Low-dose semaglutide cuts weight in HARMONY

A real-world cohort study from China found that semaglutide at doses of 0.5 or 1.0 mg weekly produced clinically meaningful weight loss in adults with obesity, even though these doses sit well below the 2.4 mg maintenance dose approved for weight management in the US and EU^[1]. The HARMONY cohort, published on 15 June 2026 in Diabetes, Obesity and Metabolism, followed 143 participants through 12 weeks and 113 through 24 weeks of treatment with once-weekly subcutaneous semaglutide plus lifestyle counselling.

What the numbers show

At 12 weeks, participants lost a mean 7.3% of body weight, which translates to 6.7 kg in absolute terms. By 24 weeks that figure climbed to 9.9% mean body-weight loss (9.1 kg). The responder rate was high: 76.1% of participants achieved at least 5% total body-weight loss by week 24^[1].

For context, the international STEP-1 trial of semaglutide 2.4 mg reported 14.9% mean weight loss at 68 weeks in a predominantly Western population. HARMONY used doses that are two to nearly five times lower than STEP-1 and ran for a shorter period. The fact that nearly 10% mean weight loss appeared in 24 weeks at these lower doses is noteworthy, though a direct comparison between the two studies is not valid given differences in design, dose, duration, and population.

It is also worth noting the distinction between percentage weight loss and absolute kilograms. A 9.1 kg loss at 24 weeks can represent a significant change in metabolic risk markers for someone with a starting BMI in the 30 to 35 range, which is where much of the Chinese obesity population sits. Body-composition data (fat mass versus lean mass) was not reported in the HARMONY abstract, so we cannot tell whether the loss was predominantly adipose tissue.

Why lower doses matter in practice

In China, semaglutide is approved under the brand name Ozempic for type-2 diabetes at doses up to 1.0 mg weekly^[3]. There is no locally approved 2.4 mg obesity indication. Clinicians are therefore working with the doses available, prescribing them off-label for weight management and pairing the medication with structured lifestyle counselling. The HARMONY data offers the first structured look at what those lower doses actually deliver in routine Chinese clinical practice^[1].

The same situation applies in other markets. In the EU, the EMA authorises Ozempic at up to 1.0 mg for type-2 diabetes, while the higher-dose obesity product (Wegovy, 2.4 mg) carries a separate marketing authorisation^[2]. In countries where Wegovy is unavailable, subject to supply constraints, or not reimbursed by national health systems, clinicians sometimes prescribe the lower diabetes dose for weight management. HARMONY quantifies, for the first time in a Chinese population, what that pragmatic approach yields.

Who responded more and who responded less

The cohort split along two clinically relevant lines. Participants without type-2 diabetes lost considerably more weight than those with diabetes: 12.4% versus 7.4% body-weight reduction at 24 weeks (p < 0.001). Separately, participants without intrapancreatic fat deposition lost more than those with it: 11.6% versus 9.0% (p = 0.043)^[1].

The diabetes finding is consistent with patterns observed in Western GLP-1 receptor agonist trials. Patients with type-2 diabetes tend to lose less weight on the same dose of semaglutide than patients without diabetes, a difference attributed in part to underlying metabolic differences and concurrent medications. The intrapancreatic fat finding is less well documented in the literature and may warrant further study.

What this study does not tell us

HARMONY is observational, not randomised. There was no placebo arm, so the contribution of lifestyle counselling versus the drug itself cannot be separated. The cohort is modest in size (143 participants at baseline, 113 completing 24 weeks) and the follow-up stops at six months. We do not know whether the weight loss holds beyond that window or what happens after discontinuation.

The publication also does not report a detailed adverse-event profile for this population, which leaves an important gap. Gastrointestinal side effects (nausea, vomiting, diarrhoea) are well documented in the international trial programme at the 2.4 mg dose. Whether lower doses produce a milder side-effect burden in Chinese adults is a question this study does not answer.

There is also the question of selection bias. Patients in a real-world cohort who stay on treatment for 24 weeks may be systematically different from those who discontinue early. The drop from 143 participants at 12 weeks to 113 at 24 weeks means roughly 20% did not complete follow-up, and we do not know why. A randomised controlled trial with a longer follow-up period would be needed to address durability, tolerability, and dropout.

Where this fits for readers

The HARMONY findings add a data point for a question many readers ask: does semaglutide work for weight loss at the lower doses available in their country? The answer from this cohort is yes, with two caveats. First, the weight loss is smaller than what the high-dose trials report. Second, we cannot tell how much of the result came from the drug versus the lifestyle programme that accompanied it.

For readers interested in how semaglutide is regulated in specific countries, including which doses are approved and for which indications, see the semaglutide regulation pages. As always, any decision about dosing or off-label use should involve a conversation with a prescribing clinician.