how often it occurs.

What dismissive coverage gets wrong

“Rare” does not mean imaginary or unimportant

A very rare risk can still be serious for the person affected. Good communication gives the scale and the action: what symptoms matter, how quickly to seek help and which particular medicine the evidence concerns.

Cause, contribution and coincidence

Underlying health matters, but it does not settle whether a medicine contributed

People prescribed GLP-1 medicines often already have conditions that independently raise some of the same risks reported after treatment. That makes careful comparison essential. It does not mean every event was caused by the medicine, and it does not mean every event can be dismissed as the patient’s existing illness.

NAION / vision loss

A raised background risk and a drug association can both be true

Type 2 diabetes independently increases NAION risk; smoking, high blood pressure and high cholesterol also raise risk. After considering the wider evidence, the MHRA still judged semaglutide to be associated with about a two-fold relative increase, roughly one additional case per 10,000 people treated for a year.

Acute pancreatitis

Other established causes must be considered

NHS guidance identifies gallstones and heavy alcohol use as the most common causes of acute pancreatitis. Other illnesses, injury, surgery and some medicines can also cause it. A suspected report may not contain enough information to determine the contribution of each factor.

Gastroparesis

Diabetes can itself slow stomach emptying

NHS guidance identifies long-term diabetes as an established cause of gastroparesis; some surgery can cause it, and sometimes no cause is found. This overlap is one reason observational associations require cautious interpretation and symptoms need a clinical assessment.

Mental health

Timing alone cannot establish cause

Distress can have multiple contributors. The MHRA review, along with EU and US reviews, did not support a causal association between the GLP-1 medicines reviewed and suicidal thoughts or behaviour. Anyone in distress still deserves prompt help, whatever the suspected cause.

What a “fatal report” means: the person died in a case submitted as a suspected adverse reaction. It does not, by itself, prove that the medicine caused the death. MHRA guidance explains that the underlying condition, other medicines and coincidental events can be difficult to separate. Serious reports still matter and can lead to stronger warnings when regulators find a concerning pattern across multiple data sources.

Checked against current UK guidance

Where the evidence stands now

Claim in circulation Current evidence status The context headlines often lose What matters to patients
“Wegovy or Ozempic can cause blindness” Very rare risk recognised
Regulators recognise NAION as a very rare adverse effect associated with semaglutide.		 Type 2 diabetes and vascular risk factors already raise NAION risk. Regulators considered background risk and still concluded semaglutide is associated with a very rare additional risk. This finding cannot automatically be applied to every GLP-1 medicine. Sudden sight loss or rapidly worsening vision needs urgent assessment.
“Weight-loss jabs cause fatal pancreatitis” Recognised uncommon risk
Acute pancreatitis is a known possible adverse effect. Rare necrotising and fatal suspected reports have occurred.		 Gallstones and heavy alcohol use are common causes of acute pancreatitis. Raw reports may include alternative or contributing causes and cannot prove causation in every case. Severe, persistent abdominal pain requires urgent medical help.
“Millions face permanent stomach paralysis” Association, scale uncertain
Delayed gastric emptying is part of how these medicines work. Gastroparesis is a distinct, serious clinical condition.		 Long-term diabetes can itself cause gastroparesis. Observational research found an association but included relatively few events and cannot provide a precise UK risk for current medicines and doses. Persistent vomiting, inability to tolerate food or fluids, or severe ongoing symptoms need clinical review.
“GLP-1 medicines increase suicide risk” Causal link not supported
UK, EU and later US reviews did not find evidence of increased suicidal thoughts or behaviour caused by GLP-1 receptor agonists.		 The original investigation was appropriate. Its existence was not proof that the signal would be confirmed. Mental-health symptoms are real and deserve urgent support regardless of their suspected cause.

Why the conclusion can change

A safety signal is the start of the investigation, not the end

  1. 1. A signal appearsA patient, clinician, study or monitoring system identifies something that might be linked to treatment.
  2. 2. It is investigatedRegulators examine case reports, trials, healthcare records, background rates and biological plausibility.
  3. 3. Evidence is weighedResearchers test whether the pattern is stronger than chance, bias, underlying illness or other treatments can explain.
  4. 4. Guidance changes, or does notA warning may be added, strengthened, narrowed to one drug, or rejected when the evidence does not support it.

Four phrases that are not interchangeable

Reported after: the event happened after treatment. Associated with: the event occurred more often in one group, but causation is not established. May cause: regulators judge a causal relationship to be reasonably possible. Causes: a much stronger claim requiring persuasive evidence. A headline can change the meaning simply by sliding from the first phrase to the last.

Case study 1

Semaglutide and NAION: the “blindness” headlines

Regulator-confirmed very rare risk

The early articles were ahead of the final regulatory conclusion, but the final answer is not “nothing to see here”

Examples of the coverage

Personal accounts can reveal important harms, but lawsuits and individual timelines cannot establish frequency or causation on their own.

What the evidence now supports

≈ 1 extra case

UK product information says epidemiological studies suggest roughly a two-fold relative increase, corresponding to approximately one additional NAION case per 10,000 person-years of semaglutide treatment in adults with type 2 diabetes.

The EMA classifies NAION as a very rare side effect of semaglutide medicines. In February 2026, the MHRA issued UK advice for sudden sight loss.

What changed after the first study?

A widely reported 2024 study from a specialist US neuro-ophthalmology centre found more NAION diagnoses among people prescribed semaglutide than in matched comparison groups. It was observational, came from one unusually selected clinical population and could not prove that semaglutide caused the events.

That did not make the signal meaningless. A later Danish, Norwegian cohort and other epidemiological evidence also found an increased risk. Regulators then reviewed the total evidence, not one headline or one paper, and updated the product information.

Background risk, not a dismissal: type 2 diabetes independently raises NAION risk, and smoking, high blood pressure and high cholesterol are additional risk factors. These factors can complicate observational research. The MHRA considered the available epidemiology and still concluded that semaglutide is associated with a very rare additional risk. Both facts belong in the same explanation.

Fair conclusion: “Semaglutide causes mass blindness” would be indefensible. “There is no confirmed link” is now also outdated. The accurate position is that semaglutide is associated with a very rare NAION risk, with urgent action required for sudden visual loss.

Open the regulator and clinical sources

Case study 2

Pancreatitis: when a real warning becomes a frightening raw number

Known possible adverse effect

The warning matters. The Yellow Card total still cannot tell an individual reader their personal risk.

Coverage to compare

The first headline was not fabricated. The weakness is that a count of suspected reports is emotionally powerful but cannot be read as a confirmed case count or incidence rate.

The regulator’s full context

1,296 reports

Between 2007 and October 2025, the MHRA received 1,296 Yellow Card reports of pancreatitis associated with GLP-1 or dual GIP/GLP-1 medicines. Of these, 19 were classified as fatal and 24 as necrotising pancreatitis.

These are suspected adverse-reaction reports. “Fatal” means the patient died in a reported case; it does not prove that the medicine caused the death.

What do trials show?

In the phase 3a Wegovy injection trials described in the UK product information, adjudication-confirmed acute pancreatitis was reported in 0.2% of participants receiving semaglutide 2.4 mg and less than 0.1% receiving placebo. Different medicines, doses and patient groups should not be treated as interchangeable.

The MHRA strengthened class-wide wording in January 2026 because post-marketing experience included rare necrotising and fatal cases. That was a meaningful safety update, not a newly discovered epidemic and not merely administrative housekeeping.

Alternative and contributing causes: acute pancreatitis can be life-threatening regardless of its cause. Gallstones and heavy alcohol use are the most common causes; other medical conditions, injury, surgery and some medicines can also contribute. It is therefore wrong to count every post-treatment case as drug-caused, and equally wrong to assume an underlying condition explains every case away.

Fair conclusion: pancreatitis is an uncommon but potentially serious recognised risk. Reporting the serious cases is legitimate. Presenting raw reports without explaining their limitations invites readers to mistake “reported after” for “proven to be caused by”.

How to read Yellow Card numbers: anyone can report a suspected reaction, duplicate and incomplete information can exist, exposure differs greatly between medicines, and publicity can stimulate reporting. The scheme is essential for detecting signals, but its raw totals cannot establish incidence or compare which medicine is safer.
Open the regulator, product and reporting sources

Case study 3

“Stomach paralysis”: a severe diagnosis is not the same as slower digestion

Plausible association; precise risk uncertain

The phrase is vivid. The evidence is more complicated than either “millions face it” or “it is just normal nausea”.

Examples of the coverage

A severe patient experience deserves to be heard. It does not show that the same outcome is common, permanent or inevitable for a wider population.

What we can say responsibly

4 + 66 cases

A 2023 US claims-database study recorded four gastroparesis events among 613 semaglutide users and 66 among 4,144 liraglutide users. It found an adjusted association versus bupropion, naltrexone, but the event counts were small and the confidence interval was wide.

This was not a trial of today’s complete UK weight-management market and it did not establish that every reported case was permanent.

Slowed emptying versus gastroparesis

GLP-1 medicines can delay gastric emptying; that action contributes to fullness and may also cause nausea, vomiting or bloating. Gastroparesis is a clinical disorder involving persistently delayed stomach emptying without a mechanical blockage. Ordinary fullness after a dose is not, by itself, a gastroparesis diagnosis.

Current UK Wegovy professional information says semaglutide should be used with caution in patients who already have gastroparesis and is not recommended if it is severe. It also recognises delayed gastric emptying and post-marketing intestinal obstruction. That is a reason for careful clinical assessment, not a reason to diagnose yourself from a headline.

Why the patient’s medical history matters: the NHS identifies long-term diabetes as a cause of gastroparesis; some surgery can also cause it, and sometimes the cause is unknown. A clinician has to consider these possibilities alongside the medicine, the timing and objective testing.

Fair conclusion: persistent or severe digestive symptoms should not be brushed off. Equally, a dramatic anecdote cannot tell millions of readers that they face a permanent condition. The evidence supports vigilance, not a population-wide prediction.

Open the clinical and product sources

Case study 4

Suicidal thoughts: an appropriate investigation that did not confirm the feared link

Causal association not supported

This case shows why “under review” must not be rewritten as “has been shown to cause”.

Contemporary reporting

The BBC and Guardian headlines accurately said “investigated” and “under review”. Problems arise when that provisional status is dropped in retelling.

What later reviews found

107,910 people

A comprehensive FDA meta-analysis across 91 placebo-controlled trials included 107,910 participants and found no increased risk of suicidal thoughts or behaviour, or relevant psychiatric events such as depression, anxiety or psychosis.

The MHRA had already concluded in September 2024 that available evidence did not support a causal association for the GLP-1 receptor agonists it reviewed.

Fair conclusion: the initial reports deserved investigation. The later evidence did not confirm that GLP-1 receptor agonists cause suicidal thoughts or behaviour. That regulatory conclusion does not make somebody’s distress less real or remove the need for urgent support.

Cause and care are different questions: symptoms can have multiple contributors, and an event occurring after a dose does not establish why it happened. The regulator’s conclusion is about population-level causation. The person’s need for support is immediate and does not depend on proving a cause first.

Open the regulator and support sources

A practical five-minute check

How to read the next frightening GLP-1 headline

Question 1

Which medicine?

Semaglutide evidence cannot automatically be applied to tirzepatide, liraglutide or an entire drug class. Look for the active ingredient, not only “fat jab” or “GLP-1”.

Question 2

What kind of evidence?

A patient account, spontaneous report, claims-database study, randomised trial and regulator review answer different questions. None should be presented as though it were all of the others.

Question 3

Relative or absolute risk?

“Risk doubled” can mean 1 became 2 in 10,000, or 1 became 2 in 10. Ask for the starting risk, time period and comparison group.

Question 4

Report or confirmed case?

A Yellow Card is a suspected association that helps regulators spot patterns. Do not divide raw reports by prescriptions and call the result a side-effect rate.

Question 5

What did regulators conclude?

Go past “experts warn” to the current MHRA notice, EMA decision and UK product information. Check the date: an honest 2023 article can still be outdated in 2026.

Question 6

What should a patient do?

Useful reporting identifies symptoms and the correct level of action. Fear without practical guidance is incomplete health communication.

The part both extremes miss

Safety is a benefit, risk decision, not a vote on whether a medicine is “good” or “bad”

Obesity and type 2 diabetes carry serious health risks. GLP-1 and dual GIP/GLP-1 medicines can provide substantial benefits for eligible people, while also causing common gastrointestinal effects and carrying less common but important risks. Approval does not mean risk-free. A warning does not mean the medicine is unsuitable for everyone.

Your own balance depends on the medicine, indication, dose, medical history, other treatments and symptoms. That decision belongs with an appropriate prescriber who knows your circumstances, not with a newspaper headline, a comparison website or a social-media thread.

For a broader symptom guide, read Mounjaro side effects: what to expect and when to get help.

Evidence library

Direct links used for this review

We prioritised current UK regulator notices and UK product information, then used peer-reviewed studies to explain how the evidence developed. News articles are included as examples of framing, not as clinical authorities.

Primary UK safety source

MHRA: what you need to know about GLP-1 medicines

Patient-level UK overview covering common effects, pancreatitis, NAION, pregnancy, contraception, mental health and surgery.

Current product information

Wegovy injection SmPC

Professional UK prescribing information, including adverse-effect frequencies, warnings and study data.

Suspected-reaction reporting

MHRA Yellow Card scheme

Report a suspected side effect or falsified medicine. A Yellow Card is a suspicion, not proof that the product caused the event.

Background causes and symptoms

NHS: acute pancreatitis

Independent NHS guidance on warning symptoms, seriousness and common causes including gallstones and heavy alcohol use.

Background causes and diagnosis

NHS: gastroparesis

Symptoms, diagnostic testing and established causes, including long-term diabetes and some surgery.

Monj transparency

Independence and funding statement

How Monj is funded, how provider rankings are kept independent and who operates the website.

All clinical, regulator and news links in one list

Regulators and product information

Peer-reviewed research

News and headline examples

Common questions

GLP-1 safety headline FAQs

Are frightening GLP-1 headlines simply made up?

Usually not. Most start with a real patient report, study, regulator investigation or product warning. The distortion often occurs when uncertainty, absolute risk, study limitations or the specific medicine involved disappears from the headline.

Can Wegovy or Ozempic cause blindness?

Semaglutide, the active ingredient in Wegovy and Ozempic, is associated with a very rare eye condition called NAION. UK product information describes approximately one additional case per 10,000 person-years of treatment in adults with type 2 diabetes. Sudden or rapidly worsening vision requires urgent medical assessment.

Does the same NAION warning apply to Mounjaro?

The current confirmed regulatory conclusion is specific to semaglutide. Mounjaro contains tirzepatide, a different dual GIP/GLP-1 medicine. The MHRA has said it is reviewing evidence concerning other GLP-1 medicines, so it would be wrong to present the semaglutide estimate as a proven Mounjaro risk.

Can Mounjaro, Wegovy or other GLP-1 medicines cause pancreatitis?

Acute pancreatitis is a recognised possible adverse effect of GLP-1 and dual GIP/GLP-1 medicines. It is uncommon overall, but rare severe, necrotising and fatal suspected reports have occurred. Severe, persistent abdominal pain, particularly pain that travels to the back, needs urgent medical help.

Does a fatal Yellow Card report prove a GLP-1 medicine caused the death?

No. It means the patient died in a case reported as a suspected adverse reaction. Medical history, the underlying condition, other medicines and alternative causes may all be relevant. The report still matters: regulators look for patterns across reports, trials, healthcare data and other evidence, and serious reports can contribute to stronger warnings.

Do underlying health conditions mean the medicine was not responsible?

No. An existing condition can raise the background risk or contribute to an event, but it does not automatically exclude a medicine as another cause or contributor. Researchers and regulators compare treated and untreated groups, examine alternative explanations and look across several kinds of evidence. The fairest answer may remain “uncertain” for one person even when a population-level warning is justified.

Is slowed digestion the same thing as stomach paralysis?

No. Delayed gastric emptying is one effect of GLP-1 treatment. Gastroparesis is a clinical disorder involving persistently delayed stomach emptying without a blockage. Feeling fuller or temporarily nauseous does not automatically mean somebody has gastroparesis.

Do GLP-1 medicines cause suicidal thoughts?

Current MHRA, EMA and FDA reviews do not support a causal association between the GLP-1 receptor agonists reviewed and suicidal thoughts or behaviour. Anyone experiencing suicidal thoughts still needs prompt support, regardless of what may have caused them.

Should I stop treatment because of a news article?

Do not change or stop prescribed treatment solely because of a headline. Read the current UK patient leaflet and contact your prescriber or pharmacist. If you have an urgent warning symptom such as sudden sight loss or severe persistent abdominal pain, seek urgent medical help rather than waiting for a routine appointment.

Who made this page

Editorial ownership

Written and researched by Nick Johnson, Monj founder. Monj is operated by Medstack Ltd and is an independent UK comparison and patient-information platform. We do not sell, prescribe or dispense medicines.

How it is maintained

Review standard

Claims were checked against MHRA and EMA safety decisions, current UK product information and the underlying peer-reviewed studies. We recheck this page after material regulator updates and during scheduled editorial review.

Last evidence review: 20 August 2026. Read our independence statement.

Found an error or a newer regulator update? Please contact Monj so the evidence can be checked and corrected. This page was last evidence-checked on 20 August 2026.

Back to the top

Sources referenced on this page

Linked from the text above:

Standing primary references for every Monj page:

Share this page

Share on Reddit Share on WhatsApp Share on Bluesky

Monj is an independent price comparison index operated by Medstack Ltd. We are not a pharmacy and we do not sell, prescribe, supply or dispense medicines. Listings are ranked strictly by the lowest total patient price, with alphabetical order used to break ties. No pharmacy can pay for placement, ranking or badges. Some outbound links earn a referral commission, and this never affects position or inclusion. Read the full independence statement.