A combination of semaglutide and a second injected hormone analogue produced about 6.6 percentage points more weight loss than semaglutide on its own, according to a meta-analysis of eight randomised trials in 6,898 people published on 15 September. The combination, CagriSema, is not something a reader in the United States can fill a prescription for: the FDA label database returns no product label for cagrilintide, its second ingredient, checked on 20 September 2026. The same review found CagriSema raised the rate of adverse events, and of gastrointestinal adverse events, against every comparator it was measured against.
What the review pooled
A team at Peking Union Medical College Hospital in Beijing searched Embase, MEDLINE, PubMed and the Cochrane Library to 14 June 2026 and found eight randomised controlled trials, together enrolling 6,898 participants, described as adults with overweight or obesity and type 2 diabetes. They ran separate random-effects pairwise comparisons rather than a network analysis, setting CagriSema against semaglutide alone, against cagrilintide alone, and against placebo.
On percentage body weight, CagriSema beat semaglutide by 6.60 points (95% CI -8.51 to -4.68), cagrilintide by 8.15 points (-11.40 to -4.89) and placebo by 14.30 points (-17.41 to -11.19). The confidence interval on that first number is worth reading rather than skipping: the pooled data are compatible with anything from about 4.7 to about 8.5 extra percentage points. The direction is not in doubt, because the interval does not reach zero. The size is much less settled than the headline figure suggests.
The glucose results are smaller than the weight results. CagriSema lowered HbA1c by 0.09 percentage points more than semaglutide (-0.14 to -0.04, P=0.0004), by 0.87 points more than cagrilintide and by 1.55 points more than placebo. Absolute weight, waist circumference and body mass index all moved further on CagriSema than on any comparator. C-reactive protein and lipid measures improved against cagrilintide and placebo; blood pressure control was better than cagrilintide and placebo and about the same as semaglutide.
Why the headline number needs a caveat
Heterogeneity in this review is high. The I-squared statistic – roughly, the share of the variation between trials that is more than chance would produce – was 82% for the weight comparison against semaglutide, 95% against cagrilintide and 99% against placebo. The authors attribute that to differences in baseline characteristics and background diabetes therapy across the eight trials. When trials disagree that much, a single pooled average describes none of them especially well, and the interval around it is the honest summary.
The HbA1c edge over semaglutide is a good example of a result that is statistically significant and clinically slight. Nine hundredths of a percentage point takes an HbA1c of 7.50% to 7.41%. That is a real difference in a dataset of nearly 7,000 people and not a difference a patient would notice.
For anyone comparing what is actually on the shelf, this changes nothing yet. Our reading of the current evidence on the drugs available now still puts tirzepatide at the top among approved options, and CagriSema is not among them.
What is not yet known
The authors name their own main limits: a small number of eligible trials, and no long-term cardiovascular endpoints for CagriSema at all. The stated population matters too. Weight loss on incretin drugs is consistently smaller in people who also have type 2 diabetes than in people who do not, so these figures should not be read across to a reader without diabetes. The underlying trials are sponsor-run. The review appeared in Diabetes, Metabolic Syndrome and Obesity, an open-access journal from Dove Medical Press. We read the published abstract and the reported results in full; we did not open the eight source trials, and the numbers above are the review’s.
Sources: Shao C and others, “Efficacy and Safety of CagriSema for Metabolic Outcomes: Systematic Review and Pairwise Meta-Analysis of Randomized Controlled Trials”, Diabetes, Metabolic Syndrome and Obesity, 15 September 2026 (PMID 42763732). Regulatory status checked against the openFDA drug label database, 20 September 2026.
