There is a huge amount of interest among medical researchers in finding a way to maintain weight loss after stopping injectable weight-loss medications. It’s worth noting that similar interest doesn’t exist for any medications we prescribe for other chronic health problems. High cholesterol, for instance. For four decades, we’ve been relying on Lipitor to lower our cholesterol. In all that time, I’ve never heard anyone suggest looking into maintaining Lipitor’s effectiveness after stopping it; everyone just accepts that we have to keep taking it. Why is it so unacceptable to us to keep taking weight loss medication? I don’t think it’s the cost. It’s true that semaglutide and tirzepatide can be expensive, but even people who get them for free or nearly free ask me about tapering them off. It’s an interesting phenomenon, and it says something about how we regard obesity.
So far, we have looked at a couple of tapering strategies that have been tried, namely, reducing the dose and increasing the time interval between doses. Neither of those strategies have worked in any meaningful way. The relevant studies provide no cause to be optimistic that tapering will work in the real world. We now arrive at another strategy of interest, namely, switching from injectable weight-loss medications to something else for maintenance.
Exiting Injectable Therapy by Switching to a Substitute Medication
A few studies have considered whether it might be possible to maintain weight loss after stopping injectables by substituting another drug. Metformin has been looked at, and there is at least one new, experimental medication that is being considered for this purpose.
As far as metformin goes, there is one published report out of Slovenia for a trial that looked at using it as a substitute for semaglutide. Metformin, by the way, is an old drug. It’s a twice-daily pill that has been in regular use in the treatment of type-2 diabetes since the early sixties.
The trial in question treated twenty-five women (all of whom suffered from polycystic-ovary syndrome) with sixteen weeks of mid-dose semaglutide and high dose-metformin.[i] Alas, there was no control group. After sixteen weeks, semaglutide therapy was discontinued and metformin continued. They were followed for two more years after cessation of semaglutide.
The study found that the study subjects lost 9% of their total body weight while on the combination of semaglutide and metformin for sixteen weeks. Two years after stopping semaglutide (and continuing metformin), the average total-body weight loss retained by the women was 5%. The authors’ remarks about the study were refreshingly restrained compared to many,[ii] but they did suggest that their method was “a potential novel strategy for partially overcoming [the weight-regain problem],” and they concluded that, “the role of metformin in attenuation of weight regain after semaglutide discontinuation needs to be explored in randomized controlled studies.”
I would not have drawn that conclusion. The medications simply performed as expected, and there was nothing in the results that could be interpreted as evidence that metformin will maintain semaglutide weight loss. For one thing, they treated the subjects with a short (sixteen week) course of semaglutide. It is therefore unsurprising that the subjects lost less than semaglutide patients normally do—just 9% of their total body weight. It’s well understood that reaching maximum weight loss (the “plateau”) with injectable medication takes longer, so their patients only saw part of it.
Neither is it surprising that the metformin-treated patients maintained 5% total-body weight loss after two years. It is already well established that metformin can—for those who can tolerate it, see below—produce around 5% total body weight loss. (The first peer-reviewed article on using metformin to treat obesity was published in 1965.[iii] There are many such papers.) The finding that the study subjects landed at 5% total-body weight loss after two years on high-dose metformin thus represents the expected effectiveness of the drug. It’s not news.
You might be thinking to yourself that getting 5% weight loss out of a pill sounds rather good, but don’t expect the authors’ enthusiasm for metformin to catch on. Metformin is a decent diabetes drug, but it’s notorious for causing diarrhea,[iv] which can be overwhelming. It’s a commonly cited reason for discontinuing metformin.[v] But for that, it would have caught on already.
There is another clinical trial still running that involves using metformin to prevent post-GLP-1 weight regain. I came across it while browsing trials on Clinicaltrials.gov.[vi] The trial is looking at preventing post-GLP-1 weight regain with three strategies: metformin alone, metformin plus naltrexone, and metformin plus rapamycin. Unfortunately, there aren’t any results posted yet, so I suppose we must stay tuned.
I came upon another an interesting article about a drug under consideration for the GLP-1 problem. The drug is so new that it doesn’t even have a real name yet. For now, it’s just called TIX100. It’s an oral form of a new drug class, namely, inhibitors of thioredoxin-interacting protein. Researchers from the University of Alabama discovered that TIX100 prevents weight regain after cessation GLP-1 therapy—in mice, that is.[vii] At this point, we don’t know if the medication is even safe for humans to take, so expect it to be a long, long time before TIX100 is available to humans for any purpose.
It’s worth stopping to consider the fact that the investigators who discovered this odd new class of enzyme inhibitors quickly thought of testing whether it might solve the GLP-1 problem, which is a problem nobody knew existed just a few years ago. It illustrates the extent to which market forces sometimes influence the thought processes of medical researchers.
[i] i.Jensterle M, Ferjan S, Janez A. The maintenance of long-term weight loss after semaglutide withdrawal in obese women with PCOS treated with metformin: a 2-year observational study. Front Endocrinol (Lausanne). 2024 Apr 11;15:1366940. doi: 10.3389/fendo.2024.1366940. PMID: 38665260; PMCID: PMC11043580.
[ii] I think so, anyway. In fairness, I do not know how faithful the translation was from Slovenian to English.
[iii] PEDERSEN J. THE EFFECT OF METFORMIN ON WEIGHT LOSS IN OBESITY. Acta Endocrinol (Copenh). 1965 Jul;49:479-86. doi: 10.1530/acta.0.0490479. PMID: 14332328.
[iv] Drug compendia traditionally list adverse drug reactions in descending order of frequency. For metformin, diarrhea is always at the top of the list.
[v] I can say from long experience: convincing patients to stick with metformin for weight-loss is a proposition that is “flush” with difficulty. Ha ha.
[vi] The story behind Clinicaltrials.gov is interesting. The bottom line is that there is a need to register clinical trials before they have results, because it avoids a situation where a study sponsor with a financial interest in the outcome might be tempted to hide significant and important results if they were embarrassing or potentially costly for the sponsor.
[vii] S.Jo, J.Chen, G.Jing, G.Xu, P.Li, and A.Shalev, “Prevention of Weight Regain After GLP1RA Cessation With Oral TIX100,” Diabetes, Obesity and Metabolism28, no. 6 (2026): 5339–5342, https://doi.org/10.1111/dom.70633.