GLP-1 therapy in a digital clinic cut women's weight 16 kg, with no comparison group

Women who started a GLP-1 drug through a structured online clinic lost an average of 16.1 kg over six months, according to an observational study of more than 8,000 patients in the United Kingdom [s1]. The result, published in Nature Medicine, is large — but it comes from a programme with no comparison group, and the authors are explicit that this limits what can be read into it [s1].

The study followed 8,127 women who began GLP-1 receptor agonist therapy — 99.8% of them on tirzepatide — through a digital service that paired the prescription with mandatory blood-test monitoring, weekly video check-ins and clinician-guided dose increases [s1]. Mean weight reduction was 16.1 kg (95% confidence interval 15.9 to 16.2), equal to 15.5% of starting body weight [s1]. Among the women, 83.1% lost at least 10% of their weight and 22.4% lost at least 20%, and the authors report that the responses were consistent across age, baseline weight and socioeconomic group [s1]. The bibliographic record for the paper is dated 16 October 2026 [s2].

What was measured

Alongside weight, the programme tracked metabolic blood markers in the subset of women who had paired before-and-after results [s1]. Mean glycated haemoglobin — HbA1c, the standard measure of average blood sugar — fell by 3.3 mmol/mol (95% CI 3.1 to 3.4) [s1]. Readings normalised in 84.5% of women who had started in the diabetes range and in 96.3% of those who had started in the prediabetes range [s1]. Lipids improved too: total cholesterol fell by 0.77 mmol/l, LDL ("bad") cholesterol by 0.61 mmol/l and triglycerides by 0.47 mmol/l [s1]. Retention was 91.4% at six months, and the authors report no serious treatment-related adverse events [s1].

These are the kind of numbers that drug trials are built to produce, and GLP-1 drugs such as tirzepatide and semaglutide have already shown substantial weight loss in randomised studies. What is new here is the setting: the study is an attempt to show that the same drugs, wrapped in a remote monitoring-and-coaching model, can be delivered at scale outside a trial or a traditional clinic [s1].

Why the design matters

The honest frame is the one the authors themselves supply. This was an observational study, not a randomised trial, and that distinction governs how much weight the headline figure can bear [s1]. There was no control group, so the 16.1 kg cannot be cleanly separated from what the same women might have achieved with the drug alone, with a different programme, or with the motivation that led them to sign up in the first place [s1].

Three further limits are worth keeping in view, each named in the paper. The population was self-paying, which tends to select for people with the means and determination to stick with treatment — not the general population of women who might be offered these drugs [s1]. Follow-up ran to six months, so the study says nothing about whether the weight stays off, a question that matters because stopping GLP-1 drugs is commonly followed by regain. And the people who remained in the programme are, by definition, the ones who did not drop out; a 91.4% retention figure is high, but outcomes are measured in those who stayed [s1].

The missing comparator is the crux. A single-arm study can show that an outcome occurred; it cannot show that the programme caused it, because there is no parallel group of similar women managed another way. That is not a flaw unique to this study — it is the defining feature of this kind of evidence, and it is why implementation studies are read as a signal that a care model is feasible rather than proof that it is better than the alternatives.

What it adds

Taken on those terms, the study does say something useful. It provides prospective, real-world evidence that a digital-first model — remote prescribing with enforced blood monitoring and regular human contact — can keep a large group of women engaged and can reproduce, in routine use, the scale of weight and metabolic change seen in trials [s1]. For a treatment whose real-world delivery is expanding faster than the evidence on how best to deliver it, that is a meaningful contribution, and the enforced biochemical monitoring is a notable safety feature in a market where pills and injections are often sold with far less oversight.

What it does not do is establish that this programme beats simpler or cheaper ways of prescribing the same drug, or that the results would hold in women who are not paying out of pocket, or that the weight loss endures past half a year [s1]. Readers weighing a digital weight-loss service should treat the 16 kg as what a motivated, self-selected group achieved under close monitoring over six months — an encouraging real-world snapshot, not a controlled comparison.

This article is informational and does not constitute medical advice. Decisions about weight-loss medication belong with a patient and their clinician.