Which GLP-1 causes the most digestive side effects?

If you are choosing a GLP-1 drug, or wondering whether a different one would be easier on your stomach, you want a straight answer: which one causes the fewest side effects? A large 2025 analysis finally ranked them. The short version is that the differences are real but smaller than most people expect, and no GLP-1 drug is gentle.

TL;DR

Researchers pooled 39 trials covering 33,354 non-diabetic adults and ranked six GLP-1 and related drugs by digestive side effects. For nausea, the injectables are almost tied, all raising the risk two- to three-fold versus placebo. Tirzepatide stood out for more vomiting, diarrhea and constipation, while semaglutide was the only drug linked to more acid reflux. The pill, orforglipron, had the highest nausea estimate but from thinner data. The most useful finding is not which drug wins, it is that side effects are dose-dependent and worst during the dose-climb, then ease off.

What the study actually measured

The analysis, published in the International Journal of Obesity in 2025, is a systematic review and network meta-analysis. That means the authors combined 39 separate trials and used statistics to compare drugs that were rarely tested head-to-head, ranking each one against placebo. The population matters: everyone in these trials was an adult with overweight or obesity and without diabetes, which is exactly the group most people picture when they think about GLP-1 drugs for weight loss.

Every number below is a relative risk versus placebo. A relative risk of 3.0 means people on that drug reported the symptom about three times as often as people on a dummy injection. It is a comparison of how much more likely a side effect is, not the percentage of users who get it. Keep that distinction in mind, because a big-looking number does not mean most people will be affected.

The nausea ranking

Nausea is the side effect people fear most and the one most likely to make someone quit. Here is how the six drugs ranked, from the highest nausea signal to the lowest:

• Orforglipron: relative risk 4.77
• Liraglutide: relative risk 3.09
• Semaglutide: relative risk 2.95
• Tirzepatide: relative risk 2.90
• Exenatide: relative risk 2.66
• Cagrilintide: relative risk 2.30

Look closely and the story is not "one drug is much worse." The four injectables in the middle - liraglutide, semaglutide, tirzepatide and exenatide - are packed into a narrow band between 2.66 and 3.09. Those ranges overlap once you account for statistical uncertainty, so treating semaglutide as meaningfully worse than tirzepatide for nausea is reading too much into a rounding difference. Cagrilintide, which is an amylin drug rather than a classic GLP-1, had the gentlest nausea signal.

Orforglipron, the once-daily pill, sits at the top with a relative risk of 4.77. That looks alarming, but its confidence interval was very wide (2.02 to 11.31), which is the statistical way of saying the estimate rests on fewer trials and fewer events. It may settle lower as more data comes in. If you want the fuller picture on this drug, our explainer on orforglipron and the new oral GLP-1 pill covers where it fits.

Vomiting, diarrhea and constipation: where tirzepatide stands out

Nausea is only the headline. When you look at the other gut symptoms, one drug separates from the pack: tirzepatide, the dual GIP/GLP-1 agonist sold as Mounjaro and Zepbound.

Drug	Nausea	Vomiting	Diarrhea	Constipation
Tirzepatide (Mounjaro, Zepbound)	2.90	13.23	3.35	3.36
Semaglutide (Ozempic, Wegovy)	2.95	4.21	1.77	2.10
Liraglutide (Saxenda)	3.09	3.87	1.82	2.24
Orforglipron (pill)	4.77	4.43	NS	NS
Exenatide	2.66	NS	NS	NS
Cagrilintide (amylin)	2.30	NS	NS	NS

Numbers are relative risk versus placebo. "NS" means the result was not statistically significant, so the drug did not clearly raise that symptom in this analysis. Tirzepatide had the highest signal for vomiting (relative risk 13.23), diarrhea (3.35) and constipation (3.36). The vomiting figure sounds dramatic, but its confidence interval was enormous (4.85 to 36.09), meaning it comes from a small number of events and should be read as "clearly elevated" rather than a precise 13-fold number.

Semaglutide and liraglutide were more moderate on diarrhea and constipation, each raising them by roughly two-fold or less. This lines up with what clinics see and with the SURMOUNT-5 comparison, where tirzepatide delivered more weight loss than semaglutide but also more people reporting gut symptoms. Our write-up of the SURMOUNT-5 head-to-head trial covers the trade-off between results and tolerability.

The reflux outlier: semaglutide

One symptom flipped the ranking. Semaglutide was the only drug in the analysis clearly linked to more gastroesophageal reflux, the burning, backing-up sensation people call heartburn, with a relative risk of 2.43. Tirzepatide, the villain on vomiting and diarrhea, did not show a significant reflux signal.

This is a useful reminder that "most side effects" depends entirely on which side effect you are prone to. If reflux is your weak spot, the drug with the worst overall gut profile might actually suit you better than the one that raises the symptom you personally get. There is no single "gentlest GLP-1," only a gentlest one for your body.

Why the numbers look scarier than your experience

Two facts stop these figures from being as frightening as they first read.

First, relative risk is not the same as your odds of suffering. A relative risk of 3 for nausea does not mean three in four users vomit into a bucket. It means nausea was reported about three times as often as on placebo, and placebo groups report a fair amount of nausea too. Most people who get GLP-1 nausea describe it as manageable queasiness, not constant sickness. For a grounded look at how common and how serious each effect really is, our overview of what the science actually shows on GLP-1 side effects puts the risks in proportion.

Second, and more practically useful, the analysis found a clear dose-response pattern. For every drug, side effects rose fastest at the lowest starting doses, then tended to plateau as the dose climbed. In plain terms, the misery is front-loaded. Most of the nausea, vomiting and diarrhea shows up while you are titrating upward in the first weeks, and the body adapts rather than getting steadily worse. That is the whole logic behind the slow, stepwise dosing schedule every GLP-1 drug uses.

What this means if you are choosing or switching

The temptation after reading a ranking like this is to ask your prescriber to switch you to whichever drug scored best. That is usually the wrong lesson, for three reasons.

1. The differences are small for most symptoms. Across the injectables, the nausea gap is a rounding error. You are unlikely to notice the difference between a relative risk of 2.9 and 3.1.
2. Switching trades one profile for another. Every GLP-1 drug raised nausea two- to nearly five-fold versus placebo. There is no side-effect-free option to switch to, so you may swap tirzepatide's diarrhea for semaglutide's reflux without coming out ahead.
3. Dosing usually beats switching. Because the effects are dose-dependent and worst during titration, a slower climb or a short hold at your current dose fixes far more problems than a change of drug. That decision belongs with your prescriber, not with a solo dose adjustment.

The drug you can actually tolerate long enough to reach an effective dose usually beats the one that looked marginally gentler on paper. If weight-loss results are also part of your decision, our guide to which GLP-1 is best for you weighs efficacy alongside tolerability.

Managing the symptoms, whichever drug you take

Since no GLP-1 drug escapes gut side effects, the winning move is managing them rather than chasing a magic bullet. The basics have decent evidence: eat smaller and more frequent meals, stop at the first sign of fullness, go easy on high-fat food that lingers in a slow stomach, and sip fluids between meals rather than with them. For constipation, fiber paired with real hydration and a short walk after eating help. We laid out the full playbook in how to manage GLP-1 side effects without stopping the drug.

There is a quieter problem underneath the symptoms, and it is the one worth the most attention. When nausea and early fullness cut how much you eat, protein and key micronutrients get hard to reach from food alone, which is where muscle loss and nutrient gaps start. That is the gap GLP-1 Shield is built to close: not a band-aid for nausea, but the nutrient support a shrinking appetite makes hard to get otherwise. If you want the specifics, our GLP-1 supplement protocol lays out what to take and how much.

In one sentence: no GLP-1 drug is gentle, the differences between them are smaller than the raw numbers suggest, and how you dose and eat matters more than which drug you pick.

Frequently asked questions

Which GLP-1 causes the least nausea?

In this 39-trial network meta-analysis of non-diabetic adults, cagrilintide (an amylin drug) had the lowest nausea signal, with a relative risk of 2.30 versus placebo. The injectable GLP-1 drugs clustered close together: exenatide 2.66, tirzepatide 2.90, semaglutide 2.95 and liraglutide 3.09. The gaps are small and overlap, so no injectable is clearly gentler than another. Orforglipron, the pill, had the highest nausea estimate but from fewer trials.

Does tirzepatide cause more side effects than semaglutide?

For nausea the two are almost identical, with a relative risk of 2.90 for tirzepatide and 2.95 for semaglutide versus placebo. Tirzepatide stood out for vomiting (RR 13.23, though from few events), diarrhea (3.35 versus 1.77) and constipation (3.36 versus 2.10). Semaglutide was the only drug linked to more acid reflux (RR 2.43). Which feels worse depends on the symptom you get, not an overall winner.

Will switching GLP-1 drugs stop my nausea?

Not reliably. In this analysis every GLP-1 drug raised nausea two- to nearly five-fold versus placebo, so switching trades one drug's side-effect profile for another rather than removing the problem. Because the risk is dose-dependent and eases after the titration phase, a slower dose climb or a brief hold, agreed with your prescriber, usually helps more than switching drugs.

Do GLP-1 side effects get better over time?

Usually yes. The analysis found a dose-response pattern: nausea, vomiting, diarrhea and constipation rise fastest at the lower starting doses, then tend to plateau at higher doses as the body adapts. That is why most people feel worst during the first weeks and after each dose increase, and steadier once they settle at a maintenance dose.