Hemorrhagic stroke (HS), including non-traumatic intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH), constitutes a substantial proportion of cerebrovascular incidents, accounting for around 30% of stroke cases. The triglyceride-glucose index (TyG-i) represents a precise insulin resistance (IR) indicator, a crucial metabolic disturbance. Existing literature has demonstrated an association between TyG-i and all-cause mortality (ACM) among individuals suffering from ischemic stroke (IS). Yet, the TyG-i prognostic implications for severe HS patients necessitating intensive care unit (ICU) admission are not clearly understood. Considering the notably elevated mortality and morbidity associated with HS relative to IS, investigating this association is warranted. Our primary aim was to investigate TyG-i and ACM association among critically ill HS patients within an ICU context. Herein, patients with severe HS were identified by accessing the Medical Information Mart for Intensive Care-IV (MIMIC-IV, version 2.2) database, using the International Classification of Diseases (ICD)-9/10 as diagnostic guidelines. Subsequently, we stratified the subjects into quartiles, relying on their TyG-i scores. Moreover, we measured mortality at ICU, in-hospital, 30 days, 90 days, and 1 year as the outcomes. Cox proportional hazards regression analysis and restricted cubic splines (RCS) were deployed for elucidating the relation between the TyG-i and ACM while utilizing the Kaplan-Meier (K-M) method to estimate survival curves. The findings’ robustness was assessed by conducting subgroup analysis and interaction tests employing likelihood ratio tests. The analysis included 1475 patients, with a male predominance of 54.4%. Observed mortality rates in the ICU, hospital, 30 days, 90 days, and 1 year were 7.3%, 10.9%, 13.8%, 19.7%, and 27.3%, respectively. Multivariate Cox regression analysis results manifested that heightened TyG-i was significantly related to ACM at 30 days (adjusted hazard ratio [aHR]: 1.32; 95% confidence interval [CI]: 1.05–1.67; P = 0.020), 90 days (aHR: 1.27; 95% CI: 1.04–1.55; P = 0.019), and 1 year (aHR: 1.22; 95% CI: 1.03–1.44; P = 0.023). The results of RCS analysis demonstrated a progressive elevation in ACM risk with rising TyG-i levels. Interaction tests found no significant effect modification in this relationship. In summary, TyG-i exhibits a significant correlation with ACM among patients enduring critical illness due to HS. This correlation underscores the probable utility of TyG-i as a prognostic tool for stratifying HS patients according to their risk of mortality. Applying TyG-i in clinical settings could enhance therapeutic decision-making and the management of disease trajectories. Additionally, this investigation augments existing research on the linkage between the TyG-i and IS, elucidating the TyG-i’s role in predicting mortality across diverse stroke categories.

中文翻译：

---

甘油三酯-葡萄糖指数：危重出血性脑卒中患者全因死亡率的新型评估工具——MIMIC-IV数据库的回顾性分析

出血性中风（HS），包括非外伤性脑出血（ICH）和蛛网膜下腔出血（SAH），在脑血管事件中占很大比例，约占中风病例的30%。甘油三酯-葡萄糖指数 (TyG-i) 代表精确的胰岛素抵抗 (IR) 指标，这是一种至关重要的代谢紊乱。现有文献已证明 TyG-i 与缺血性中风 (IS) 患者的全因死亡率 (ACM) 之间存在关联。然而，TyG-i 对需要入住重症监护病房 (ICU) 的严重热射病患者的预后影响尚不清楚。考虑到与 IS 相比，HS 相关的死亡率和发病率显着升高，因此有必要调查这种关联。我们的主要目的是调查 ICU 危重 HS 患者中 TyG-i 和 ACM 的关联。在此，使用国际疾病分类 (ICD)-9/10 作为诊断指南，通过访问重症监护医疗信息集市 IV (MIMIC-IV，版本 2.2) 数据库来识别严重热射病患者。随后，我们根据受试者的 TyG-i 分数将他们分为四分位数。此外，我们还测量了 ICU、住院期间、30 天、90 天和 1 年的死亡率作为结果。采用 Cox 比例风险回归分析和限制三次样条 (RCS) 来阐明 TyG-i 和 ACM 之间的关系，同时利用 Kaplan-Meier (KM) 方法估计生存曲线。通过使用似然比检验进行亚组分析和交互检验来评估研究结果的稳健性。该分析包括 1475 名患者，其中男性占 54.4%。ICU、医院、30天、90天和1年观察到的死亡率分别为7.3%、10.9%、13.8%、19.7%和27.3%。多变量 Cox 回归分析结果表明，30 天时 TyG-i 升高与 ACM 显着相关（调整后风险比 [aHR]：1.32；95% 置信区间 [CI]：1.05–1.67；P = 0.020）、90 天（aHR） ：1.27；95% CI：1.04–1.55；P = 0.019）和 1 年（aHR：1.22；95% CI：1.03–1.44；P = 0.023）。RCS 分析结果表明，随着 TyG-i 水平的升高，ACM 风险逐渐升高。交互作用测试发现这种关系没有显着的影响改变。总之，TyG-i 在因 HS 导致的危重疾病患者中表现出与 ACM 显着相关。这种相关性强调了 TyG-i 作为根据死亡风险对 HS 患者进行分层的预后工具的可能实用性。在临床环境中应用 TyG-i 可以增强治疗决策和疾病轨迹的管理。此外，这项研究增强了有关 TyG-i 和 IS 之间联系的现有研究，阐明了 TyG-i 在预测不同中风类别死亡率方面的作用。