When glucose is not directly available to the body (from food or liver glycogen), body fat is broken down. This can occur during fasting, during or after increased physical exertion/exercise, and in cases of insulin deficiency in people with diabetes. In the latter case, blood glucose levels are usually elevated, often above 250 mg/dL (13.9 mmol/L). This condition is known as diabetic ketoacidosis (DKA). It results from the breakdown of fats into ketone bodies, which lower blood pH and cause electrolyte disturbances.
Users of insulin pumps are at particularly high risk—regardless of whether they are using a “classic” continuous subcutaneous insulin infusion (CSII) or automated insulin delivery (AID) systems. In all cases, because only rapid-acting insulin is used, there is a risk of ketoacidosis if insulin delivery is interrupted for any reason (e.g., infusion set dislodgement, occlusion of the infusion set, etc.).
Since patients with type 1 diabetes often have little or no endogenous insulin secretion, diabetic ketoacidosis can develop quite rapidly [1,2]. The fact that DKA does not occur more frequently in insulin pump users than in those undergoing intensified insulin injection therapy (ICT) [3] is presumably due to better patient education and, as a result, increased awareness of this issue.
In addition to diabetic ketoacidosis (DKA), euglycemic ketoacidosis (EKA) can also occur, which differs from classic DKA. In this case, there is no marked hyperglycemia, i.e., blood glucose levels usually remain well below 250 mg/dL (13.9 mmol/L). However, elevated levels of ketone bodies are still present in both blood and urine. EKA occurs predominantly in patients treated with SGLT-2 inhibitors. This is particularly noteworthy given the sharp increase in the use of these drugs (in 2024, SGLT-2 inhibitors were the second highest-grossing diabetes medications worldwide after GLP-1 receptor agonists [4]). Therefore, ketone self-monitoring is necessary not only for people with type 1 diabetes without relevant residual insulin secretion but also for people with type 2 diabetes treated with SGLT-2 inhibitors.
In ketoacidosis, the ketone bodies acetoacetate (acetoacetic acid) and β-hydroxybutyrate are formed. Both can be detected:
In the foreseeable future, continuous ketone monitoring (CKM) will also become available [5]. Through ketone self-monitoring, metabolic decompensation can be detected at an early stage.
If a continuous glucose monitoring (CGM) sensor is used, the development of ketoacidosis can often be detected at an early stage and possibly prevented proactively [6]. As of 2025, sensors that measure glucose and ketones simultaneously are not yet available. Systems are currently under development by companies such as Abbott, PercuSense, and Indigo Diabetes.
Based on experience, ketone self-monitoring is still rather neglected by many diabetes patients.
Ketone self-monitoring is required when signs that may be associated with ketoacidosis are detected:
For the detection of ketones in urine, combination tests (glucose and ketones) are available, such as the Keto-Diabur-Test 5000. The procedure is simple:
For the detection of ketones in blood, the procedure is the same as for blood glucose measurement, except that a ketone test strip is inserted into the device and β-hydroxybutyrate is measured. Typical devices such as the FreeStyle Precision NEO measure both blood glucose and ketones using two different types of test strips. As with blood glucose self-monitoring, various errors can affect ketone measurement, such as improperly stored test strips, expired test strips, touching the test fields with fingers, inadequate hand washing, insufficient blood volume on the test strip, etc. The interpretation of ketone self-monitoring results is shown in Table SK 2.
Urine test (e.g., Keto-Stix) | Blood measurement β-Hydroxybutyrate | Statement |
negative or + | <0,6mmol/l | Hyperglycemia |
++ | 0,6-1,5 mmol/l | mild ketoacidosis |
+++ | >1,5 mmol/l | severe ketoacidosis |
Table SK 2: Interpretation of the Results of a Ketone Test
[1] Reichel A, Rietzsch H, Köhler HJ, Pfützner A, Gudat U, Schulze J. Cessation of insulin infusion at night-time during CSII therapy: comparison of regular human insulin and insulin lispro. Experimental and Clinical Endocrinology & Diabetes. 1998;106(3):168–172. doi: 10.1055/s-0029-1211971.
[2] Zinman B. Insulin pump therapy and rapid-acting insulin: what have we learned? International Journal of Clinical Practice Supplement. 2001 Sep;(123):47–50.
[3] Karges B, Schwandt A, Heidtmann B, Kordonouri O, Binder E, Schierloh U, Boettcher C, Kapellen T, Rosenbauer J, Holl RW. Association of insulin pump therapy vs insulin injection therapy with severe hypoglycemia, ketoacidosis, and glycemic control among children, adolescents, and young adults with type 1 diabetes. JAMA. 2017 Oct 10;318(14):1358–1366.
[4] Kelly Close Report, January 21, 2025.
[5] Alva S, Castorino K, Cho H, Ou J. Feasibility of continuous ketone monitoring in subcutaneous tissue using a ketone sensor. Journal of Diabetes Science and Technology. 2021 Jul;15(4):768–774. doi: 10.1177/19322968211008185.
[6] Zhang JY, Shang T, Koliwad SK, Klonoff DC. Continuous ketone monitoring: a new paradigm for physiologic monitoring. Journal of Diabetes Science and Technology. 2021:19322968211009860.