Can a Weight-Loss Drug Burn More Energy Without Making You Eat Less?

The question nobody asks about weight loss

Most weight-loss talk is about eating less. That makes sense: semaglutide and tirzepatide work mainly by turning down appetite, and they do it well.

But weight depends on two numbers. Energy in is the calories you eat. Energy out is what you burn, mostly just staying alive. The resting part is called the resting metabolic rate.

The two are linked. In a 2016 NIH study, 14 former Biggest Loser contestants had a resting metabolic rate about 500 calories a day below what their size and age predicted six years later. That group lost an extreme amount of weight, so it marks the far end, not a typical result. It still explains the research question: could a medicine raise energy out, not only lower energy in?

The AMPK fuel gauge

Nearly every cell carries a protein called AMPK, a kind of fuel gauge. When energy runs low, as in exercise or fasting, AMPK switches on and tells the cell to burn stored fuel and take up glucose. A drug that nudges the same signal might get cells to use more fuel. Researchers call these compounds exercise mimetics.

The name oversells it. Exercise also builds muscle and bone, trains the heart and changes the brain. A drug that touches one signaling network does none of that.

ATX-304 is an investigational oral small molecule, not a peptide. Its developer, Amplifier Therapeutics, a Cambrian Bio company, says it activates AMPK and increases mitochondrial proton conductance. In plain terms, it makes the cell's power plants lose a little more energy as heat.

What ATX-304 researchers actually found

In June 2026, Cambrian presented Phase 1b results at the American Diabetes Association meeting. Phase 1b is an early step, mostly checking safety.

The study enrolled 23 adults with obesity and prediabetes. It was randomized, double-blind and placebo-controlled: eight weeks of ATX-304 or placebo, then an optional eight-week open-label extension. The developer reports:

• Resting metabolic rate rose 8%.
• Liver fat (measured by MRI), visceral fat (deep belly fat) and triglycerides fell.
• Adiponectin, a fat-tissue hormone tied to healthier blood sugar handling, rose.
• Weight loss was minimal at the studied exposure.

Side effects were mostly mild and similar to placebo, with no rise in core body temperature or heart rate. Those were watched because burning more energy could, in theory, make the body run hot.

These are separate measurements, and none equals pounds lost. Liver and visceral fat can shrink while the scale barely moves. For scale only: if someone's resting burn were 1,800 calories a day, 8% would be about 144 calories, a small snack. That is an illustration, not a result. The company did not give the change in calories or say whether the 8% is against placebo or each person's starting point. The conference abstract itself reports a rise of up to 33% from baseline, without saying what up to refers to, and the two figures have not been reconciled.

The curious mouse experiment

ATX-304's earlier version was called O304. A peer-reviewed 2018 JCI Insight paper tested it in mice on a high-fat diet. Mice given O304 gained almost no weight for two weeks even though they ate more. Then the groups were switched. Mice moved onto O304 lost weight quickly while eating relatively more. Mice moved off it gained weight while eating relatively less. The pattern held at 30°C, a warm room that removes the need to burn extra energy for heat.

The weight followed the treatment, not the appetite, which is what a rise in energy out would look like.

It is not proof for people. That experiment used five mice per group, and animal results often fail to carry over. The paper's human study lasted 28 days in 65 adults with type 2 diabetes and looked at blood sugar and blood vessels, not weight.

How different is this from GLP-1 medicines?

Semaglutide and tirzepatide act on gut-hormone receptors that cut appetite, so people eat less (see how long they take to work). ATX-304 is designed to change how cells use fuel instead.

No head-to-head trial exists, so nobody can say which works better or whether the effects would add up. A 2025 conference abstract by company-affiliated researchers reported that in obese mice, ATX-304 plus semaglutide produced more weight loss than semaglutide alone, and that the loss was fat, not lean mass. That was eight mice per group. No human combination study has been reported, so no combination is established. The combinations page covers what is known about pairing weight-loss medicines. Muscle loss is a real concern with current medicines (see protecting muscle on GLP-1 medicines), but a mouse finding does not answer it for people.

What about other energy-expenditure approaches?

ATX-304 is one idea among several. This table compares mechanisms and how far the evidence has come. It is not a ranking, and none of these is an approved medicine.

SANA's small trial reported weight and blood sugar changes within two weeks, which ATX-304's Phase 1b did not show at the studied exposure. The studies differ in size, length and design, so they cannot be compared directly. Products sold online under these names are not tested drugs, and the findings here do not apply to them.

The weight-loss question remains unanswered

Cambrian plans two Phase 2 studies called REWIRE. REWIRE-1 will test higher exposures and look at muscle function and lipid metabolism. REWIRE-2 aims to show proof of concept for weight loss in people with obesity. No start dates have been given, and the Atlas found no registry listing for either as of November 8, 2026.

A real answer needs weight loss against placebo over months, a split between fat and muscle, evidence the effect lasts, careful safety monitoring at higher exposures (heart rate and body temperature especially), and peer-reviewed results. For registered weight-management trials, see the pipeline page.

Bottom line

A promising metabolic signal is not yet a proven weight-loss medicine.

According to its developer, ATX-304 raised resting energy use in a small early study and moved fat measures the right way. Whether that becomes meaningful, lasting weight loss with acceptable safety is for Phase 2 to show. Until then it is an idea to watch, not a treatment option. Ask your prescriber about anything new.

How strong is this evidence?

• The June 2026 Phase 1b numbers come from an ADA conference abstract (1782-P) and Cambrian's June 18 release on it. Abstracts have not had full peer review. The abstract reports resting metabolic rate up to 33% from baseline and the release says 8%. Neither gives group sizes or starting values, and the two figures have not been reconciled.
• An EU registry entry (CTIS 2023-505967-36-00) lists a two-part ATX-304 Phase 1B that ended in May 2025 with 72 planned participants. It does not say how that relates to the 23 reported.
• The semaglutide result is a mouse abstract, and its authors are company-affiliated, as are several authors of the SANA paper. The REWIRE studies are announced intentions.
• For chemical identifiers, study-by-study summaries and a longer source list, see the ATX-304 entry on PeptideBlog, a separate reference site.
• The Atlas gives no dosing guidance for any compound here.

Frequently asked questions

Is ATX-304 a weight-loss drug?

Not yet. It is investigational with no approved use, and the one study that measured resting metabolic rate reported minimal weight loss. A weight-loss study is planned.

Can I get ATX-304 or take it with Wegovy or Zepbound?

No. It is only given to people in the developer's trials, and no combination study in people has been reported. Talk to your prescriber before adding anything.

Does a faster metabolism mean faster weight loss?

Not automatically. The body often pushes back with appetite. Mice on O304 ate more yet stayed lighter, and whether people respond the same way is untested.

When will we know if it works?

No dates have been given. Phase 2 studies often take a year or more, and Phase 3 would follow before any approval request.