Review of GLP-1 drugs added to insulin in type 1 diabetes finds modest A1c and weight effects

Adding a GLP-1 receptor agonist to insulin lowered HbA1c by an average of 0.23 percentage points and body weight by 3.93 kg in people with type 1 diabetes, according to an umbrella review published September 10, 2026, in Diabetes, Obesity and Metabolism ^[1][2]. The same analysis found high-certainty increases in nausea, vomiting and dropout due to side effects, and no significant improvement in time in range ^[1][2]. The authors concluded that their findings "do not support routine use" of these drugs in type 1 diabetes, but "suggest a potential role in selected individuals for whom weight reduction and lower insulin requirements are therapeutic priorities" ^[1][2].

The review, led by Okjoo Lee of Soonchunhyang University Hospital Bucheon in South Korea and Yejun Son of Kyung Hee University, with Dong Keon Yon of Kyung Hee University Medical Center as corresponding author, examined 18 systematic reviews with meta-analyzes covering 56 unique primary studies — 35 randomized controlled trials and 21 nonrandomized studies ^[1][2]. The authors searched PubMed/MEDLINE, Embase, the Cochrane Central Register of Controlled Trials and OpenAlex from their start through June 22, 2026 ^[1][2].

An umbrella review is a study of existing reviews rather than of individual patients. The authors calculated a corrected covered area of 19.6%, which they describe as a very high degree of overlap between the reviews — meaning the same trials were counted repeatedly across different papers, largely because of the ADJUNCT ONE and ADJUNCT TWO trials ^[1][2]. That overlap is why the team re-analyzed the underlying trial data themselves instead of simply averaging the published results, and why their certainty ratings vary by outcome ^[1][2].

The specific numbers

Results were reported for GLP-1 receptor agonists as a class, not for individual products. Alongside the HbA1c reduction of 0.23 percentage points (95% confidence interval, −0.30 to −0.17; moderate certainty) and weight reduction of 3.93 kg (−4.29 to −3.56; moderate certainty), the review found total daily insulin dose fell by 5.74 international units per day (−7.30 to −4.17; low certainty) ^[1][2]. Time in range rose by 1.99 percentage points, but the confidence interval crossed zero (−1.17 to 5.15), so the change was not statistically significant, and the certainty was rated very low ^[1][2].

On safety, the review found no statistically significant increase in severe hypoglycemia (risk ratio, 0.83; 95% CI, 0.36–1.91; low certainty) or diabetic ketoacidosis (risk ratio, 0.67; 95% CI, 0.16–2.86; low certainty) ^[1][2]. Nausea was about 2.9 times as likely with a GLP-1 drug (risk ratio, 2.88; 95% CI, 2.20–3.76; high certainty), vomiting about 3.1 times as likely (3.11; 1.94–4.97; high certainty), and withdrawal because of adverse events about 2.1 times as likely (2.10; 1.42–3.12; high certainty) ^[1][2]. Diarrhea was not significantly more common (1.88; 0.82–4.33; low certainty) ^[1][2].

The authors declared no conflicts of interest. Funding came from South Korea's Ministry of Health and Welfare, its Ministry of Science and ICT, and the Soonchunhyang University Research Fund ^[1][2].

Why it matters for patients

People with type 1 diabetes manage their condition with insulin, and this review does not describe any GLP-1 medication as approved for type 1 diabetes. Use of these drugs in type 1 diabetes would therefore be outside their approved labeling, and the review does not state how often doctors prescribe them that way ^[1][2].

The benefit shown here is real but small on the A1c side: a 0.23 percentage point average drop is well below the change usually needed to move most people into a different A1c category. The weight and insulin-dose changes are larger. But the review found no significant gain in time in range, the measure of how much of the day blood sugar stays in target, and rated that evidence very low in certainty ^[1][2]. At the same time, the chance of nausea, vomiting or stopping treatment because of side effects was roughly two to three times higher, and those findings carried the review's highest certainty rating ^[1][2].

Whether the drugs change the risk of severe low blood sugar or diabetic ketoacidosis remains unclear — the estimates pointed toward lower risk, but the ranges included the possibility of no difference or harm ^[1][2].

What happens next

The review does not announce new trials or a timeline for further research, and it does not report which GLP-1 products or doses were used in the 56 underlying studies ^[1][2]. The paper is listed as an Early View article, first published September 10, 2026, and a peer review history is available through the publisher ^[1].