The Gastric Balloon: Six Months of Help, Then What?
A bag inside a bag works while it is in there, and a network meta-analysis agrees to six decimal places.

One way medicine treats obesity is to put a balloon in a person’s stomach, leave it there for a while, and then take it out. That is the gastric balloon, more or less: an object that takes up room where dinner would otherwise go, one of the “endoscopic bariatric therapies” that sit somewhere between a prescription and an operation. (The title above says six months, which is the general shape of it. The research mostly measures things “at time of IGB removal,” which is the clinical way of saying the same thing without committing to a calendar.)
Does a gastric balloon work? While it is in, yes. A network meta-analysis of randomized trials found that at the time of balloon removal, all currently available FDA-approved endoscopic bariatric therapies produced “statistically significantly higher percentage total weight loss than control interventions.” After it comes out is the less cheerful part. A narrative review on GLP-1 drugs, surgery and balloons put balloon weight loss at roughly 10 to 15% of total body weight, with typically temporary results after removal. A review titled “Endobariatrics: well past infancy and maturing rapidly” put it more flatly: “Weight loss achieved with intragastric balloons is not sustained long-term after device removal.”
How a gastric balloon works, the dumb version
Here is the version a reasonably bright ten-year-old could produce. The stomach is a bag. If you put a smaller bag inside the bigger bag, there is less room in the bigger bag. Less room means less food. Less food means less weight. Then you take the smaller bag out, there is room again, and the food (and the weight) come back.
That’s it. That’s the model.
I mean, it is a little unfair to call it dumb, because as far as the evidence in front of me goes, it is basically correct. It makes two predictions: the balloon helps while it is there, and the help fades once it is gone. Both of those predictions turn out to be what the literature says. The dumb version gets there in one paragraph.
The balloon is, structurally speaking, filler. (Filler is a respected term in plenty of fields. It is also what people call the part of a thing they suspect is not doing the real work.) The question the sophisticated version asks is whether the filler does anything beyond filling.
The two balloons the trials actually compare
The network meta-analysis compares two balloons, and they are made by two different companies in two different states, which is the kind of detail that research papers include and that I find oddly comforting. The Orbera balloon is made by Apollo Endosurgery, headquartered in Austin, Texas. The Spatz3 is made by Spatz Medical, headquartered in Fort Lauderdale, Florida, and it is described as an “adjustable” gastric balloon.
Adjustable. That is the product’s whole pitch, presumably: a balloon you can tune rather than a balloon you simply have. I don’t know the mechanics of the adjustment from the sources here, and I am not going to make them up. What the sources do tell us is how the two balloons rank against each other, and against endoscopic sleeve gastroplasty (ESG), which is a procedure that reshapes the stomach with stitches instead of filling it with an object. ESG has its own article on this site; here it mostly serves as the yardstick.
You can picture the product meeting. “Hey guys, what if the balloon was adjustable,” says someone, and everyone nods, because who would argue against more control over the thing in a patient’s stomach.
The researchers searched MEDLINE, EMBASE and Cochrane CENTRAL for balloon studies “from inception” and for ESG studies from 2013 until May 2023. “From inception” is a wonderful phrase. It means they went back to the beginning of the balloon. All of it.
What the network meta-analysis found, to six decimal places
A network meta-analysis is the sophisticated version. Instead of one trial comparing one thing to a control, it stitches many randomized trials together so that treatments never tested head to head can still be ranked against each other through the shared comparisons. The output includes a P score, which (as far as I can tell from how the paper uses it) is a ranking where higher means more likely to be the best at weight loss.
At the time of removal, the Spatz3 had a P score of 0.822894.
I want to sit with that for a second. Not 0.82. Not “about four-fifths of the way to the top.” Zero point eight two two eight nine four, six decimal places, the precision of a lab scale weighing a grain of salt, applied to the question of how much weight people lost with a balloon in their stomach compared with other people who had a different balloon in theirs. Somewhere a statistician produced the 4 at the end and decided it belonged in the record. That’s fine. It is correct, in the sense that the arithmetic produced it. Whether the sixth decimal place of a weight-loss ranking carries information is, I assume, a question the statistician also had, briefly.
The rest of the table goes like this. ESG had the highest P score at removal (0.889272) and at the 12-month follow-up (0.899469). The Spatz3 scored 0.822894 at removal and 0.842773 at 12 months. The Orbera scored 0.536968 at removal and 0.507165 at 12 months. At the follow-up closest to 12 months, both ESG and the Spatz3 were more effective than the Orbera, “with no statistically significant difference between ESG and Spatz3,” in the network meta-analysis’s words.
That last part is the interesting bit. The adjustable balloon is, statistically, hard to tell apart from the stitched-up stomach at a year. The Orbera ranks lower. Adjustability, whatever it involves, seems to buy something.
What happens to the weight after the balloon comes out
Now notice what the sophisticated version has not done. It has measured weight loss at removal, and it has measured weight loss at roughly 12 months, and it has ranked balloons against each other at both points to six decimal places. It has not overturned the dumb version’s second prediction. The reviews say it plainly: balloon weight loss is “not sustained long-term after device removal,” and the narrative review calls the results typically temporary.
How much comes back, and how fast, is not settled by the facts in front of me. Nothing here gives a regain curve for balloons the way some drug studies give one. (For contrast, and only for contrast: people in the SURMOUNT-CN extension who stopped tirzepatide regained approximately 9 to 12% of body weight within 6 months but kept about half of the net weight lost. Things that stop being used tend to stop working. This is not a unique property of balloons.)
So here is where the two versions stand. The dumb version: bag in a bag, works while it’s in, stops when it’s out. The sophisticated version: a network meta-analysis of randomized trials across every balloon study since inception, P scores out to the millionths, adjustable versus fixed, ESG as a benchmark, and the conclusion that balloons beat controls while present and do not hold the loss after removal. The longer road has better scenery. Same destination.
Between us, you and I both knew how this was going to end the moment the word “removal” appeared in the outcome measure. Nobody designs a study around “time of removal” for a treatment they expect to keep working after it leaves.
Where the balloon sits next to ESG and everything else
The comparison that matters for this site is not balloon versus balloon but balloon versus the things that stay. A 2022 review characterized ESG as potentially more effective and durable than intragastric balloons, with fewer adverse events than bariatric surgery. In the MERIT trial, ESG produced 13.6% mean total body weight loss at 52 weeks versus 0.8% for controls, and 68% of ESG participants maintained at least 25% excess weight loss at 104 weeks. The narrative review put sleeve gastrectomy at roughly 20 to 30% total body weight loss, “most durable,” and GLP-1 drugs at around 10 to 15%, the same range it gave balloons. Those other comparisons get their own articles here, so I will leave them there.
What this produces is a funny niche. The balloon lands in roughly the same weight-loss range as the drugs while it is in, and its standout version ranks alongside ESG at a year, but its defining feature is that it is designed to leave. Every other treatment on the menu either stays (stitches, surgery) or can be continued (drugs, or whatever). The balloon has an end date built into the product.
This invites an obvious thought experiment, which I am making up entirely and which no source here describes. Suppose the answer to “it stops working when you take it out” were simply “never stop putting one in.” A balloon, then a removal, then another balloon, every few months, forever. A subscription. Balloon-as-a-service. You would want a loyalty program, obviously, and an annual plan at a discount, and eventually some consultant would point out that the customer’s stomach has been continuously occupied for nine years and ask whether this is, functionally, just a permanent device with very frequent maintenance visits. At which point you have reinvented the thing the balloon was supposed to be an alternative to, at the cost of a great many endoscopies. Again, none of that is real. It is just where the logic goes if you follow it far enough without stopping.
The movie “Up” is about a man who ties enough balloons to his house that it floats away, and the whole plot works because the balloons stay tied on. Nobody in the film unties them at the six-month mark and expects the house to keep flying.
I should say that I write explainers about the inside of other people’s stomachs and have no business second-guessing a network meta-analysis, lol, so I will not. The paper is careful and the numbers are what they are. The dumb version and the sophisticated version agree. The sophisticated version just shows its work.
Of all the findings in that network meta-analysis, the one most certain to still hold a year after removal is that Spatz Medical is in Fort Lauderdale and Apollo Endosurgery is in Austin.
Sources
- Comparative effectiveness of balloons, adjustable balloons, and endoscopic sleeve gastroplasty: a network meta-analysis of randomized trials
- GLP-1 medications versus surgery and balloon: evaluating cost-benefit in weight loss
- Endobariatrics: well past infancy and maturing rapidly
- 1672-P: Weight Changes after Tirzepatide Cessation in Chinese Adults with Overweight/Obesity—A Real-World Follow-up Study of SURMOUNT-CN Trial
- Endoscopic sleeve gastroplasty for treatment of class 1 and 2 obesity (MERIT): a prospective, multicentre, randomised trial
Bariatric Alt explains published research. It is not medical advice, and it never recommends a surgeon, a clinic, a dose or a product. Talk to a clinician about which option fits you.