Tomohiko Kimura

Background and aims: Predicting short‐term blood glucose changes after initiation of continuous intravenous insulin infusion (CII) in hyperglycemic crises remains challenging because glucose trajectories are influenced by both patient‐specific metabolic severity and dynamically adjusted treatment. This study aimed to develop and externally validate machine learning models for estimating retest glucose levels and glucose‐lowering responses after CII initiation. Methods: We analyzed 574 post‐CII blood glucose measurements from 191 patients with hyperglycemic crises across five Japanese institutions. An independent external validation cohort included 29 patients from another institution. Two complementary models were developed using gradient‐boosted decision tree algorithms: a direct retest glucose prediction model and a ΔGlu treatment‐response model. We evaluated model performance using internal validation, patient‐level fivefold cross‐validation, and external validation. ΔGlu was defined as the decrease in blood glucose between two consecutive reassessment points. Results: The eight‐feature direct retest glucose model showed stable predictive performance in internal validation (R ² = 0.853, MAE = 76.9 mg/dL, RMSE = 106.0 mg/dL), patient‐level fivefold cross‐validation (R ² = 0.803 ± 0.080), and external validation (R ² = 0.802, MAE = 78.0 mg/dL, RMSE = 101.0 mg/dL). Although the eight‐feature ΔGlu treatment‐response model showed moderate internal performance, its external generalizability for ΔGlu prediction was limited. Conclusions: The direct retest glucose model showed stable performance across internal and external validation, whereas ΔGlu itself was more difficult to generalize across institutions. These complementary models may support objective monitoring of short‐term glucose trajectories during CII for hyperglycemic crises, although further prospective validation is warranted.

PDX-1 and MafA are insulin gene transcription factors and maintain mature pancreatic β-cell function. After chronic exposure to hyperglycemia, however, the expression levels of PDX-1 and MafA are substantially suppressed by provoked oxidative stress and the activation of the JNK/c-Jun pathway, which further aggravates β-cell function. In addition, glucocorticoids are often used to treat a variety of diseases, but they often induce hyperglycemia, not only by increasing insulin resistance, but also by directly impairing insulin biosynthesis in β-cells. Glucocorticoids bind and activate the glucocorticoid receptor in β-cells, while PDX-1 and MafA expression levels are substantially suppressed through the activation of the JNK/c-Jun pathway, which could explain the molecular mechanism behind glucocorticoid-induced β-cell dysfunction. There are similarities in the molecular mechanism for β-cell dysfunction induced by chronic hyperglycemia and glucocorticoid administration. On the other hand, a GLP-1 receptor agonist rapidly enhances microvascular flow in pancreatic islets in a NO-dependent manner. Such phenomena show that the islet microenvironment is substantially regulated by GLP-1 receptor signaling. Also, based on these data, it is likely that rapid microvascular adaptation contributes to the maintenance of β-cell function under diabetic conditions.

This study aimed to investigate the anti-atherosclerotic properties of tirzepatide, a dual glucose-dependent insulinotropic polypeptide (GIP)/glucagon-like peptide-1 (GLP-1) receptor agonist, in comparison with semaglutide, a selective GLP-1 receptor agonist. ApoE knockout mice were divided into early diabetes (dosed from 10 to 22 weeks of age), late diabetes (dosed from 18 to 30 weeks of age), and non-diabetic groups after streptozotocin treatment, and each group received semaglutide, tirzepatide, or saline for 12 weeks. In the early diabetes group, both agents significantly suppressed aortic plaque formation compared with control, while modestly improving glycemia and lipid levels. No significant vascular effects were observed in late diabetes or non-diabetic groups. Tirzepatide markedly reduced inflammatory mediators, including Mcp-1 , Il-6 , I-cam , and Cd68 , whereas semaglutide showed partial overlap. Notably, these anti-inflammatory effects were also detected in non-diabetic mice, suggesting vascular protection may involve arterial actions independently of metabolic control. Taken together, our findings demonstrate that tirzepatide exerts anti-atherosclerotic effects comparable to semaglutide, supporting the concept that GIP and GLP-1 signaling can confer vascular benefits. These results highlight the potential clinical relevance of dual incretin receptor agonism for cardiovascular risk reduction, although further studies are required to clarify the specific role of GIP signaling.

An 81-year-old man was brought to our institution because of general fatigue, loss of appetite, and decreased consciousness. Based on the clinical findings, the patient was diagnosed with diabetic ketoacidosis (DKA) and started on insulin therapy. Two days after normalization of the ketone body level, he developed a high fever and abdominal pain. Imaging findings revealed a necrotic lesion in the intestine, and surgery was performed. Macroscopic findings showed a necrotic lesion and no thrombi. Finally, the patient was diagnosed with non-occlusive mesenteric ischemia (NOMI). We should be aware of the possibility of NOMI in subjects with DKA, especially those who have abdominal pain even after normalization of ketone body levels.

While preserving skeletal muscle mass (SMM) is advocated in type 2 diabetes mellitus (T2DM), mass-based assessments often overlook muscle quality. This study aimed to investigate whether the balance between cellular integrity (phase angle [PhA]) and metabolic load (body fat mass [BFM]) is more clinically relevant to glycemic control than SMM-based profiling in patients with T2DM. In this cross-sectional study of 622 patients, participants were stratified into four groups based on sex-specific medians of PhA and BFM, and similarly for SMM and BFM. Multivariate logistic regression evaluated associations with glycemic targets (HbA1c <8.0% and <7.0%), adjusting for confounders. Sensitivity analyses utilized continuous Z-scores and interaction terms. High-PhA/High-BFM and High-PhA/Low-BFM groups showed significantly higher odds of HbA1c <8.0% compared with the Low-PhA/High-BFM reference group (adjusted odds ratio [OR] 2.05, 95% confidence interval [CI] 1.22–3.44; OR 2.10, 95% CI 1.21–3.64, respectively). Conversely, SMM/BFM combinations showed no significant associations. Neither profile was significantly associated with the stricter HbA1c <7.0% target. In continuous models, the PhA × BFM interaction was significantly associated with achievement of HbA1c <8.0% (p = 0.025), whereas SMM × BFM interaction was not. In conclusion, the PhA/BFM interplay shows a clearer association with glycemic control than SMM/BFM profiling in T2DM. Higher PhA may reflect a more favorable metabolic phenotype against adiposity and is associated with a lower likelihood of severe hyperglycemia (HbA1c ≥8.0%). PhA/BFM profiling offers a novel approach to identify high-risk patients overlooked by conventional mass-based assessments.

Hypokalemia is a common and potentially life-threatening complication of continuous intravenous insulin infusion (CII) in patients with hyperglycemic crises. However, no simple quantitative indicator can estimate the risk of hypokalemia at treatment initiation. This multicenter retrospective cohort study enrolled patients hospitalized for hyperglycemic crises who received CII. Clinical data available at CII initiation were collected. Machine-learning techniques were primarily employed for feature selection and model comparison to develop a simple, clinically interpretable prediction model. A logistic regression–based indicator was constructed using the most contributory variables and validated internally and externally. Hypokalemia was defined as a serum potassium level <3.5 mmol/L. The model-building and external validation cohorts included 99 and 55 patients, respectively. Among multiple candidate models, a lightweight logistic regression model using only two variables, serum potassium level and insulin infusion rate per body weight, was selected for clinical applicability. In internal validation, the model demonstrated good discriminative performance (receiver operating characteristic-area under the curve [ROC-AUC] 0.821). When applied to the external validation cohort, the ROC-AUC was 0.655, and accuracy decreased slightly. A lower threshold may increase sensitivity in screening, whereas a higher threshold may improve specificity. In conclusion, this study presents a simple and clinically applicable indicator for predicting CII-related hypokalemia using only two routinely available variables at treatment initiation. This model supports individualized electrolyte monitoring during the acute management of hyperglycemic crises. Additionally, it may facilitate safer and more efficient clinical decision-making without reliance on complex algorithms.