Association between the triglyceride-glucose index and all-cause mortality in critically ill patients with traumatic brain injury: analysis of two large cohorts

doi:10.21203/rs.3.rs-5331299/v1

Download PDF

Research Article

Association between the triglyceride-glucose index and all-cause mortality in critically ill patients with traumatic brain injury: analysis of two large cohorts

https://doi.org/10.21203/rs.3.rs-5331299/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

This study aimed to assess the role of the triglyceride-glucose (TyG) index in predicting all-cause mortality in critically ill patients with traumatic brain injury (TBI).

Methods

This was a retrospective observational cohort study, and data were obtained from the Medical Information Mart for Intensive Care-IV (MIMIC-IV) and the eICU Collaborative Research Database (eICU-CRD). The participants were divided into three groups according to the TyG index tertiles. The primary outcome was hospital all-cause mortality. Multivariate logistic proportional regression analysis and restricted cubic spline regression were used to evaluate the association between the TyG index and hospital mortality in patients with TBI.

Results

A total of 865 critically ill patients with TBI were enrolled. The hospital mortality rate and intensive care unit (ICU) mortality rates were 13.4% and 8.4%, respectively. Multivariate logistic regression analysis showed that the TyG index was independently associated with an elevated risk of hospital mortality [OR 1.69 (95% CI 1.23–2.31); P = 0.001] and ICU mortality [OR 1.71 (95% CI 1.18–2.48); P = 0.004]. The restricted cubic spline regression model revealed that the risks of hospital and ICU mortality increased linearly with increasing TyG index.

Conclusion

The TyG index is significantly associated with hospital and ICU all-cause mortality rate in critically ill patients with TBI. This finding illustrates that the TyG index may serve as a valuable tool for identifying patients with TBI who are at an elevated risk of all-cause mortality.

Triglyceride-glucose index

Insulin resistance

Traumatic brain injury

all-cause mortality

Traumatic brain injury (TBI) has become the leading cause of disability and death disease globally [1]. Despite the advent of more efficacious treatments and improved management strategies in recent years, the risk of adverse clinical outcomes in patients with TBI remains considerable, particularly in those who are critically ill [2, 3]. This emphasizes the crucial requirement to ascertain risk factors associated with outcomes of TBI, thereby enabling the expedient implementation of preventive strategies.

Insulin resistance (IR) is a pathological condition in which peripheral tissues exhibit reduced sensitivity to insulin. This condition, commonly observed in critically ill patients, is characterized by elevated insulin levels and decreased insulin responsiveness. The triglyceride-glucose (TyG) index, which is calculated according to fasting blood glucose (FBG) and triglyceride (TG) levels, has become a convincing surrogate marker of IR [4, 5]. Several cross-sectional and retrospective studies have indicated that the TyG index is associated with a greater risk of all-cause mortality in critically ill patients with conditions such as stroke, sepsis, and atrial fibrillation [6–8]. The extensive release of inflammatory factors and the imbalance of oxidative stress likely contribute to insulin resistance following TBI. However, the associations between critically ill patients who have sustained TBI and poor pathophysiological conditions remain unknown. Therefore, investigating the relationship between insulin resistance severity and TBI patient outcomes is crucial.

This study aimed to evaluate the role of the TyG index in predicting all-cause mortality among critically ill TBI patients by analyzing data from the Medical Information Mart for Intensive Care-IV (MIMIC-IV) and the eICU Collaborative Research Database (eICU-CRD). Identifying high-risk patients through this analysis could inform the development of new strategies to improve clinical outcomes.

Study population

This retrospective cohort study investigated health-related data obtained from the MIMIC-IV database and the eICU-CRD [9, 10]. The MIMIC-IV database, managed by the MIT Computational Physiology Laboratory, is a comprehensive repository of medical records of over 190,000 patients treated in various ICUs at the Beth Israel Deaconess Medical Center between 2008 and 2019. The eICU-CRD database includes records from over 200,000 patients who received ICU care in more than 200 medical centers from 2014 to 2015. One author (Liang Wu) complied with the requirements for access to these two databases and was responsible for the data extraction (certification number: 59448949). Patients who were diagnosed with TBI were enrolled in this study according to the International Classification of Disease (ICD), 9th and 10th Revision. The exclusion criteria were as follows: (1) patients aged less than 18 years at the time of first admission; (2) patients with multiple admissions to the ICU for TBI, for whom only the data from the first admission were extracted; (3) patients with insufficient data (such as triglyceride and fasting blood glucose data); and (4) patients whose ICU length of stay was < 3 h (Fig. 1).

Variable extraction

This study used a combination of PostgreSQL (version 13.7.2) and Navicat Premium (version 16) software to extract data with Structured Query Language (SQL). Potential variables from four categories were identified: (1) demographic characteristics, including age, sex, weight, height, and body mass index (BMI); (2) laboratory indicators, including red blood cell (RBC), white blood cell (WBC), hemoglobin, platelets, serum sodium, serum creatinine, FBG, and TG; (3) comorbidities, including hypertension, heart failure, diabetes, renal disease, and chronic obstructive pulmonary disease (COPD); and (4) severity of illness scores at admission, such as the Acute Physiology and Chronic Health Evaluation (APACHE) score and Glasgow Coma Scale (GCS) score. Follow-up began on the date of admission and ended on the date of death. The value of the TyG index was obtained as ln[fasting TG (mg/dl) × FBG (mg/dl)]/2 [6, 7]. All the laboratory variables and disease severity scores were extracted from the data that were collected within the first 24 hours after the patients entered the ICU.

To prevent potential bias, variables were excluded if more than 20% of the data were missing. Variables with incomplete data comprising less than 20% of the total data were processed via multiple imputations with a random forest algorithm (trained on other non-missing variables). This was achieved through the application of the 'mice' package of R statistical software [11, 12].

Clinical outcomes

The primary endpoint of the present study was hospital all-cause mortality, and the secondary endpoint was mortality in the ICU.

Statistical analysis

Continuous variables are presented as the means ± SDs or medians (interquartile ranges) according to the data distribution, whereas categorical variables are expressed as proportions. The Kolmogorov‒Smirnov test was used to evaluate the normality of continuous parameters. The analysis of continuous variables was performed via a t test or ANOVA if they presented a normal distribution and via the Mann‒Whitney U test or Kruskal‒Wallis test if they presented a nonnormal distribution. Kaplan‒Meier survival analysis was performed to assess the incidence rate of endpoints among groups on the basis of different TyG index values, and their differences were assessed through log-rank tests. To evaluate influencing factors related to the risk of all-cause death, binary logistic regression analysis was performed. Cox proportional hazard models were used to calculate the hazard ratio (HR) and 95% confidence interval (CI) between the TyG index and endpoints and were adjusted for some models. The confounding variables included variables selected based on a p value < 0.05 in the univariate analysis. In addition, clinically relevant and prognosis-associated variables were also included in the multivariate model: Model 1: unadjusted; Model 2: adjusted for age, sex, and BMI; and Model 3: adjusted for age, sex, BMI, WBC, RBC, RDW, MCV, hemoglobin, INR, PT, albumin, anion gap, bicarbonate, chloride, potassium, AST, ALT, total bilirubin, creatinine, platelets, myocardial infarction, CHF, COPD, hypertension, diabetes, renal disease, atrial fibrillation, malignancy, and stroke. Furthermore, we analyzed the nonlinear associations of the baseline TyG index with ICU and hospital all-cause mortality using a restricted cubic spline regression model with four knots. The receiver operating characteristic (ROC) curves were analyzed to determine the cutoff value of the TyG index. The TyG index was entered into the models as a continuous variable or ordinal variable (the first tertile of the TyG index was taken as a reference group). The P values for trends were calculated at the tertile level. To identify the consistency of the prognostic value of the TyG index for primary outcomes, further stratified analyses were performed on the basis of sex, age (< 65 and ≥ 65 years), BMI (< 27.4 27.4–31.2, and ≥ 31.2 kg/m2), renal disease, atrial fibrillation, and stroke. The interactions between the TyG index and the variables used for stratification were examined with likelihood ratio tests. A two-sided P value < 0.05 was considered statistically significant. All the statistical analyses were performed with R software (version 4.0.2) and SPSS 22.0 (IBM SPSS Statistics, Armonk, NY, USA).

The present study enrolled a total of 865 critically ill patients with TBI. The median age of the patients included in the study was 60.5 (20.0) years, and 552 (63.8%) of the patients were male. The median TyG index for all participants was 9.2 (IQR: 8.6–9.7). The hospital mortality and ICU mortality rates were 13.4% and 8.4%, respectively (Table 1).

Table 1

Characteristics and outcomes of participants categorized by TyG index a
TyG index tertile	T1 (N = 289)	T2 (N = 288)	T3 (N = 288)	p-value
Demographic
Age	67.1 (19.0)	63.4 (19.8)	53.4 (18.7)	< 0.001
Gender (Male)	162 (56.1)	192 (66.7)	198 (68.8)	0.003
Height (cm)	169.2 (11.5)	171.3 (10.3)	172.4 (11.3)	0.002
Weight (kg)	74.1 (21.1)	77.6 (16.6)	88.5 (22.4)	< 0.001
BMI	25.7 (5.7)	26.3 (4.6)	29.9 (7.8)	< 0.001
APACHE-IV	51.0 (20.1)	54.6 (21.3)	64.3 (26.1)	< 0.001
GCS	13.0 (3.4)	12.2 (3.8)	10.0 (4.6)	< 0.001
Laboratory tests
WBC (K/uL)	11.5 (5.3)	14.3 (6.3)	18.0 (7.7)	< 0.001
RBC (m/uL)	4.2 (0.6)	4.3 (0.7)	4.3 (0.7)	0.367
RDW (%)	14.6 (1.8)	15.0 (2.2)	15.5 (2.4)	< 0.001
MCV (fL)	91.9 (6.7)	92.7 (7.0)	93.4 (6.6)	0.027
Hemoglobin (g/dL)	12.7 (1.9)	13.0 (2.1)	13.0 (2.2)	0.105
Platelets (K/uL)	241.1 (168.5–282.0)	290.0 (190.5–331.0)	362.0 (206.8-478.8)	< 0.001
Albumin (g/dL)	3.5 (0.6)	3.4 (0.6)	3.3 (0.7)	0.003
Anion gap (mEq/L)	12.1 (4.1)	13.0 (4.3)	14.3 (4.9)	< 0.001
Bicarbonate (mEq/L)	26.5 (3.2)	27.1 (4.0)	28.0 (4.5)	< 0.001
Calcium (mg/dL)	8.9 (0.6)	9.0 (0.7)	9.1 (0.9)	0.065
Chloride (mEq/L)	107.6 (6.5)	110.1 (7.4)	114.0 (9.7)	< 0.001
Creatinine (mg/dL)	1.0 (0.7–1.1)	1.3 (0.8–1.3)	1.6 (0.9–1.5)	< 0.001
Sodium (mEq/L)	141.6 (5.9)	143.5 (6.5)	147.4 (8.8)	< 0.001
Potassium (mEq/L)	4.2 (0.5)	4.4 (0.6)	4.6 (0.7)	< 0.001
INR	1.2 (1.0-1.2)	1.4 (1.0-1.3)	1.4 (1.1–1.4)	0.012
PT	14.3 (12.4–15.2)	15.2 (12.2–15.6)	16.2 12.9–17.4)	0.001
AST (U/L)	58.0 (20.0-49.5)	112.9 (22.0–81.0)	179.5 (30.0-119.5)	< 0.001
ALT (U/L)	39.6 (18.0–39.0)	82.0 (19.0–61.0)	124.7 (27.0-101.5)	< 0.001
Total Bilirubin (mg/dL)	1.0 (0.5-1.0)	1.1 (0.5–1.1)	1.7 (0.5–1.2)	0.007
Tyg	8.3 (8.2–8.6)	9.1 (8.9–9.2)	10.1 (9.6–10.4)	< 0.001
Comorbidities
Myocardial Infarction	25 (8.7)	15 (5.2)	11 (3.8)	0.040
CHF	20 (6.9)	21 (7.3)	12 (4.2)	0.232
COPD	19 (6.6)	17 (5.9)	15 (5.2)	0.785
Hypertension	138 (47.8)	135 (46.9)	111 (38.5)	0.049
Diabetes	23 (8)	15 (5.2)	31 (10.8)	0.048
Renal Disease	21 (7.3)	13 (4.5)	18 (6.2)	0.372
Atrial Fibrillation	27 (9.3)	27 (9.4)	16 (5.6)	0.154
Malignancy	29 (10)	26 (9)	15 (5.2)	0.081
Stroke	37 (12.8)	37 (12.8)	21 (7.3)	0.049
Outcomes
LOS ICU, days	4.4 (1.1–4.6)	6.8 (1.8–9.5)	11.9 (3.9–16.8)	< 0.001
LOS hospital, days	10.0 (3.7–11.3)	13.0 (4.7–17.8)	18.3 (6.9–25.6)	< 0.001
ICU mortality	10 (3.5)	17 (5.9)	46 (16)	< 0.001
Hospital mortality	20 (6.9)	33 (11.5)	63 (21.9)	< 0.001
^a TyG index: T1 (7.27–8.74), T2 (8.74–9.44), T3 (9.44–12.34)
Data: N (%) or Mean (Q1-Q3) or Mean ± Standard Deviation
APACHE: Acute Physiology and Chronic Health Evaluation-IV; GCS: Glasgow coma scale; BMI: Body Mass Index; WBC: White Blood Cell; MCV: Mean Cell Volume; RDW: Red Cell Distribution Width; INR: International Normalized Ratio; PT: Prothrombin Time; ALT: Alanine Aminotransferase; AST: Aspartate Aminotransferase; CHF: Congestive Heart Failure; COPD: Chronic Obstructive Pulmonary Disease; LOS ICU: Length of Stay Intensive Care Unit; LOS hospital: Length of Stay hospital.

Baseline characteristics

The baseline characteristics of critically ill patients with TBI divided according to TyG index tertiles are shown in Table 1. The enrolled individuals were grouped into three groups on the basis of hospital admission TyG index levels [tertiles (T)1 (7.27–8.74), T2 (8.74–9.44), and T3 (9.44–12.34)]. The median values of the TyG index of each tertile were 8.3 (IQR: 8.2–8.6), 9.1 (IQR: 8.9–9.2), and 10.1 (IQR: 9.6–10.4), respectively. Patients in the highest tertiles of the TyG index generally had high BMIs; higher APACHE-IV scores at admission; and higher incidences of myocardial infarction, hypertension and stroke. Moreover, patients with a higher TyG index exhibited elevated WBC, RDW, MCV, platelets, anion gap, bicarbonate, chloride, creatinine, sodium, PT, AST, and ALT levels. Compared with individuals in the lower tertiles of the TyG index, those in the higher tertiles had longer hospital stays (p < 0.001) and higher hospital mortality. Since the T3 subgroup had a higher rate of all-cause mortality, we further compared the differences between T3 and T1-2. The analysis revealed that different grouping approaches yielded similar results (Additional File 1, Table S1).

The differences in baseline characteristics between survivors and nonsurvivors during the hospital stay are shown in Table 2. Patients in the nonsurvivor group were more likely to be male and have a higher APACHE-IV score; a higher incidence of hypertension and renal disease; and higher WBC, RDW, MCV, chloride, creatinine, sodium, potassium, INR, PT, and total bilirubin values. The TyG index was significantly higher in the nonsurvivor group than in the survivor group (9.6 vs. 9.1, p < 0.001). Figure S1 shows the distribution of the TyG index stratified by the mortality status of all-cause hospital mortality and ICU mortality (Additional File 2, Figure S1).

Table 2

Baseline characteristics of the survivors and non-survivors groups
Categories	Overall (N = 865)	Survivor (N = 749)	Non-survivor (N = 116)	p-value
Demographic
Age	60.5 (20.0)	61.4 (20.2)	60.9 (18.9)	0.833
Gender (Male)	552 (63.8)	469 (62.6)	83 (71.6)	0.078
Height (cm)	170.9 (11.1)	170.6 (11.2)	173.1 (9.9)	0.022
Weight (kg)	80.0 (21.0)	79.4 (20.7)	84.5 (23.1)	0.015
BMI	27.3 (6.5)	27.1 (6.2)	28.4 (7.7)	0.103
APACHE-IV	57.0 (23.7)	53.9 (21.6)	74.2 (25.9)	< 0.001
GCS	11.7 (4.2)	12.1 (3.9)	9.4 (4.9)	< 0.001
Laboratory tests
WBC (K/uL)	14.6 (7.0)	14.1 (6.6)	17.8 (8.5)	< 0.001
RBC (m/uL)	4.2 (0.7)	4.3 (0.6)	4.1 (0.8)	0.012
RDW (%)	15.0 (2.1)	14.9 (2.0)	16.0 (2.6)	< 0.001
MCV (fL)	92.7 (6.8)	92.5 (6.7)	94.0 (7.4)	0.027
Hemoglobin (g/dL)	12.9 (2.1)	13.0 (2.0)	12.5 (2.5)	0.038
Platelets (K/uL)	300.6 (189.0-345.0)	304.9 (192.0-354.0)	250.6 (152.8-309.5)	< 0.001
Albumin (g/dL)	3.4 (0.6)	3.4 (0.6)	3.3 (0.7)	0.049
Anion gap (mEq/L)	13.1 (4.5)	12.9 (4.5)	14.7 (4.8)	< 0.001
Bicarbonate (mEq/L)	27.2 (4.0)	27.2 (3.9)	26.9 (4.3)	0.485
Calcium (mg/dL)	9.0 (0.7)	9.0 (0.7)	9.0 (0.9)	0.709
Chloride (mEq/L)	110.6 (8.4)	110.0 (7.7)	114.3 (11.6	< 0.001
Creatinine (mg/dL)	1.3 (0.8–1.3)	1.2 (0.8–1.1)	2.0 (1.3–1.9)	< 0.001
Sodium (mEq/L)	144.1 (7.6)	143.5 (7.0)	148.2 (9.4)	< 0.001
Potassium (mEq/L)	4.4 (0.6)	4.4 (0.6)	4.6 (0.8)	0.015
INR	1.3 (1.0-1.3)	1.3 (1.0-1.3)	1.6 (1.1–1.7)	< 0.001
PT	15.3 (12.5–15.9)	14.8 (12.4–15.5)	18.0 (12.9–19.9)	< 0.001
AST (U/L)	117.34 (23.0-78.5)	104.8 (22.0–72.0)	193.4 28.0-123.8)	0.059
ALT (U/L)	83.0 (20.0–63.0)	75.9 20.0–61.0)	121.9 (25.0-71.8)	0.147
Total Bilirubin (mg/dL)	1.3 (0.5–1.1)	1.0 (0.5–1.1)	2.7 (0.5–1.5)	0.009
Tyg	9.2 (8.6–9.7)	9.1 (8.5–9.6)	9.6 (8.9–10.1)	< 0.001
Comorbidities
Myocardial Infarction	51 (5.9)	40 (5.3)	11 (9.5)	0.121
CHF	53 (6.1)	42 (5.6)	11 (9.5)	0.158
COPD	51 (5.9)	45 (6)	6 (5.2)	0.886
Hypertension	384 (44.4)	321 (42.9)	63 (54.3)	0.027
Diabetes	69 (8.0)	54 (7.2)	15 (12.9)	0.053
Renal Disease	52 (6.0)	37 (4.9)	15 (12.9)	0.002
Atrial Fibrillation	70 (8.1)	59 (7.9)	11 (9.5)	0.684
Malignancy	70 (8.1)	62 (8.3)	8 (6.9)	0.745
Stroke	95 (11.0)	79 (10.5)	16 (13.8)	0.378
Data: N (%) or Mean (Q1-Q3) or Mean ± Standard Deviation
APACHE: Acute Physiology and Chronic Health Evaluation-IV; GCS: Glasgow coma scale; BMI: Body Mass Index; WBC: White Blood Cell; MCV: Mean Cell Volume; RDW: Red Cell Distribution Width; INR: International Normalized Ratio; PT: Prothrombin Time; ALT: Alanine Aminotransferase; AST: Aspartate Aminotransferase; CHF: Congestive Heart Failure; COPD: Chronic Obstructive Pulmonary Disease.

TyG index and hospital and ICU mortality

Multivariate logistic regression analysis revealed that the TyG index was independently associated with an increased risk of hospital mortality [OR 1.69 (95% CI 1.23–2.31); P = 0.001] and ICU mortality [OR 1.71 (95% CI 1.18–2.48); P = 0.004]. These results were further confirmed in the fully adjusted Model 3; specifically, the OR for hospital mortality in the highest TyG tertiles was 2.37 (95% CI 1.22–4.59), and that for ICU mortality was 4.25 (95% CI 1.82–9.92), both compared with the lowest tertile. Additionally, the risk of hospital mortality and ICU mortality demonstrated a consistent increasing trend with increasing TyG index tertiles, all with p values below 0.05 (Table 3). Moreover, the restricted cubic spline regression model revealed that the risk of hospital mortality and ICU mortality increased linearly with increasing TyG index (P for nonlinearity = 0.761 and P for nonlinearity = 0.05, respectively) (Fig. 2C and D).

Table 3

The association between TyG index a groups and Hospital and ICU mortality
Exposure	Model 1		Model 2		Model 3
Exposure	OR (95%CI)	P	OR (95%CI)	P	OR (95%CI)	P
Hospital mortality
TyG as continuous	1.89 (1.51–2.36)	< 0.001	2.06 (1.59–2.66)	< 0.001	1.69 (1.23–2.31)	0.001
T1	Ref		Ref		Ref
T2	1.74 (0.97–3.11)	0.062	1.73 (0.96–3.11)	0.066	1.31 (0.68–2.51)	0.425
T3	3.77 (2.21–6.42)	< 0.001	4.02 (2.27–7.11)	< 0.001	2.37 (1.22–4.59)	0.011
P for trend	< 0.001		< 0.001		0.011
ICU mortality
TyG as continuous	1.94 (1.50–2.51)	< 0.001	2.00 (1.49–2.68)	< 0.001	1.71 (1.18–2.48)	0.004
T1	Ref		Ref		Ref
T2	1.75 (0.79–3.89)	0.170	1.76 (0.79–3.93)	0.166	1.59 (0.66–3.82)	0.302
T3	5.30 (2.62–10.74)	< 0.001	5.56 (2.63–11.72)	< 0.001	4.25 (1.82–9.92)	< 0.001
P for trend	< 0.001		< 0.001		< 0.001
^a TyG index: T1 (7.27–8.74), T2 (8.74–9.44), T3 (9.44–12.34)
Model 1: Unadjusted
Model 2: Adjusted for Age, Gender, BMI
Model 3: Adjusted for Age, Gender, BMI, WBC, RBC, RDW, MCV, Hemoglobin, INR, PT, Albumin, Anion gap, Bicarbonate, Chloride, Potassium, AST, ALT, Total bilirubin, Creatinine, Platelets, Myocardial Infarction, CHF, COPD, Hypertension, Diabetes, Renal Disease, Atrial Fibrillation, Malignancy, Stroke

The Kaplan-Meier survival curves were generated to analyze the incidence of hospital mortality and ICU mortality among the groups, and the TyG index tertiles are presented in Figure S2 (Additional File 3, Figure S2). Patients with a higher TyG index had a greater risk of hospital and ICU mortality. However, there was no significant difference at 28 days or 3 months (log-rank P = 0.072 and 0.064, respectively). We evaluated the clinical efficacy of the TyG index via ROC analysis. However, the AUC of the TyG index was not good enough (hospital mortality AUC: 0.67 and ICU mortality AUC: 0.69). The cutoff values of the TyG index were 9.169 and 9.221 for hospital mortality and ICU mortality, respectively (Additional File 4, Figure S3). The associations between TyG index groups and mortality in patients with TBI according to Cox regression were stable (Additional File 5, Table S2).

Subgroup analysis

Furthermore, to confirm the relationship between the TyG index and in-hospital mortality, stratified analyses were conducted on the basis of sex, age, BMI, hypertension, and stroke (Figs. 3 and 4). The TyG index was significantly associated with a greater risk of hospital mortality in the TBI patient subgroups of females [OR: 3.76 (95% CI 1.68–8.39)], those aged ≥ 65 years [OR: 3.55 (95% CI 2.01–6.26)], those with BMIs ranging from 27.4–31.2 [OR: 2.80 (95% CI 1.13–6.91)] and > 31.2 [OR: 3.72 (95% CI 1.12–12.39)], those without stroke [OR: 2.04 (95% CI 1.32–3.17)], those without renal disease [OR: 2.03 (95% CI 1.32–3.13)], and those without atrial fibrillation [OR: 2.01 (95% CI 1.32–3.07)] (Fig. 3). Similarly, in terms of stratified analyses of ICU mortality, the TyG index demonstrated a significant association with high risk of ICU mortality in the female subgroup [OR: 3.76 (95% CI 1.68–8.39)], those aged ≥ 65 years [OR: 3.59 (95% CI 1.59–8.08)], those without stroke [OR: 2.11 (95% CI 1.13–3.93)], those without renal disease [OR: 1.97 (95% CI 1.09–3.55)], and those without atrial fibrillation [OR: 1.86 (95% CI 1.05–3.29)] (Fig. 4).

To the best of our knowledge, this study is the first to examine the relationship between the TyG index and clinical outcomes in TBI patients. Our findings suggest that a higher TyG index is associated with increased all-cause hospital and ICU mortality in critically ill TBI patients. Even after adjusting for confounding risk factors, the TyG index remained strongly associated with all-cause hospital and ICU mortality.

The TyG index, consisting of TG and FBG, has been suggested to be a potential indicator of metabolic disorders, atherosclerotic disease, cardiovascular disease, and hemorrhagic disease [13–16]. Numerous clinical studies have explored the associations between the TyG index and the incidence and mortality of cardiovascular diseases across various populations, including both public and specific patient groups. According to Yang et al., an elevated TyG index was associated with an increased risk of mortality in the hospital and ICU among patients who experienced cardiac arrest [17]. On the basis of the observations of Lee et al., the TyG index may help predict short-term functional outcomes in patients with acute ischemic stroke who have undergone reperfusion therapy [18]. Liu et al. revealed an association between the TyG index and mortality among surgical ICU patients [19]. In patients with sepsis, the TyG index could help in the early identification of insulin resistance, thereby improving risk evaluation and guiding further interventions [6]. Additionally, another study indicated that the TyG index might serve as a valuable predictor of future cardiovascular events in individuals with coronary artery disease [20]. These studies indicate that the TyG index holds promise for predicting the clinical outcomes of critically ill patients, patients with infection, and patients with cerebrovascular-related diseases.

The results of this study suggest an association between elevated TyG index levels and both the severity and clinical outcomes of TBI. TBI may lead to insulin resistance and impaired lipid metabolism, along with uncontrolled hyperglycemia and fluctuations in blood glucose during the acute phase [21–23]. The outcome of TBI is strongly correlated with the intensity of inflammatory responses, which are significantly related to insulin resistance. Our study revealed a positive association between the TyG index and the TBI.

The precise biological pathways that link the TyG index to TBI prognosis have not been fully elucidated. Research indicates that insulin resistance is associated with the TyG index. Insulin resistance may trigger the development of the dyslipidemia triad, which in turn facilitates endothelial dysfunction, the local release of chronic inflammatory mediators, endothelial injury, and eventually contributes to atherosclerosis development [24–26]. First, the TyG index may reflect the status of glucose metabolism, inflammation, and oxidative stress. In a state of insulin resistance, the production of nitric oxide (NO) stimulated by insulin is reduced. As a result, the inhibition of MAPK pathway activation due to hyperinsulinemia is impaired. Moveover, increased blood glucose levels combined with hyperinsulinemia further enhance the activation of the MAPK pathway, amplifying its downstream effects, such as heightened local inflammatory responses. Second, an increased TyG index may increase the level of free fatty acids, which may accompany insulin resistance. Chronic excessive calorie levels in the presence context of insulin resistance leads to increased visceral fat accumulation within the body [27]. Subsequently, adipocytes release chemokines, which induce monocytes to differentiate into macrophages that secrete large amounts of TNFα, reducing the storage of insulin-related substances. This leads to elevated circulating triglyceride levels and promotes an increase in free fatty acids [28]. All these pathophysiological changes can lead to poor clinical outcomes.

However, studies on the association between the TyG index and critically ill patients are rare. An elevated TyG index is significantly associated with an increased risk of mortality in critically ill patients with stroke and cardiovascular disease [6, 8, 29]. In our study, significant variations in APACHE scores were observed across patients grouped by their TyG index, suggesting a strong correlation between the TyG index and the severity of the disease. Furthermore, for TBI, we identified the TyG index as an important independent predictor of mortality in critically ill patients, contributing to clinical management aimed at reducing the risk of future adverse outcomes.

In addition, this study further analyzed the risk stratification of various subgroups. Our study revealed that the linear relationships between the TyG index and hospital mortality in TBI patients were consistent across older patients, female patients, those with higher BMIs, those without renal disease, those without atrial fibrillation, and those without stroke. This phenomenon could be explained by reverse causality: patients who are already diagnosed with these comorbidities are more likely to have received proper treatment or made healthy lifestyle changes. Interestingly, individuals with a higher TyG index tended to be younger, but the relationship between the TyG index and all-cause mortality was stronger in older patients, which is different from findings of previous studies. Consistent with common practice, clinicians tend to focus more on older TBI patients, as they are likely to have more comorbidities and face a higher risk of mortality. This study also revealed a clear linear association between the TyG index and hospital mortality, suggesting that the TyG index could serve as a practical tool for identifying critically ill TBI patients at high risk of mortality. Therefore, maintaining proper triglyceride and glucose levels, along with improved management of the TyG index, contributes to reducing the risk of future major adverse clinical events. In conclusion, our analysis suggests that the TyG index should not be viewed as the sole diagnostic measure but rather as a complementary tool alongside other clinical and laboratory parameters. This combined approach offers a more comprehensive evaluation of an individual's metabolic health and aids in the risk stratification of clinical outcomes, such as mortality in severe TBI, in clinical practice.

However, it is important to acknowledge the limitations of this study. First, our analysis was retrospective and conducted with observational data, and this study could not definitively establish causality. Although multivariate adjustment and subgroup analyses were used, residual confounding factors could still have influenced the clinical outcomes. Second, TBI itself could affect lipid metabolism and blood glucose fluctuations, and the TyG index may change significantly when a patient is in the hospital. Instead of relying solely on measurements that were performed at admission, future research should focus on monitoring this index dynamically throughout the treatment process. Finally, our study did not perform a hyperinsulinemia-euglycemic clamp test, so we cannot evaluate the association between the TyG index and the gold standard of insulin resistance assessment. Further research is needed to investigate the key mechanisms underlying insulin resistance in patients with TBI.

In summary, our findings demonstrated that an elevated TyG index is significantly associated with increased hospital all-cause mortality in patients with TBI. Monitoring the TyG index could be valuable in supporting clinical decision-making and improving disease management. Given its accessibility and strong reproducibility, the TyG index can be considered a dependable tool for use across all levels of the health care system.

TBI Traumatic brain injury

ICU Intensive care unit

TyG Triglyceride-glucose

IR Insulin resistance

MIMIC-IV Medical Information Mart for Intensive Care

eICU-CRD eICU Collaborative Research Database

CHF Congestive Heart Failure

COPD Chronic Obstructive Pulmonary Disease

TG Triglyceride

FBG Fasting blood glucose

SQL Structured Query Language

GCS Glasgow coma scale

APACHE Acute Physiology and Chronic Health Evaluation

BMI Body Mass Index

WBC White Blood Cell

MCV Mean Cell Volume

RDW Red Cell Distribution Width

INR International Normalized Ratio

PT Prothrombin Time

ALT Alanine Aminotransferase

AST Aspartate Aminotransferase

LOS Length of Stay

Acknowledgments

None.

Author contributions

Liang Wu and Weiming Liu: conceptualization, methodology; Yunfei Li and Yu Deng: software, data curation; Yunfei Li: writing-original draft preparation; Guoyi Gao: visualization, investigation; Liang Wu and Yunfei Li: writing-reviewing and editing.

Funding

Capital’s Funds for Health Improvement and Research (CFH2024-1-2042).

Data Availability

The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

The study was performed according to the guidelines of Helsinki Declaration. The use of the MIMIC-IV and eICU-CRD databases was approved by the review committee of Massachusetts Institute of Technology and Beth Israel Deaconess Medical Center. The data is publicly available (in the MIMIC-IV and eICU-CRD databases), therefore, the ethical approval statement and the requirement for informed consent were waived for this study.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interest.

Author details

¹Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, No. 119 Nansihuan Xilu, Fengtai District, Beijing, 100070, China. ²Beijing Key Laboratory of Central Nervous System Injury, Beijing Neurosurgical Institute, Capital Medical University

Maas AIR, Menon DK, Adelson PD, Andelic N, Bell MJ, Belli A, et al. Traumatic brain injury: integrated approaches to improve prevention, clinical care, and research. The Lancet Neurology. 2017;16:987–1048.
Maas AIR, Menon DK. Highlights in traumatic brain injury research in 2023. Lancet Neurol. 2024;23:15–7.
Taylor CA, Bell JM, Breiding MJ, Xu L. Traumatic Brain Injury-Related Emergency Department Visits, Hospitalizations, and Deaths - United States, 2007 and 2013. MMWR Surveill Summ. 2017;66:1–16.
Irace C, Carallo C, Scavelli FB, De Franceschi MS, Esposito T, Tripolino C, et al. Markers of insulin resistance and carotid atherosclerosis. A comparison of the homeostasis model assessment and triglyceride glucose index. Int J Clin Pract. 2013;67:665–72.
Guerrero-Romero F, Simental-Mendía LE, González-Ortiz M, Martínez-Abundis E, Ramos-Zavala MG, Hernández-González SO, et al. The product of triglycerides and glucose, a simple measure of insulin sensitivity. Comparison with the euglycemic-hyperinsulinemic clamp. J Clin Endocrinol Metab. 2010;95:3347–51.
Cai W, Xu J, Wu X, Chen Z, Zeng L, Song X, et al. Association between triglyceride-glucose index and all-cause mortality in critically ill patients with ischemic stroke: analysis of the MIMIC-IV database. Cardiovasc Diabetol. 2023;22:138.
Zheng R, Qian S, Shi Y, Lou C, Xu H, Pan J. Association between triglyceride-glucose index and in-hospital mortality in critically ill patients with sepsis: analysis of the MIMIC-IV database. Cardiovasc Diabetol. 2023;22:307.
Hu Y, Zhao Y, Zhang J, Li C. The association between triglyceride glucose-body mass index and all-cause mortality in critically ill patients with atrial fibrillation: a retrospective study from MIMIC-IV database. Cardiovasc Diabetol. 2024;23:64.
Pollard TJ, Johnson AEW, Raffa JD, Celi LA, Mark RG, Badawi O. The eICU Collaborative Research Database, a freely available multi-center database for critical care research. Sci Data. 2018;5:180178.
Johnson AEW, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10:1.
Austin PC, White IR, Lee DS, van Buuren S. Missing Data in Clinical Research: A Tutorial on Multiple Imputation. Can J Cardiol. 2021;37:1322–31.
Blazek K, van Zwieten A, Saglimbene V, Teixeira-Pinto A. A practical guide to multiple imputation of missing data in nephrology. Kidney Int. 2021;99:68–74.
Zhang R, Shi S, Chen W, Wang Y, Lin X, Zhao Y, et al. Independent effects of the triglyceride-glucose index on all-cause mortality in critically ill patients with coronary heart disease: analysis of the MIMIC-III database. Cardiovasc Diabetol. 2023;22:10.
Zhou Y, Pan Y, Yan H, Wang Y, Li Z, Zhao X, et al. Triglyceride Glucose Index and Prognosis of Patients With Ischemic Stroke. Front Neurol. 2020;11:456.
Zhao Q, Cheng Y-J, Xu Y-K, Zhao Z-W, Liu C, Sun T-N, et al. Comparison of various insulin resistance surrogates on prognostic prediction and stratification following percutaneous coronary intervention in patients with and without type 2 diabetes mellitus. Cardiovasc Diabetol. 2021;20:190.
Yang Y, Liang S, Liu J, Man M, Si Y, Jia D, et al. Triglyceride-glucose index as a potential predictor for in-hospital mortality in critically ill patients with intracerebral hemorrhage: a multicenter, case-control study. BMC Geriatr. 2024;24:385.
Boshen Y, Yuankang Z, Xinjie Z, Taixi L, Kaifan N, Zhixiang W, et al. Triglyceride-glucose index is associated with the occurrence and prognosis of cardiac arrest: a multicenter retrospective observational study. Cardiovasc Diabetol. 2023;22:190.
Lee M, Kim C-H, Kim Y, Jang MU, Mo HJ, Lee S-H, et al. High Triglyceride Glucose Index Is Associated with Poor Outcomes in Ischemic Stroke Patients after Reperfusion Therapy. Cerebrovasc Dis. 2021;50:691–9.
Liu D, Ren B, Tian Y, Chang Z, Zou T. Association of the TyG index with prognosis in surgical intensive care patients: data from the MIMIC-IV. Cardiovasc Diabetol. 2024;23:193.
Jin J-L, Cao Y-X, Wu L-G, You X-D, Guo Y-L, Wu N-Q, et al. Triglyceride glucose index for predicting cardiovascular outcomes in patients with coronary artery disease. J Thorac Dis. 2018;10:6137–46.
Sekar S, Viswas RS, Miranzadeh Mahabadi H, Alizadeh E, Fonge H, Taghibiglou C. Concussion/Mild Traumatic Brain Injury (TBI) Induces Brain Insulin Resistance: A Positron Emission Tomography (PET) Scanning Study. Int J Mol Sci. 2021;22:9005.
Korkmaz N, Kesikburun S, Atar MÖ, Sabuncu T. Insulin resistance and related factors in patients with moderate and severe traumatic brain injury. Ir J Med Sci. 2023;192:1177–82.
Mowery NT, Gunter OL, Guillamondegui O, Dossett LA, Dortch MJ, Morris JA, et al. Stress insulin resistance is a marker for mortality in traumatic brain injury. J Trauma. 2009;66:145–51; discussion 151-153.
Markus MRP, Rospleszcz S, Ittermann T, Baumeister SE, Schipf S, Siewert-Markus U, et al. Glucose and insulin levels are associated with arterial stiffness and concentric remodeling of the heart. Cardiovasc Diabetol. 2019;18:145.
Laakso M, Kuusisto J. Insulin resistance and hyperglycaemia in cardiovascular disease development. Nat Rev Endocrinol. 2014;10:293–302.
Ormazabal V, Nair S, Elfeky O, Aguayo C, Salomon C, Zuñiga FA. Association between insulin resistance and the development of cardiovascular disease. Cardiovasc Diabetol. 2018;17:122.
Fitzgibbons TP, Czech MP. Epicardial and perivascular adipose tissues and their influence on cardiovascular disease: basic mechanisms and clinical associations. J Am Heart Assoc. 2014;3:e000582.
Samuel VT, Shulman GI. The pathogenesis of insulin resistance: integrating signaling pathways and substrate flux. J Clin Invest. 2016;126:12–22.
Liao Y, Zhang R, Shi S, Zhao Y, He Y, Liao L, et al. Triglyceride-glucose index linked to all-cause mortality in critically ill patients: a cohort of 3026 patients. Cardiovasc Diabetol. 2022;21:128.

No competing interests reported.

AdditionalFile1TableS1.docx
AdditionalFile2FigureS1.jpg
Figure S1The distribution of the TyG index stratified by the mortality status of all-cause hospital death (A) and ICU death (B).
AdditionalFile3FigureS2.jpg
Figure S2Kaplan-Meier survival analysis curves for all-cause mortality. Footnote TyG index tertiles T1 (7.27-8.74), T2 (8.74-9.44), T3 (9.44-12.34). Kaplan-Meier curves show the cumulative probability of all-cause mortality according to groups at hospital mortality (A), 28 days (B), and 90 days (C).
AdditionalFile4FigureS3.jpg
Figure S3Receiver operating characteristic curve (ROC) of the TyG index for the hospital mortality and ICU mortality in TBI patients. ICU: intensive care unit; TyG: triglyceride-glucose, TBI, traumatic brain injury
AdditionalFile5TableS2.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Association between the triglyceride-glucose index and all-cause mortality in critically ill patients with traumatic brain injury: analysis of two large cohorts

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Materials

Study population

Variable extraction

Clinical outcomes

Statistical analysis

Results

Baseline characteristics

TyG index and hospital and ICU mortality

Subgroup analysis

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1