Study analysis · BMC Medicine · 2025
The largest look yet at semaglutide: 15% lower relative risk of death, 23% lower relative risk of heart attack — and a 3-fold relative increase in nausea.
A big review found that semaglutide helps prevent death and heart attacks in people with heart risks, but it often causes stomach upset like nausea and vomiting.
Overview
What the study found
The study in plain English — the bottom line, every takeaway we extracted, and what to do with them.
In simple terms
This study put together results from 50 randomized trials, where people were randomly assigned to semaglutide or a placebo. Because of that random assignment, we can be quite confident that semaglutide itself caused the good effects (like fewer deaths and heart attacks) and the bad effects (like nausea and vomiting) in people similar to those studied. It doesn't prove semaglutide works for everyone or is safe for all uses.
What’s the bottom line?
Scientists combined 50 randomized trials to see how semaglutide affects people at higher risk of heart problems.
How strong is this study?
The researchers did a careful job: they searched many databases, followed a pre-written plan, and used strong statistical methods to check the results. However, more than half the trials had some risk of bias, and most were funded by the company that makes semaglutide, so we can trust the main findings but should be cautious about smaller or less certain results.
40 / 100
- COI disclosure+40/40
- Data availabilitydata not shared
- Code availabilitycode not shared
38 / 100
- Randomizationnot randomized
- Blindingblinding unclear
- Control groupno control group
- Sample size (n=54972)+20/20
- Follow-up+10/10
100 / 100
100 / 100
- P-values+15/15
- Effect size+20/20
- Confidence intervals+15/15
- Pre-registration+15/15
Each component is scored out of 100 and then capped by the study design — a case series cannot reach the ceiling a randomised trial can, however well it is reported.
Where it sits
RCT reviewsReviews of RCTs (Meta-analyses)
Max 100Randomized TrialsRandomized Trials
Max 90Reviews of Cohort StudiesReviews of Cohort Studies
Max 85Cohort StudiesCohort Studies
Max 72Reviews of Case-Control StudiesReviews of Case-Control Studies
Max 63Case-Control StudiesCase-Control Studies
Max 58Cross-Sectional & Case SeriesCross-Sectional & Case Series
Max 50Expert OpinionExpert Opinion
Max 566 / 100
Probability of being correct
The highest quality evidence. Systematic reviews and meta-analyses that pool randomized controlled trials, giving the most reliable summary of experimental evidence.
This design can establish causation. The review pools randomized placebo-controlled trials, so the design can support causal inference. However, 54% of included trials were at high risk of bias, 88% were funded by Novo Nordisk, some outcomes had heterogeneity or potential publication bias, and multiple secondary/exploratory comparisons were not adjusted for multiple testing. These limitations reduce certainty for specific adverse events but do not negate the causal design for the primary outcomes with high GRADE certainty.
No Conflicts
No conflicts of interest identified
No conflicts of interest or funding source were declared for the systematic review authors. All included trials appear to be industry-funded, but this does not constitute a direct conflict for the review itself.
Independent Analysis Safeguards
- PROSPERO registration (CRD42024499511)
- PRISMA reporting guidelines
- Trial Sequential Analysis
- GRADE assessment
- Independent data extraction and risk of bias assessment by multiple authors
- Pre-specified eight-step assessment for statistical significance
The review itself has no declared funding or competing interests. However, the included trials are predominantly funded by pharmaceutical companies (mostly Novo Nordisk), which may introduce bias in the primary evidence base. The review authors used rigorous methodological safeguards, but no explicit safeguards against industry influence on the review process are described.
Key takeaways
- 01
Semaglutide lowered relative risk of death by 15% (RR 0.85), heart attack by 23% (RR 0.77), and serious adverse events by 7% (RR 0.93).
- 02
It raised relative risk of nausea 3.00-fold, vomiting 4.12-fold, and diarrhea 1.88-fold.
- 03
Stroke and hospitalization were not clearly changed.
- 04
In absolute terms, over the study follow-up, death happened in 5.9% of placebo patients vs 4.1% of semaglutide patients — about 18 fewer deaths per 1,000 people.
- 05
Heart attacks: 3.3% vs 2.1% — about 12 fewer per 1,000.
- 06
Serious adverse events: 33.1% vs 26.5% — about 66 fewer per 1,000.
- 07
The GI side effects were common: numbers needed to harm were 6 for nausea, 11 for vomiting, and 11 for diarrhea (roughly 167, 91, and 91 extra cases per 1,000 people).
Surprising findings
- Semaglutide reduced serious adverse events by 7% relative, even though it increased many non-serious GI side effects.Common belief is that a drug with frequent nausea, vomiting, and diarrhea would overall make people feel worse and have more health problems. Instead, serious events were fewer.
- No significant reduction in stroke, despite clear reductions in death and heart attack.If a drug improves cardiovascular health, many assume it would also prevent strokes. This study found RR 0.93 with a confidence interval crossing 1.0.
- Nausea number needed to harm is only 6.That means for every 6 people taking semaglutide, one extra person experiences nausea compared with placebo. It is far more common than many patients expect.
- 88% of included trials were funded by Novo Nordisk, but the authors did not downgrade the evidence for for-profit bias.Industry funding is a classic red flag for bias. The authors argued trial quality was high enough to avoid downgrading, which is a debatable call.
- The rare eye side effect ischemic optic neuropathy was not significantly increased.Recent observational studies raised alarm about semaglutide and NAION. This meta-analysis found only 5 events vs 1 event, with a very wide confidence interval.
Practical takeaways
If you are considering semaglutide for cardiovascular risk, ask your doctor for both the relative and absolute benefits: about 15% relative lower risk of death and 23% relative lower risk of heart attack, but only 18 fewer deaths and 12 fewer heart attacks per 1,000 people.
Absolute benefits depend on your baseline risk and follow-up duration. The study included people at increased cardiovascular risk, not everyone.
high confidenceExpect GI side effects. Nausea is especially common — 1 extra case for every 6 people treated. Ask about dose titration, dietary changes, and anti-nausea options.
These are non-serious adverse events; most are not dangerous, but they can affect quality of life and adherence.
moderate confidenceDo not assume semaglutide prevents strokes. The study found no statistically significant reduction in stroke.
The confidence interval was wide and the result was moderate certainty; it does not rule out a small benefit.
moderate confidenceBe skeptical of headlines claiming semaglutide causes blindness. This meta-analysis found no significant increase in ischemic optic neuropathy.
The event was very rare, with only 5 vs 1 events, so the evidence is very low certainty and cannot exclude a rare risk.
low confidenceWhen reading about semaglutide, check who funded the trials. 88% were supported by Novo Nordisk, which does not invalidate the results but should be considered.
The authors did not downgrade evidence for for-profit bias because of high methodological quality, but independent replication is still valuable.
high confidenceWhy this study matters
Mortality benefit: 15% relative reduction, 18 fewer deaths per 1,000
In 50 randomized trials with 54,972 participants, semaglutide reduced all-cause mortality with RR 0.85 (95% CI 0.79–0.91; high certainty; I²=0.0%). Absolute risk was 5.9% on placebo vs 4.1% on semaglutide — about 18 fewer deaths per 1,000 people over follow-up of 15.5 weeks to 47.5 months.
This is a hard outcome, not just weight loss. It suggests the drug may keep people alive, but the absolute benefit is modest.
Heart attack risk down 23% relative
Myocardial infarction was reduced with RR 0.77 (95% CI 0.69–0.85; high certainty; I²=0.0%). Absolute risk fell from 3.3% to 2.1%, about 12 fewer heart attacks per 1,000 people.
Most people think of semaglutide as a weight-loss drug, but its cardiovascular benefit may be a major selling point.
Serious adverse events down 7% relative, 66 fewer per 1,000
Serious adverse events were reduced with RR 0.93 (95% CI 0.88–0.98; high certainty; I²=24.1%). Absolute risk was 33.1% on placebo vs 26.5% on semaglutide — about 66 fewer serious adverse events per 1,000.
It seems counterintuitive that a drug with many side effects could reduce serious health events overall.
GI side effects are common: nausea NNH 6, vomiting/diarrhea NNH 11
Nausea increased with RR 3.00 (95% CI 2.63–3.42; NNH 6), vomiting with RR 4.12 (95% CI 3.47–4.90; NNH 11), and diarrhea with RR 1.88 (95% CI 1.68–2.11; NNH 11). That translates to roughly 167 extra nausea cases, 91 extra vomiting cases, and 91 extra diarrhea cases per 1,000 people.
These are not rare side effects. About 1 in 6 people get nausea, which can affect daily life and adherence.
Stroke and hospitalization: no clear benefit
Stroke was not significantly reduced: RR 0.93 (95% CI 0.82–1.06; p=0.27), absolute 1.6% vs 2.0%, about 4 fewer per 1,000. All-cause hospitalization also showed no significant difference: RR 0.93 (95% CI 0.79–1.10; p=0.42; low certainty).
People may assume all cardiovascular events fall equally, but heart attack and death improved while stroke did not.
Rare eye side effect not confirmed
Ischemic optic neuropathy occurred in 5/10,451 semaglutide users vs 1/10,450 placebo users, RR 2.85 (95% CI 0.43–18.70; p=0.28; very low certainty). The study does not support a significant risk, but the confidence interval is wide.
Recent headlines suggested a possible link between semaglutide and blindness. This meta-analysis found no statistically significant increase.
Funding and bias red flags
88% of trials (44/50) were supported by Novo Nordisk, the maker of semaglutide. 54% of trials (27/50) were at high risk of bias, though 84.4% of participants came from low-risk-of-bias trials. Authors did not downgrade evidence for for-profit bias, and Egger’s test suggested possible publication bias for non-serious adverse events (p=0.002).
Trust in evidence matters. Even a high-quality meta-analysis can be influenced by who funds the underlying trials.
Want the whole report?
Detailed mode opens the full scientific breakdown — every score component, the methodology, conflicts of interest, the evidence analysis behind each claim, and the raw study data.
Overview
What the study found
The study in plain English — the bottom line, every takeaway we extracted, and what to do with them.
Not medical advice. For informational purposes only. Always consult a healthcare professional. Terms
Scientists combined 50 randomized trials to see how semaglutide affects people at higher risk of heart problems.
Research results
Semaglutide lowered relative risk of death by 15% (RR 0.85), heart attack by 23% (RR 0.77), and serious adverse events by 7% (RR 0.93). It raised relative risk of nausea 3.00-fold, vomiting 4.12-fold, and diarrhea 1.88-fold. Stroke and hospitalization were not clearly changed.
What this means - more context
In absolute terms, over the study follow-up, death happened in 5.9% of placebo patients vs 4.1% of semaglutide patients — about 18 fewer deaths per 1,000 people. Heart attacks: 3.3% vs 2.1% — about 12 fewer per 1,000. Serious adverse events: 33.1% vs 26.5% — about 66 fewer per 1,000. The GI side effects were common: numbers needed to harm were 6 for nausea, 11 for vomiting, and 11 for diarrhea (roughly 167, 91, and 91 extra cases per 1,000 people).
To systematically assess adverse effects and cardiovascular outcomes of semaglutide (oral or subcutaneous) versus placebo in patients at increased risk of cardiovascular events.
In this systematic review/meta-analysis of 50 randomized trials (54,972 participants), semaglutide reduced relative risk of all-cause mortality by 15% (RR 0.85; absolute risk 5.9% placebo vs 4.1% semaglutide, about 18 fewer deaths per 1,000), myocardial infarction by 23% (RR 0.77; absolute 3.3% vs 2.1%, about 12 fewer per 1,000), and serious adverse events by 7% (RR 0.93; absolute 33.1% vs 26.5%, about 66 fewer per 1,000). It increased relative risk of non-serious GI adverse events: nausea 3.00-fold (NNH 6), vomiting 4.12-fold (NNH 11), diarrhea 1.88-fold (NNH 11). No significant differences in stroke or all-cause hospitalization; ischemic optic neuropathy was not significantly increased.
Methods Used
Systematic review with meta-analysis and Trial Sequential Analysis. Searched six databases and trial registries to 31 March 2025; included randomized trials comparing semaglutide (oral or subcutaneous) with placebo in patients at increased cardiovascular risk. 50 trials, 54,972 participants; risk of bias via RoB 2, GRADE certainty, eight-step significance thresholds, PROSPERO CRD42024499511.
Main Finding
In patients at increased cardiovascular risk, semaglutide reduced relative risk of all-cause mortality 15% (RR 0.85, 95% CI 0.79–0.91; absolute 5.9% to 4.1%, 18 fewer deaths per 1,000), myocardial infarction 23% (RR 0.77, 95% CI 0.69–0.85; absolute 3.3% to 2.1%, 12 fewer per 1,000), and serious adverse events 7% (RR 0.93, 95% CI 0.88–0.98; absolute 33.1% to 26.5%, 66 fewer per 1,000) versus placebo over follow-up of 15.5 weeks to 47.5 months. It increased non-serious GI adverse events: nausea RR 3.00, vomiting RR 4.12, diarrhea RR 1.88. Stroke and hospitalization were not significantly reduced.
Confidence Level
High certainty of evidence for all-cause mortality and myocardial infarction; high certainty for serious adverse events (moderate heterogeneity I²=24.1%); moderate certainty for GI adverse events; low certainty for hospitalization; very low certainty for ischemic optic neuropathy. Limitations: 54% of trials at high risk of bias (though 84.4% of participants from low-risk trials), 88% of trials funded by Novo Nordisk, and short/variable follow-up.
Study Flags
Red Flags
- •54% of included trials (27/50) were at high risk of bias, although 84.4% of participants came from low-risk-of-bias trials.
- •88% of trials (44/50) were supported by Novo Nordisk, the maker of semaglutide; authors did not downgrade evidence for for-profit bias.
- •Follow-up varied widely (15.5 weeks to 47.5 months), and thresholds were not adjusted for all secondary/exploratory comparisons; possible publication bias for non-serious adverse events (Egger p=0.002).
Surprising Findings
Semaglutide reduced serious adverse events by 7% relative, even though it increased many non-serious GI side effects.
Common belief is that a drug with frequent nausea, vomiting, and diarrhea would overall make people feel worse and have more health problems. Instead, serious events were fewer.
Practical Takeaways
If you are considering semaglutide for cardiovascular risk, ask your doctor for both the relative and absolute benefits: about 15% relative lower risk of death and 23% relative lower risk of heart attack, but only 18 fewer deaths and 12 fewer heart attacks per 1,000 people.
RCT reviewsReviews of RCTs (Meta-analyses)
Max 100Randomized TrialsRandomized Trials
Max 90Reviews of Cohort StudiesReviews of Cohort Studies
Max 85Cohort StudiesCohort Studies
Max 72Reviews of Case-Control StudiesReviews of Case-Control Studies
Max 63Case-Control StudiesCase-Control Studies
Max 58Cross-Sectional & Case SeriesCross-Sectional & Case Series
Max 50Expert OpinionExpert Opinion
Max 566 / 100
Probability of being correct
The highest quality evidence. Systematic reviews and meta-analyses that pool randomized controlled trials, giving the most reliable summary of experimental evidence.
Human Meta-Analysis
Subject
Moderate probability
on the GRADE evidence scale
This study put together results from 50 randomized trials, where people were randomly assigned to semaglutide or a placebo. Because of that random assignment, we can be quite confident that semaglutide itself caused the good effects (like fewer deaths and heart attacks) and the bad effects (like nausea and vomiting) in people similar to those studied. It doesn't prove semaglutide works for everyone or is safe for all uses.
The study has a COI section but no disclosure was found. A small penalty has been applied.
Strengths
- Comprehensive search of six databases and trial registries.
- Pre-registered protocol (PROSPERO) and PRISMA reporting.
- Large sample size: 50 trials, 54,972 participants.
Weaknesses
- 54% of included trials were at high risk of bias.
- 88% of trials were supported by Novo Nordisk, the manufacturer.
- Multiple comparisons for secondary/exploratory outcomes were not adjusted.
Methodology
Evidence Keywords
Statistical Reporting
Not medical advice. For informational purposes only. Always consult a healthcare professional. Terms
Scientists combined 50 randomized trials to see how semaglutide affects people at higher risk of heart problems.
Research results
Semaglutide lowered relative risk of death by 15% (RR 0.85), heart attack by 23% (RR 0.77), and serious adverse events by 7% (RR 0.93). It raised relative risk of nausea 3.00-fold, vomiting 4.12-fold, and diarrhea 1.88-fold. Stroke and hospitalization were not clearly changed.
What this means - more context
In absolute terms, over the study follow-up, death happened in 5.9% of placebo patients vs 4.1% of semaglutide patients — about 18 fewer deaths per 1,000 people. Heart attacks: 3.3% vs 2.1% — about 12 fewer per 1,000. Serious adverse events: 33.1% vs 26.5% — about 66 fewer per 1,000. The GI side effects were common: numbers needed to harm were 6 for nausea, 11 for vomiting, and 11 for diarrhea (roughly 167, 91, and 91 extra cases per 1,000 people).
To systematically assess adverse effects and cardiovascular outcomes of semaglutide (oral or subcutaneous) versus placebo in patients at increased risk of cardiovascular events.
In this systematic review/meta-analysis of 50 randomized trials (54,972 participants), semaglutide reduced relative risk of all-cause mortality by 15% (RR 0.85; absolute risk 5.9% placebo vs 4.1% semaglutide, about 18 fewer deaths per 1,000), myocardial infarction by 23% (RR 0.77; absolute 3.3% vs 2.1%, about 12 fewer per 1,000), and serious adverse events by 7% (RR 0.93; absolute 33.1% vs 26.5%, about 66 fewer per 1,000). It increased relative risk of non-serious GI adverse events: nausea 3.00-fold (NNH 6), vomiting 4.12-fold (NNH 11), diarrhea 1.88-fold (NNH 11). No significant differences in stroke or all-cause hospitalization; ischemic optic neuropathy was not significantly increased.
Methods Used
Systematic review with meta-analysis and Trial Sequential Analysis. Searched six databases and trial registries to 31 March 2025; included randomized trials comparing semaglutide (oral or subcutaneous) with placebo in patients at increased cardiovascular risk. 50 trials, 54,972 participants; risk of bias via RoB 2, GRADE certainty, eight-step significance thresholds, PROSPERO CRD42024499511.
Main Finding
In patients at increased cardiovascular risk, semaglutide reduced relative risk of all-cause mortality 15% (RR 0.85, 95% CI 0.79–0.91; absolute 5.9% to 4.1%, 18 fewer deaths per 1,000), myocardial infarction 23% (RR 0.77, 95% CI 0.69–0.85; absolute 3.3% to 2.1%, 12 fewer per 1,000), and serious adverse events 7% (RR 0.93, 95% CI 0.88–0.98; absolute 33.1% to 26.5%, 66 fewer per 1,000) versus placebo over follow-up of 15.5 weeks to 47.5 months. It increased non-serious GI adverse events: nausea RR 3.00, vomiting RR 4.12, diarrhea RR 1.88. Stroke and hospitalization were not significantly reduced.
Confidence Level
High certainty of evidence for all-cause mortality and myocardial infarction; high certainty for serious adverse events (moderate heterogeneity I²=24.1%); moderate certainty for GI adverse events; low certainty for hospitalization; very low certainty for ischemic optic neuropathy. Limitations: 54% of trials at high risk of bias (though 84.4% of participants from low-risk trials), 88% of trials funded by Novo Nordisk, and short/variable follow-up.
Study Flags
Red Flags
- •54% of included trials (27/50) were at high risk of bias, although 84.4% of participants came from low-risk-of-bias trials.
- •88% of trials (44/50) were supported by Novo Nordisk, the maker of semaglutide; authors did not downgrade evidence for for-profit bias.
- •Follow-up varied widely (15.5 weeks to 47.5 months), and thresholds were not adjusted for all secondary/exploratory comparisons; possible publication bias for non-serious adverse events (Egger p=0.002).
Surprising Findings
Semaglutide reduced serious adverse events by 7% relative, even though it increased many non-serious GI side effects.
Common belief is that a drug with frequent nausea, vomiting, and diarrhea would overall make people feel worse and have more health problems. Instead, serious events were fewer.
Practical Takeaways
If you are considering semaglutide for cardiovascular risk, ask your doctor for both the relative and absolute benefits: about 15% relative lower risk of death and 23% relative lower risk of heart attack, but only 18 fewer deaths and 12 fewer heart attacks per 1,000 people.
RCT reviewsReviews of RCTs (Meta-analyses)
Max 100Randomized TrialsRandomized Trials
Max 90Reviews of Cohort StudiesReviews of Cohort Studies
Max 85Cohort StudiesCohort Studies
Max 72Reviews of Case-Control StudiesReviews of Case-Control Studies
Max 63Case-Control StudiesCase-Control Studies
Max 58Cross-Sectional & Case SeriesCross-Sectional & Case Series
Max 50Expert OpinionExpert Opinion
Max 566 / 100
Probability of being correct
The highest quality evidence. Systematic reviews and meta-analyses that pool randomized controlled trials, giving the most reliable summary of experimental evidence.
Human Meta-Analysis
Subject
Moderate probability
on the GRADE evidence scale
This study put together results from 50 randomized trials, where people were randomly assigned to semaglutide or a placebo. Because of that random assignment, we can be quite confident that semaglutide itself caused the good effects (like fewer deaths and heart attacks) and the bad effects (like nausea and vomiting) in people similar to those studied. It doesn't prove semaglutide works for everyone or is safe for all uses.
The study has a COI section but no disclosure was found. A small penalty has been applied.
Strengths
- Comprehensive search of six databases and trial registries.
- Pre-registered protocol (PROSPERO) and PRISMA reporting.
- Large sample size: 50 trials, 54,972 participants.
Weaknesses
- 54% of included trials were at high risk of bias.
- 88% of trials were supported by Novo Nordisk, the manufacturer.
- Multiple comparisons for secondary/exploratory outcomes were not adjusted.
Methodology
Evidence Keywords
Statistical Reporting
Scoring
How strong is this study?
The researchers did a careful job: they searched many databases, followed a pre-written plan, and used strong statistical methods to check the results. However, more than half the trials had some risk of bias, and most were funded by the company that makes semaglutide, so we can trust the main findings but should be cautious about smaller or less certain results.
40 / 100
- COI disclosure+40/40
- Data availabilitydata not shared
- Code availabilitycode not shared
38 / 100
- Randomizationnot randomized
- Blindingblinding unclear
- Control groupno control group
- Sample size (n=54972)+20/20
- Follow-up+10/10
100 / 100
100 / 100
- P-values+15/15
- Effect size+20/20
- Confidence intervals+15/15
- Pre-registration+15/15
Each component is scored out of 100 and then capped by the study design — a case series cannot reach the ceiling a randomised trial can, however well it is reported.
Where it sits
RCT reviewsReviews of RCTs (Meta-analyses)
Max 100Randomized TrialsRandomized Trials
Max 90Reviews of Cohort StudiesReviews of Cohort Studies
Max 85Cohort StudiesCohort Studies
Max 72Reviews of Case-Control StudiesReviews of Case-Control Studies
Max 63Case-Control StudiesCase-Control Studies
Max 58Cross-Sectional & Case SeriesCross-Sectional & Case Series
Max 50Expert OpinionExpert Opinion
Max 566 / 100
Probability of being correct
The highest quality evidence. Systematic reviews and meta-analyses that pool randomized controlled trials, giving the most reliable summary of experimental evidence.
This design can establish causation. The review pools randomized placebo-controlled trials, so the design can support causal inference. However, 54% of included trials were at high risk of bias, 88% were funded by Novo Nordisk, some outcomes had heterogeneity or potential publication bias, and multiple secondary/exploratory comparisons were not adjusted for multiple testing. These limitations reduce certainty for specific adverse events but do not negate the causal design for the primary outcomes with high GRADE certainty.
No Conflicts
No conflicts of interest identified
No conflicts of interest or funding source were declared for the systematic review authors. All included trials appear to be industry-funded, but this does not constitute a direct conflict for the review itself.
Independent Analysis Safeguards
- PROSPERO registration (CRD42024499511)
- PRISMA reporting guidelines
- Trial Sequential Analysis
- GRADE assessment
- Independent data extraction and risk of bias assessment by multiple authors
- Pre-specified eight-step assessment for statistical significance
The review itself has no declared funding or competing interests. However, the included trials are predominantly funded by pharmaceutical companies (mostly Novo Nordisk), which may introduce bias in the primary evidence base. The review authors used rigorous methodological safeguards, but no explicit safeguards against industry influence on the review process are described.
Standing
Who’s using this study?
The videos and claims on this site that lean on this study, and the researchers who wrote it.
1 video from Dr Brad Stanfield cite this study, drawing 1 claim from it.
- Very strong evidence
Randomized or controlled trials support this claim, alongside consistent supporting evidence.
Evidence
Authored by
15 researchersIf this is your work, this is how we attribute it on Fit Body Science. Christina Dam Bjerregaard Sillassen is listed as the lead author.