Dapagliflozin plus semaglutide: BMI classes

A retrospective observational cohort published in Current Medical Research and Opinion on 30 June 2026 tracked 328 patients in Saudi Arabia receiving dapagliflozin alone or in combination with once-weekly semaglutide across five BMI categories^[1]. The headline finding: patients in obesity class III showed the greatest absolute BMI reduction, but when the researchers adjusted for starting weight, the proportional response was uniform across all weight classes.

That uniformity matters for how clinicians read real-world data. A patient losing 1 kg/m² from a starting BMI of 45 is not responding better pharmacologically than a patient losing 0.5 kg/m² from a starting BMI of 27. They are both losing roughly the same fraction of their baseline. What the absolute numbers reflect is starting BMI, not differential drug effect. The authors argue this point explicitly^[1].

Study design and population

The cohort enrolled 328 adults distributed across five BMI bands: normal weight (n=23), overweight (n=57), obesity class I (n=87), obesity class II (n=89), and obesity class III (n=72). Follow-up ran from 90 to 365 days. The intervention was dapagliflozin monotherapy in most participants, with a subset receiving add-on once-weekly semaglutide. The study was conducted in Saudi Arabia, where obesity prevalence and the metabolic disease burden are both high, and where SGLT2 inhibitors have been integrated into clinical practice for type-2 diabetes management.

Dapagliflozin is a sodium-glucose cotransporter-2 (SGLT2) inhibitor. It lowers blood glucose by blocking glucose reabsorption in the renal proximal tubule, resulting in glucosuria. The mechanism carries a secondary weight effect: caloric loss through excreted glucose, along with diuresis and a reduction in visceral fat. A 2025 narrative review of SGLT2 inhibitors across metabolic syndrome components confirmed these secondary effects include weight loss and reduced waist circumference, though the magnitude is generally modest compared with GLP-1 class agents such as semaglutide or tirzepatide^[3].

The BMI findings by class

Absolute BMI reduction was greatest in the obesity class III group at approximately -1.05 kg/m², compared with approximately -0.09 kg/m² in the normal-weight group^[1]. Class I and class II patients fell between those poles. The gradient follows starting weight closely: the heavier the patient, the more kilograms per square metre lost in absolute terms.

Once the researchers modelled proportional response, the gradient flattened. The adjusted analysis found that roughly 5% weight reduction rates did not differ significantly across BMI categories. This finding supports the hypothesis that the pharmacological signal from dapagliflozin is consistent regardless of obesity class, and that absolute differences are a mathematical consequence of baseline body mass rather than a sign that the drug works better in more severe obesity.

Where semaglutide added effect

The add-on semaglutide subgroup showed particular promise in obesity class I patients^[1]. This is not a surprise mechanistically: semaglutide works as a GLP-1 receptor agonist, reducing appetite and slowing gastric emptying through central and peripheral pathways that are distinct from the glucose-excretion mechanism of dapagliflozin. Combining an SGLT2 inhibitor with a GLP-1 receptor agonist covers two largely non-overlapping biological routes to weight reduction.

Semaglutide's weight-loss track record in real-world settings is increasingly well documented. A 2026 prospective cohort from the United Arab Emirates followed 278 patients with type-2 diabetes on semaglutide for 12 months, reporting a mean weight reduction of 4.77 kg at one year, with 43% of participants achieving at least 5% weight loss^[2]. The Saudi Arabia study does not report separate semaglutide weight-loss figures with the same granularity, but the direction of benefit in the combination arm is consistent with that regional evidence base.

What this study does not tell us

The retrospective design is the main caveat. Patients were not randomised to dapagliflozin versus dapagliflozin plus semaglutide; the combination arm reflects clinical decisions by prescribers who may have selected it for patients with heavier baseline BMI or more difficult glycaemic control. That selection bias means the class-level BMI comparisons cannot be read as evidence that one obesity class responds better to the combination than another. The sample in some subgroups, notably the normal-weight arm (n=23), is too small to draw firm conclusions.

The study also does not report long-term outcomes beyond one year. Dapagliflozin's weight effect is known to plateau: most of the body weight change occurs in the first few months, then stabilises. Whether the modest BMI reductions seen here are durable beyond 12 months in a Saudi population on dapagliflozin with or without semaglutide is a question the study cannot answer.

Finally, this is a Middle Eastern real-world cohort. Genetic variation, dietary patterns, physical activity norms, and rates of insulin resistance differ from European or North American trial populations. The authors do not claim the proportional uniformity finding would replicate in a different healthcare system.

What to take from it

The practical implication is narrow but useful. Clinicians should not expect dapagliflozin to produce larger proportional BMI reductions in patients with class III obesity simply because the absolute numbers look bigger. The drug is not working harder; the denominator is larger. For patients where deeper weight reduction is the goal, combining dapagliflozin with a GLP-1 receptor agonist like semaglutide appears to add meaningful effect, particularly in class I obesity. The full clinical context for semaglutide in weight management is on the semaglutide peptide page.