📄 medicine

Risk of Diabetic Ketoacidosis with Glucagon-Like Peptide-1 Receptor Agonists: A Systematic Review and Meta-analysis

This systematic review and meta-analysis of 32 studies involving over 327,000 users indicates that while diabetic ketoacidosis is an uncommon complication of Glucagon-Like Peptide-1 Receptor Agonists overall (0.7%), the risk is significantly higher in patients with type 1 diabetes (2.0%) compared to those with type 2 diabetes (0.3%).

Original authors: Nasser Alkhushaym, Wadha Almutairi, Shahad Alanazi, Reem Balharith, Abdullah Alhifany, Imad R Musaa

Published 2026-07-08

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Nasser Alkhushaym, Wadha Almutairi, Shahad Alanazi, Reem Balharith, Abdullah Alhifany, Imad R Musaa

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A Safety Check on Popular Diabetes Meds

Imagine GLP-1 receptor agonists (drugs like Ozempic, Wegovy, Trulicity, and Byetta) as highly efficient traffic controllers for your body's sugar and appetite. They help lower blood sugar, help you lose weight, and keep your heart healthy. They are very popular right now.

However, doctors have been worried about a specific, rare, and dangerous traffic jam called Diabetic Ketoacidosis (DKA). Think of DKA as a "chemical fire" in the body. It happens when the body runs out of insulin, stops burning sugar for fuel, and starts burning fat instead. This produces toxic waste products (ketones) that make the blood acidic, like pouring vinegar into a swimming pool.

This paper is a massive safety inspection. The researchers gathered 32 different studies (involving over 327,000 people) to answer one question: How often does this "chemical fire" happen when people use these traffic controller drugs?

How They Did It: The Great Library Search

The researchers acted like detectives. They didn't just look at one study; they went to the world's biggest medical libraries (PubMed, Scopus, Cochrane) and looked at everything published up to June 2025.

The Filter: They started with 625 potential studies. They threw out the ones that were just reviews, didn't have real data, or were about the wrong drugs.
The Final Count: They kept 32 studies that had solid data on real people taking these drugs.
The Method: They used a statistical tool (a "random-effects model") which is like averaging the results of many different weather reports to get a true picture of the climate, rather than relying on just one day's forecast.

What They Found: The Numbers

The study broke the results down into two main groups, because the risk isn't the same for everyone.

1. The Overall Picture (Everyone Combined)

The Stat: Out of every 100 people taking these drugs, about 0.7 developed DKA.
The Analogy: This is like flipping a coin 140 times and only getting "heads" (the bad outcome) once. It is uncommon.

2. The Split: Type 1 vs. Type 2 Diabetes
The researchers realized that the "traffic controllers" affect two different types of drivers differently:

Type 1 Diabetes Drivers: These people's bodies make no insulin at all. They rely entirely on external insulin.
- The Risk: The DKA rate was 2.0%.
- The Analogy: If you have a car with no engine (Type 1) and you try to drive it with a new GPS system (the drug), you have to be very careful not to turn off the engine (insulin) too much. The risk here is higher (about 1 in 50 people).
Type 2 Diabetes Drivers: These people's bodies still make some insulin, but it doesn't work well.
- The Risk: The DKA rate was 0.3%.
- The Analogy: These cars still have an engine, just a sluggish one. The risk of the "chemical fire" is very low (about 3 in 1,000 people).

Why Does This Happen? (The "Why" Behind the "What")

The paper suggests a few reasons why the "chemical fire" might start, especially in Type 1 patients:

The "Too Little Fuel" Trap: These drugs make you feel full and stop you from eating. Sometimes, patients (or their doctors) might cut their insulin dose too quickly because they are eating less.
- Analogy: Imagine you are driving a car that needs gas (insulin). If you suddenly stop eating (fuel) but also turn down the gas supply (insulin) too much, the car sputters and catches fire.
The "Silent" Fire (Euglycemic DKA): Sometimes, the blood sugar doesn't look high, but the body is still burning fat and creating acid.
- Analogy: It's like a house fire where the smoke alarm (high blood sugar) doesn't go off, but the house is still burning inside. This makes it harder to spot.
Other Triggers: Getting sick, having surgery, or stopping insulin abruptly can also trigger this, especially when combined with these new drugs.

The Bottom Line

Is it dangerous? Yes, DKA is a serious medical emergency.
Is it common? No. For the vast majority of people, it is a rare event.
Who needs to be careful? People with Type 1 Diabetes who are using these drugs alongside insulin need to be extra vigilant. They must not cut their insulin doses too fast, even if they are eating less because of the drug.
Who is safer? People with Type 2 Diabetes have a very low risk, but they still need to monitor their health.

In short: These drugs are powerful tools, but like any powerful tool, they require a manual. The study confirms that while the "chemical fire" (DKA) is rare, it happens more often in Type 1 patients if the insulin "fuel" is reduced too quickly. The key is to keep the fuel line open and monitor the engine closely.

Technical Summary: Risk of Diabetic Ketoacidosis with Glucagon-Like Peptide-1 Receptor Agonists

Problem Statement
Glucagon-Like Peptide-1 Receptor Agonists (GLP-1 RAs) have become a cornerstone in managing type 2 diabetes mellitus (T2DM) and obesity due to their efficacy in glycemic control, weight reduction, and cardiovascular benefits. However, safety concerns have emerged regarding their potential to induce diabetic ketoacidosis (DKA), including euglycemic variants. While DKA is classically associated with absolute insulin deficiency in type 1 diabetes (T1DM), it can occur in T2DM under physiological stress. The specific risk magnitude of DKA associated with GLP-1 RA use across different patient populations remains a subject of investigation, necessitating a comprehensive synthesis of current evidence to guide clinical practice.

Methodology
This study is a systematic review and meta-analysis conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

Data Sources: A comprehensive search was performed across PubMed, Scopus, and the Cochrane Library, covering publications from inception through June 2025.
Search Strategy: Keywords included specific GLP-1 RAs (semaglutide, liraglutide, dulaglutide, exenatide, lixisenatide) and outcomes related to diabetic ketoacidosis (DKA, euglycemic ketoacidosis).
Selection Criteria: The review included human studies (randomized controlled trials and observational studies) involving adults (>18 years) treated with GLP-1 RAs that reported DKA outcomes. Non-English publications, editorials, letters without original data, and animal studies were excluded.
Data Extraction: Three independent reviewers extracted data on study design, country, specific GLP-1 RA used, total user numbers, and DKA incidence.
Statistical Analysis: Meta-analytic calculations were performed using Comprehensive Meta-Analysis (Version 4.0). A random-effects model was employed to account for heterogeneity across studies. Heterogeneity was quantified using Cochran's Q and the $I^2$ statistic, with significant heterogeneity defined as $P < 0.05$ for Q and $I^2 > 50\%$ . Subgroup analyses were conducted separately for patients with T1DM and T2DM.

Key Contributions and Results
The analysis screened 625 unique studies, ultimately including 32 studies comprising a total of 327,465 GLP-1 RA users.

Overall Incidence: The pooled meta-analysis estimated the frequency of DKA among all GLP-1 RA users to be 0.7% (95% CI: 0.4–1.4). Substantial heterogeneity was observed ( $I^2 = 98\%$ ).
Subgroup Analysis (T1DM): In the subgroup of 17 studies involving patients with T1DM, the estimated frequency of DKA was significantly higher at 2.0% (95% CI: 1.1–3.3; $I^2 = 53\%$ ).
Subgroup Analysis (T2DM): In the subgroup of 12 studies involving patients with T2DM, the estimated frequency was 0.3% (95% CI: 0.1–0.8; $I^2 = 99\%$ ).

Discussion of Mechanisms and Context
The authors discuss plausible mechanisms for the observed risk, particularly in T1DM. A primary driver appears to be the reduction of insulin doses following GLP-1 RA initiation, either intentionally to prevent hypoglycemia or secondary to reduced caloric intake due to nausea and appetite suppression. The paper notes that DKA events in T1DM trials (e.g., ADJUNCT ONE and ADJUNCT TWO) were often triggered by factors such as intercurrent infection, pump malfunction, or inadequate perioperative insulin therapy, alongside insulin dose reductions.

The interaction with other therapies is also highlighted. While combination therapy with SGLT2 inhibitors remains an area of investigation, the paper notes that randomized evidence in selected populations has been reassuring, suggesting risk depends more on patient selection and insulin management than the drug combination itself. Additionally, the paper acknowledges that reduced caloric intake can promote ketone production even without marked hyperglycemia, potentially leading to euglycemic DKA.

Significance and Limitations
The study concludes that while DKA among GLP-1 RA users is uncommon overall (0.7%), it is clinically more relevant in susceptible groups, specifically adults with T1DM receiving adjunct therapy and individuals undergoing rapid insulin de-escalation. The authors emphasize that careful patient selection and gradual insulin titration are warranted.

The paper modestly acknowledges several limitations:

Study Design: Most included studies were not primarily powered to detect differences in DKA risk, as DKA is a rare outcome.
Bias: Real-world observational studies are susceptible to residual confounding and indication bias.
Data Granularity: The analysis relied on aggregate study-level data rather than individual patient data, limiting the exploration of dose-response relationships.
Generalizability: The evidence base is dominated by specific regions and healthcare systems, which may limit applicability to under-represented populations.

The authors state that these limitations highlight the need for large, prospective, systematic studies to more precisely define the relationship between GLP-1 receptor agonists and DKA risk.