Does starting heavier change what these drugs do for you? A 2026 retrospective study of adults with type 2 diabetes at three centers in Riyadh looked at exactly that, splitting patients by whether their body mass index was above or below 30 and following them for a year (n=1,016, 73.4% obese) [1].
The weight answer is yes, and it needs a caveat the authors supply themselves. Obese patients lost more absolute weight and more BMI at 3, 6, 9 and 12 months, but they also started heavier, and the paper says plainly that this may account for the difference. That is the same arithmetic caution that applies to every headline weight figure, including the pooled averages in what the trials actually found.
The proportional measure is harder to explain away. The share reaching at least 5% weight loss was 21.9% vs 12.0% at three months (p=0.0016) and 27.2% vs 15.5% at six months (p=0.0086), obese against non-obese. A percentage-of-body-weight threshold is not inflated by a higher starting point the way a kilogram count is, so that gap is a real signal rather than an artifact of where people began.
The glycemic answer is no, and it is the half worth carrying away. After adjustment, HbA1c reductions and the proportions reaching HbA1c below 7.0% or dropping by at least 0.5% or 1.0% were statistically comparable between the groups. Whatever these drugs do to blood sugar, they appear to do it without much regard to how heavy someone was when they started. That separation between the two effects also runs through the kidney evidence, where benefit does not track the scale either.
Two limits matter before any of this is carried anywhere. Patients were started on either an SGLT2 inhibitor or a GLP-1 receptor agonist, and the comparison is obese against non-obese rather than drug against drug — so nothing here is a statement about GLP-1 on its own. And the authors call their own findings hypothesis-generating and ask for prospective work, which is the appropriate weight to give a retrospective record review.
For a reader deciding whether to start, the useful version is narrow: a lower starting BMI is not a reason to expect nothing, and a higher one is not a promise. The variation between individuals still dwarfs the variation between these groups, which is the point made at length in weight loss in ordinary care and worth checking against the expected weight loss tool rather than against a single trial number.