Association of estimated glucose disposal rate with all-cause and cardiovascular mortality in individuals with cardiovascular-kidney-metabolic syndrome: Two prospective cohorts

Highlights

• •Elevated estimated glucose disposal rate levels were found to be associated with a decreased likelihood of all-cause and cardiovascular mortality among adults with cardiovascular-kidney-metabolic syndrome.
• •The dose-response curve between estimated glucose disposal rate and all-cause and cardiovascular mortality risks showed a negative linear relationship in individuals with cardiovascular-kidney-metabolic syndrome.
• •Elevated estimated glucose disposal rate can serve as a promising predictor and therapeutic target for reducing mortality risks among cardiovascular-kidney-metabolic syndrome adults in clinical practice.

Abstract

Background

Given evidence on cardiovascular disease (CVD) risks conferred by comorbidity risk factors, the American Heart Association (AHA) introduced a novel framework, named cardiovascular-kidney-metabolic (CKM) syndrome. Accumulative evidence suggested that the estimated glucose disposal rate (eGDR) was significantly associated with mortality risk. However, it remains unknown whether this association exists in individuals with CKM syndrome. This study aimed to investigate whether eGDR can predict all-cause and cardiovascular mortality risks among adults with CKM syndrome.

Methods

This study used data from two prospective cohorts: the National Health and Nutrition Examination Survey (NHANES) and the China Health and Retirement Longitudinal Study (CHARLS). The exposure was eGDR at baseline, which was calculated using a combination of waist circumference, hypertension, and hemoglobin A1c. Cox regression models and restricted cubic splines were used to calculate the hazard ratio (HR) and 95 % confidence interval (95 % CI) after adjusting for potential confounders. The main outcomes were all-cause and cardiovascular mortality.

Results

A total of 34,809 participants (female: 50.8 %, mean age: 46.7 years) and 9,036 from CHARLS (female: 53.3 %, mean age: 59.6 years) were included. The median follow-up periods were 8.3 years in NHANES and 9.0 years in CHARLS. Per 1-SD increase in eGDR was associated with 18 %-24 % lower risks of all-cause mortality after adjusting for confounders (NHANES, HR 0.76, 95 % CI 0.71-0.82; CHARLS: HR 0.82, 95 % CI 0.77-0.88). In NHANES, the adjusted HR (95 % CI) for per 1-SD increase in eGDR was 0.60 (0.53–0.68) for cardiovascular mortality. The dose-response curve between eGDR and mortality risks showed a negative linear relationship in individuals with CKM syndrome.

Conclusion

A higher level of eGDR was associated with reduced risk of all-cause and cardiovascular mortality among adults with CKM syndrome. eGDR can serve as a promising predictor and therapeutic target for reducing mortality risks among CKM adults in clinical practice.

Graphical abstract

eGDR, estimated glucose disposal rate; CKM, cardiovascular-kidney-metabolic; NHANES, National Health and Nutrition Examination Survey; CHARLS, China Health and Retirement Longitudinal Study; WC, waist circumference; HTN, hypertension; HbA1c, hemoglobin A1c; HR, hazard ratio; CI, confidence interval.

1. Introduction

Epidemiological findings from observational research and clinical trials have underscored the significant overlap among cardiovascular disease (CVD), chronic kidney disease (CKD), and metabolic risk factors [[1], [2], [3], [4]]. In addition, individuals with CKD and/or diabetes exhibit worsened cardiovascular prognoses [5]. Recognizing this complex interplay, the AHA has recently issued an expert consensus on identifying and classifying individuals with cardiovascular-kidney-metabolic (CKM) syndrome in clinical practice [6]. This classification scheme is pivotal for enhancing multidisciplinary approaches to the prevention, risk stratification, and management of these conditions [7].

The prevalence of CKM syndrome in the worldwide is considerable. In 2021, the global prevalence of CVD, CKD, and diabetes was 612.1 million, 673.7 million and 525.7 million people, respectively, representing a surge increase since 1990 (111.4 % for CVD; 92.0 % for CKD; 277.9 % for diabetes) [8]. In the United States, the prevalence of CVD (coronary heart disease, heart failure, and stroke only), CKD, and metabolic syndrome is approximately 10 %, 15 %, and 35 % respectively [9]. Ahmad et al. [10] reported that heart disease (20.6 %), stroke (4.7 %), diabetes (3.0 %), and kidney disease (1.6 %) were the leading causes of death in the US, 2020. In China, the prevalence rates of CVD, CKD, and diabetes are 6.0 %, 11.2 %, and 10.8 %, respectively [11]. In addition, rural and urban cases of CVD accounted for 48.0 % and 45.9 % of mortality causes, respectively, with two of every five deaths attributed to CVD.

The estimated glucose disposal rate (eGDR), a composite indicator combining WC (waist circumference), hypertension, and hemoglobin A1c (HbA1c), correlates strongly with insulin resistance measured by the hyperinsulinemic-euglycemic clamp method [12,13]. Previous studies have consistently linked eGDR to mortality risks in individuals with and without diabetes [[14], [15], [16], [17]]. In a nationwide population based observational cohort study of 104,697 diabetes patients, Zabala et al. [14] reported that the hazard ratios (95 % confidence intervals) for all-cause mortality in individuals with eGDR categories between 4-6, 6-8 and > 8 were: 0.77 (0.69-0.87), 0.68 (0.58-0.80), and 0.60 (0.48-0.76), compared to the reference < 4. In a cohort study of 17,787 individuals, per-1 unit increment of eGDR was associated with 12 % and 10 % decrease risk of all-cause mortality in individuals with diabetes and without diabetes, respectively [16]. Nonetheless, the association between eGDR and mortality among adults with CKM syndrome remains to be elucidated.

To fill this knowledge gap, we utilized data from two large prospective cohorts in the current study, including the National Health and Nutrition Examination Survey (NHANES) and the China Health and Retirement Longitudinal Study (CHARLS). We aimed to examine the association between eGDR and the risk of all-cause and cardiovascular mortality among adults with CKM syndrome.

2. Methods

2.1. Study design and population

The NHANES [18] is a continuous, national survey that utilizes a complex, stratified, and multistage probability sampling approach targeting the noninstitutionalized US civilian population. This study incorporated data from 10 NHANES cycles, spanning from 1999–2000 to 2017–2018. The CHARLS [19] is an ongoing nationwide cohort investigation of persons in China 45 years of age or older and their spouses, employing a multistage clustering sampling technique to recruit participants from 28 provinces. A total of 17,705 individuals were initially enrolled during the baseline period of 2011-2012 (Wave 1), with subsequent biennial follow-ups in 2013 (Wave 2), 2015 (Wave 3), 2018 (Wave 4), and 2020 (Wave 5). The NHANES and CHARLS were approved by the National Center for Health Statistics Research Ethics Review Board and the Ethics Review Committees of Peking University, respectively. Informed consent was obtained from each participant in these two cohorts. This study complied with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline [20].

Fig. 1 illustrates the process of selecting the study population. The exclusion criteria were as follows: (1) participants under the age of 20 years in the NHANES or 45 years in the CHARLS; (2) participants with incomplete data required for the calculation of eGDR; (3) participants with missing data necessary for determining CKM syndrome stages; (4) participants lost to follow-up. Finally, a total of 34,809 participants from NHANES and 9,036 from CHARLS were included in our study.

Fig. 1.

Selection process of the study population. NHANES, National Health and Nutrition Examination Survey; CHARLS, China Health and Retirement Longitudinal Study; eGDR. estimate glucose disposal rate; CKM, cardiovascular-kidney-metabolic.

2.2. Assessment of eGDR

We computed eGDR using the following formula:[13]

$$\begin{array}{ccc}\hfill eGDR& =& 21.158-0.09\times WC\left(cm\right)-3.407\times hypertension\hfill \\ & & \left(yes=1/no=0\right)-0.551\times HbA1c(\%)\hfill \end{array}$$

Hypertension was defined based on self-reported physician-diagnosed hypertension, the use of antihypertensive drugs, systolic blood pressure (SBP) ≥ 140 mmHg, or diastolic blood pressure ≥ 90 mmHg [21].

2.3. Definitions of CKM syndrome stages

We delineated CKM syndrome stages according to the criteria from the 2023 AHA Presidential Advisory on CKM Health, with specific criteria outlined in N Methods [6]. The CKM syndrome is divided into five stages based on metabolic risk factors and the established disease: stage 0 (no CKM risk factors), stage 1 (excess or dysfunctional adiposity), stage 2 (metabolic risk factors or moderate-to-high-risk CKD [6]), stage 3 (subclinical CVD including very high-risk CKD or high predicted 10-y CVD risk[22]), and stage 4 (clinical CVD). We estimated the 10-year CVD risk using the AHA Predicting Risk of CVD Events (PREVENT) equations [22]. The PREVENT equations were initially developed for variables within specific ranges: age 30-79 years, total cholesterol 130-320 mg/dL, high-density lipoprotein cholesterol (HDL-C) 20-100 mg/dL, SBP 90-200 mmHg, BMI 18.5-<40 kg/m², and eGFR 15-150 mL/min/1.73m². Consequently, the 10-year CVD risk was not estimated for adults out of these values. However, to minimize the underestimation of CKM syndrome stage 3, participants with values exceeding the threshold were not excluded from the 10-year CVD risk. To approximate PREVENT risk strata, values for these variables above or below the specified bounds were adjusted to the upper or lower bounds of allowable values, respectively (for instance, age 90 years was set as 79 years). CKM stage 0 was designated as no CKM syndrome, and early CKM stages were defined as stages 1 or 2. Advanced CKM stages were defined as stages 3 or 4, as they identify individuals with or at high risk of CVD.

2.4. Definition of mortality

We merged baseline data from NHANES 1999–2018 with mortality data from the National Death Index death certificate records until December 31, 2019, using a probabilistic matching algorithm to determine mortality status. The outcomes included all-cause mortality and cardiovascular mortality. Cardiovascular mortality was defined as deaths resulting from heart or cerebrovascular disease (codes I00–I09, I11, I13, I20–I51, and I60–I69) using the International Classification of Diseases, Tenth Revision. In the CHARLS, mortality data were collected from life history surveys conducted in 2013, 2015, 2018, and 2020, with a follow-up period of approximately 9 years. Interviewers visited the residences of participants and, in the event of the participant’s death, collected relevant information by interviewing household members who lived with the deceased. As the reason for mortality was not recorded in the CHARLS, only NHANES contains cardiovascular mortality data.

2.5. Assessment of covariates

The covariates included age, gender, race and ethnicity (only in NHANES), residence (only in CHARLS), education, marital status, labor force status, smoking status, drinking status, BMI, history of arthritis, chronic lung disease, and cancer, high-density lipoprotein cholesterol (HDL-C), white blood cell (WBC) count, hemoglobin, platelets, serum creatinine (Scr), blood urea nitrogen (BUN), and uric acid (UA). Race and ethnicity were categorized into Mexican American, non-Hispanic Black, non-Hispanic White, other Hispanic, and other races. The residence was divided into rural and urban. Education was classified into three levels: less than high school, high school, and college or above. Marital status was divided into married or living with a partner and other marital status (including separated, divorced, widowed, or unmarried). Labor force status was divided into working and not working. Smoking status was categorized as ever and never smokers. Similarly, drinking status was categorized as ever and never drinkers.

2.6. Statistical analyses

For continuous variables, mean values and their standard deviations (SD) were reported. Categorical variables were characterized by counts and proportions. We compared the baseline characteristics of the three eGDR tertile groups using analysis of variance for continuous variables and the chi-square test for categorical ones. Missing rates for covariates were summarized in Supplementary Tables 1 and 2. Missing data of covariates were imputed using multiple imputations with chained equations [23]. We performed 5 imputations and generated 5 imputed datasets. We derived separate effect estimates for each dataset and pooled them according to Rubin's rules [23]. The multiple imputations were conducted using the R package "mice".

Kaplan–Meier survival curves were utilized to gauge the cumulative mortality risk across the three eGDR groups, compared using the log-rank test. We used Cox proportional hazard regression to examine the association between eGDR and the risk of all-cause and cardiovascular mortality, yielding the hazard ratio (HR) and its 95 % confidence interval (CI). Two models were fitted for the Cox regression. Model 1 was unadjusted. Model 2 was adjusted for age, gender, race and ethnicity (only in NHANES), residence (only in CHARLS), education, marital status, labor force status, smoking status, drinking status, BMI, history of arthritis, chronic lung disease, and cancer, HDL-C, WBC count, hemoglobin, platelets, Scr, BUN, and UA. The proportional hazards assumption for these Cox regression models was verified by Schoenfeld residuals [24]. The potential non-linear relationship between eGDR and mortality was explored using restricted cubic spline (RCS) regression with three knots at the 10th, 50th, and 90th percentiles. We conducted subgroup analyses to evaluate whether CKM syndrome stages, age, sex, smoking, and drinking status influenced the associations between eGDR and mortality risk, with interaction significance tested using likelihood ratio tests.

Several sensitivity analyses were performed: (i) to minimize the reverse causality, the main analyses were repeated after excluding participants who died within the initial two-year follow-up period; (ii) we repeated the main analyses after excluding participants with continuous values outside of the specified ranges set by the AHA PREVENT equations; (iii) given cancer's significant contribution to mortality, the main analyses were repeated after excluding participant with a history of cancer; (iv) we repeated the main analyses after further adjusting for the uses of antihypertensive, antidiabetic, and lipid-lowering drugs; (v) acknowledging the competing risk between non-cardiovascular and cardiovascular mortality, the competing risk analysis was conducted using the Fine-Gray model [25]; (vi) to compare the results of raw datasets and imputed datasets, we repeated the main analyses using data not imputed.

All statistical analyses were executed using R software (Version 4.3.2). Two-sided p-values were employed, with a threshold of <0.05 denoting statistical significance.

3. Results

3.1. Baseline characteristics of the study population

A total of 34,809 participants from NHANES (female: 50.8 %, mean age: 46.7 years) and 9,036 from CHARLS (female: 53.3 %, mean age: 59.6 years) were included according to the inclusion and exclusion criteria. Baseline characteristics of these participants stratified by eGDR are presented in Table 1, Table 2. In the NHANES and CHARLS, participants with higher tertiles of eGDR were younger, more likely to be working, and less likely to have a history of arthritis than participants with the lower tertile. The participant with higher tertiles of eGDR had lower levels of BMI, WBC, Scr, UA, and a higher level of HDL-C in contrast to their counterparts with the lower tertile. Additionally, the baseline characteristics of participants, utilizing non-imputed data, are detailed in Supplementary Tables 3 and 4. These results were similar to those presented in Table 1, Table 2. Baseline characteristics of these participants stratified by CKM syndrome stages are presented in Supplementary Tables 5 and 6. Participants with higher CKM stages were older, less likely to be married and working, and had lower education level than participants with no CKM. Participant with higher CKM stages also had higher levels of BMI, WBC, Scr, UA, and a lower level of eGDR and HDL-C.

Table 1.

Baseline characteristics of individuals classified by tertiles of eGDR in the NHANES.

Variable	eGDR tertiles (n = 34,809)			P value
	Tertile 1 (<6.30)	Tertile 2 (6.30-9.49)	Tertile 3 (≥9.49)
eGDR, mean (SD), mg/kg/min	4.5 (1.4)	8.1 (1.0)	10.6 (0.7)	<0.001
Age, mean (SD), years	56.3 (14.8)	47.7 (16.0)	38.3 (13.8)	<0.001
Age group, n (%)				<0.001
20-44 years	2,158 (22.4)	4,962 (44.7)	8,062 (69.4)
45-64 years	4,853 (45.7)	4,147 (38.8)	2,751 (25.3)
≥65 years	4,591 (31.9)	2,490 (16.5)	795 (5.3)
Gender, n (%)				<0.001
Male	6,249 (55.1)	5,873 (51.6)	4,996 (42.3)
Female	5,353 (44.9)	5,726 (48.4)	6,612 (57.7)
Race and ethnicity, n (%)				<0.001
Mexican American	1,846 (6.8)	2,253 (9.8)	1,963 (8.4)
Non-Hispanic Black	990 (4.6)	1,046 (5.4)	1,123 (6.6)
Non-Hispanic White	5,162 (70.3)	5,128 (69.1)	4,953 (67.2)
Other Hispanic	2,832 (12.7)	2,110 (9.5)	1,924 (8.9)
Other races	772 (5.6)	1,062 (6.2)	1,645 (8.9)
Education, n (%)				<0.001
Below high school	3,441 (18.9)	3,035 (17.0)	2,470 (14.0)
High school	2,836 (26.1)	2,780 (25.1)	2,369 (19.8)
College or above	5,325 (55.1)	5,784 (57.9)	6,769 (66.2)
Marital status, n (%)				<0.001
Married/Living with a partner	7,113 (66.4)	7,375 (67.3)	6,875 (61.5)
Other marital status	4,489 (33.6)	4,224 (32.7)	4,733 (38.5)
Labor force status, n (%)				<0.001
Working	5,046 (52.9)	6,844 (66.6)	8,170 (75.0)
Not working	6,556 (47.1)	4,755 (33.4)	3,438 (25.0)
Smoking status, n (%)				<0.001
Ever smokers	5,820 (51.1)	5,380 (47.2)	4,553 (41.0)
Never smokers	5,782 (48.9)	6,219 (52.8)	7,055 (59.0)
Drinking status, n (%)				<0.001
Ever drinkers	8,354 (76.5)	8,619 (79.4)	8,810 (81.1)
Never drinkers	3,248 (23.5)	2,980 (20.6)	2,798 (18.9)
BMI, mean (SD), kg/m2	33.4 (6.9)	29.8 (5.4)	24.1 (3.2)	<0.001
Arthritis, n (%)	4,958 (41.7)	2,899 (24.5)	1,268 (11.4)	<0.001
Chronic lung disease, n (%)	1,235 (10.7)	842 (7.6)	451 (4.2)	<0.001
Cancer, n (%)	1,525 (14.2)	982 (8.8)	533 (5.5)	<0.001
HDL-C, mean (SD), mg/dL	48.8 (14.5)	51.9 (16.2)	58.3 (16.6)	<0.001
WBC, mean (SD), 10^9/L	7.6 (3.3)	7.4 (2.3)	6.9 (2.0)	<0.001
Hemoglobin, mean (SD), g/dL	14.4 (1.5)	14.4 (1.5)	14.2 (1.4)	<0.001
Platelets, mean (SD), 10^9/L	248.0 (68.7)	252.2 (63.9)	248.7 (60.5)	<0.001
Scr, mean (SD), mg/dL	0.93 (0.40)	0.87 (0.28)	0.82 (0.20)	<0.001
BUN, mean (SD), mg/dl	15.1 (6.3)	13.3 (4.8)	12.3 (4.1)	<0.001
UA, mg/dL	6.0 (1.4)	5.5 (1.3)	4.9 (1.2)	<0.001
CKM syndrome stages				<0.001
No CKM syndrome	0 (0.0)	20 (0.3)	3,887 (37.1)
Early CKM stages	7,311 (70.6)	9,934 (89.4)	7,344 (60.6)
Advanced CKM stages	4,291 (29.4)	1,645 (10.3)	377 (2.2)

Categorical variables are presented as unweighted numbers (weighted percentages) and continuous variables as means ± SDs.

eGDR, estimated glucose disposal rate; NHANES, National Health and Nutrition Examination Survey; SD, standard deviation; BMI, body mass index; HDL-C, high-density lipoprotein cholesterol; WBC, white blood cell; BUN, blood urea nitrogen; UA, uric acid; CKM, cardiovascular-kidney-metabolic.

Table 2.

Baseline characteristics of individuals classified by tertiles of eGDR in the CHARLS.

Variable	eGDR tertiles (n = 9,036)			P value
	Tertile 1 (<7.78)	Tertile 2 (7.78-10.73)	Tertile 3 (≥10.73)
eGDR, mean (SD), mg/kg/min	6.5 (0.9)	9.5 (0.9)	11.5 (0.5)	<0.001
Age, mean (SD), years	61.3 (9.3)	59.2 (9.4)	58.2 (9.1)	<0.001
Age group, n (%)				<0.001
45-64 years	1,983 (65.8)	2,196 (73.1)	2,327 (77.1)
≥65 years	1,029 (34.2)	810 (26.9)	691 (22.9)
Gender, n (%)				<0.001
Female	1,317 (43.7)	1,378 (45.8)	1,525 (50.5)
Male	1,695 (56.3)	1,628 (54.2)	1,493 (49.5)
Residence, n (%)				<0.001
Rural	1,812 (60.2)	1,957 (65.1)	2,179 (72.2)
Urban	1,200 (39.8)	1,049 (34.9)	839 (27.8)
Education, n (%)				0.319
Below high school	2,719 (90.3)	2,716 (90.4)	2,732 (90.5)
High school	244 (8.1)	259 (8.6)	248 (8.2)
College or above	49 (1.6)	31 (1.0)	38 (1.3)
Marital status, n (%)				<0.001
Married/Living with a partner	2,575 (85.5)	2,636 (87.7)	2,704 (89.6)
Other marital status	437 (14.5)	370 (12.3)	314 (10.4)
Labor force status, n (%)				<0.001
Working	1,538 (51.1)	1,776 (59.1)	2,125 (70.4)
Not working	1,474 (48.9)	1,230 (40.9)	893 (29.6)
Smoking status, n (%)				<0.001
Ever smokers	1,093 (36.3)	1,161 (38.6)	1,306 (43.3)
Never smokers	1,919 (63.7)	1,845 (61.4)	1,712 (56.7)
Drinking status, n (%)				0.239
Ever drinkers	1,145 (38.0)	1,168 (38.9)	1,211 (40.1)
Never drinkers	1,867 (62.0)	1,838 (61.1)	1,807 (59.9)
BMI, mean (SD), kg/m2	25.6 (3.8)	23.9 (3.7)	21.0 (2.5)	<0.001
Arthritis, n (%)	1,138 (37.8)	1,081 (36.0)	1,036 (34.3)	0.020
Chronic lung disease, n (%)	312 (10.4)	305 (10.1)	343 (11.4)	0.260
Cancer, n (%)	30 (1.0)	31 (1.0)	30 (1.0)	0.987
HDL-C, mean (SD), mg/dL	47.5 (14.2)	50.6 (15.2)	55.6 (15.5)	<0.001
WBC, mean (SD), 10^9/L	6.5 (2.7)	6.2 (1.8)	6.1 (1.9)	<0.001
Hemoglobin, mean (SD), g/dL	14.7 (2.3)	14.4 (2.3)	14.0 (2.1)	<0.001
Platelets, mean (SD), 10^9/L	214.6 (73.9)	211.8 (81.3)	209.6 (72.5)	0.017
Scr, mean (SD), mg/dL	0.80 (0.29)	0.78 (0.21)	0.77 (0.20)	<0.001
BUN, mean (SD), mg/dl	15.8 (4.6)	15.7 (4.5)	15.7 (4.6)	0.968
UA, mean (SD), mg/dL	4.7 (1.3)	4.5 (1.3)	4.2 (1.2)	<0.001
CKM syndrome stages				<0.001
No CKM syndrome	0 (0.0)	55 (1.8)	913 (30.3)
Early CKM stages	1,866 (62.0)	2,357 (78.4)	1,717 (56.9)
Advanced CKM stages	1,146 (38.0)	594 (19.8)	388 (12.9)

Categorical variables are presented as numbers (percentages) and continuous variables as means ± SDs.

eGDR, estimated glucose disposal rate; CHARLS, China Health and Retirement Longitudinal Study; SD, standard deviation; BMI, body mass index; HDL-C, high-density lipoprotein cholesterol; WBC, white blood cell; BUN, blood urea nitrogen; UA, uric acid; CKM, cardiovascular-kidney-metabolic.

3.2. Association between eGDR and mortality

Over a median follow-up duration of 8.3 years, 4,020 (11.5 %) all-cause deaths and 1,587 (4.6 %) cardiovascular deaths occurred in the NHANES. In the CHARLS, there were 1,175 all-cause deaths (13.0 %) with a median follow-up of 9.0 years. Kaplan-Meier survival analysis revealed significant disparities in all-cause and cardiovascular mortality rates across the three eGDR tertiles (Supplementary Fig.s 1 and 2). Participants with a higher eGDR level had significantly lower cumulative incidence rates of all-cause and cardiovascular mortality (P value < 0.001 for all log-rank tests).

Table 3 shows the correlation between eGDR and all-cause mortality risks. A 1-SD increment in eGDR corresponded to 18 %-24 % decreased risks of all-cause mortality after adjusting for confounders (NHANES, HR 0.76, 95 % CI 0.71-0.82; CHARLS: HR 0.82, 95 % CI 0.77-0.88). When categorizing eGDR into tertiles, the upper tertile of eGDR was associated with significantly lower risks of all-cause mortality compared to the lower tertile (NHANES, HR 0.71, 95 % CI 0.60–0.83; CHARLS, HR 0.68, 95 % CI 0.58–0.80). For those with the middle tertile of eGDR, a significant reduction in all-cause mortality risks was evident only in the CHARLS (HR 0.78, 95 % CI 0.68–0.90), not in the NHANES (HR 0.93, 95 % CI 0.85–1.02). Moreover, multivariable-adjusted RCS analyses indicated negative linear associations between eGDR with all-cause mortality in both NHANES (P for overall < 0.001, P for nonlinear = 0.283) and CAHRLS (P for overall < 0.001, P for nonlinear = 0.251) (Fig. 2). The distributions of eGDR are described in Supplementary Fig. 3.

Table 4.

Association of the eGDR with cardiovascular mortality in the NHANES.

Variable	HR1 (95 % CI)a	P1a	HR2 (95 % CI)b	P2b
NHANES, n = 34,809
eGDR (per 1-SD)	0.43 (0.41-0.46)	<0.001	0.60 (0.53-0.68)	<0.001
eGDR tertile
Tertile 1	1 (reference)		1 (reference)
Tertile 2	0.40 (0.36-0.45)	<0.001	0.81 (0.71-0.93)	0.002
Tertile 3	0.09 (0.07-0.11)	<0.001	0.50 (0.39-0.66)	<0.001

In the NHANES, tertile 1 of eGDR < 6.30, 6.30 ≤ tertile 2 of eGDR < 9.49, tertile 3 of eGDR ≥ 9.49.

^aHR1 and P1 were unadjusted.

^bHR2 and P2 were adjusted for age, gender, race and ethnicity, education, marital status, labor force status, smoking status, drinking status, body mass index, history of arthritis, chronic lung disease, and cancer, high-density lipoprotein cholesterol, white blood cell count, hemoglobin, platelets, serum creatinine, blood urea nitrogen, and uric acid. NHANES, National Health and Nutrition Examination Survey; eGDR, estimated glucose disposal rate; SD, standard deviation; HR, hazard ratio; CI, confidence interval.

Table 3.

Association of the eGDR with all-cause mortality in the NHANES and CHARLS.

Variable	HR1 (95 % CI)a	P1a	HR2 (95 % CI)b	P2b
NHANES, n = 34,809
eGDR (per 1-SD)	0.52 (0.50-0.54)	<0.001	0.76 (0.71-0.82)	<0.001
eGDR tertile
Tertile 1	1 (reference)		1 (reference)
Tertile 2	0.52 (0.47-0.57)	<0.001	0.93 (0.85-1.02)	0.133
Tertile 3	0.17 (0.15-0.20)	<0.001	0.71 (0.60-0.83)	<0.001
CHARLS, n = 9,036
eGDR (per 1-SD)	0.83 (0.78-0.88)	<0.001	0.82 (0.77-0.88)	<0.001
eGDR tertile
Tertile 1	1 (reference)		1 (reference)
Tertile 2	0.75 (0.66-0.86)	<0.001	0.78 (0.68-0.90)	<0.001
Tertile 3	0.69 (0.60-0.79)	<0.001	0.68 (0.58-0.80)	<0.001

In the NHANES, tertile 1 of eGDR < 6.30, 6.30 ≤ tertile 2 of eGDR < 9.49, tertile 3 of eGDR ≥ 9.49. In the CHARLS, tertile 1 of eGDR < 7.78, 7.78 ≤ tertile 2 of eGDR < 10.73, tertile 3 of eGDR ≥ 10.73.

^aHR1 and P1 were unadjusted.

^bHR2 and P2 were adjusted for age, gender, race and ethnicity (only in NHANES), residence (only in CHARLS), education, marital status, labor force status, smoking status, drinking status, body mass index, history of arthritis, chronic lung disease, and cancer, high-density lipoprotein cholesterol, white blood cell count, hemoglobin, platelets, serum creatinine, blood urea nitrogen, and uric acid. NHANES, National Health and Nutrition Examination Survey; CHARLS, China Health and Retirement Longitudinal Study; eGDR, estimated glucose disposal rate; SD, standard deviation; HR, hazard ratio; CI, confidence interval.

Fig. 2.

Dose-response relationship between eGDR and all-cause mortality in the NHANES and CHARLS. The horizontal dotted line represents the HR of 1.0, where the eGDR values were 8.18 mg/Kg/min in the NHANES and 9.80 in the CHARLS. Hazard ratios were adjusted for the same variables as multivariable-adjusted models in Table 3. The lowest 1 % and highest 1 % of eGDR values were not shown in the Fig.s for small sample sizes. NHANES, National Health and Nutrition Examination Survey; CHARLS, China Health and Retirement Longitudinal Study; eGDR. estimate glucose disposal rate; HR, hazard ratio; CI, confidence interval.

Table 4 presents the association between eGDR and cardiovascular mortality in the NHANES. A 1-SD rise in eGDR was linked to a 40 % decrease in cardiovascular mortality risk after adjusting for confounders (HR 0.60, 95 % CI 0.53-0.68). Compared with participants with the lower eGDR tertile, those with middle or upper tertiles had 19 % (HR 0.81, 95 % CI 0.71–0.93) and 50 % (HR 0.50, 95 % CI 0.39–0.66) lower risks of cardiovascular mortality, respectively. In addition, multivariable-adjusted RCS analyses showed a negative linear association between eGDR with cardiovascular mortality (P for overall < 0.001, P for nonlinear = 0.982) (Fig. 3).

Fig. 3.

Dose-response relationship between eGDR and cardiovascular mortality in the NHANES. The horizontal dotted line represents the HR of 1.0, where the eGDR value was 8.18 mg/Kg/min in the NHANES. Hazard ratios were adjusted for the same variables as multivariable-adjusted model in Table 4. The lowest 1 % and highest 1 % of eGDR values were not shown in the Fig.s for small sample sizes. NHANES, National Health and Nutrition Examination Survey; CHARLS, China Health and Retirement Longitudinal Study; eGDR. estimate glucose disposal rate; HR, hazard ratio; CI, confidence interval. NHANES, National Health and Nutrition Examination Survey; eGDR. estimate glucose disposal rate; HR, hazard ratio; CI, confidence interval.

In the subgroup analysis, the associations between eGDR and all-cause mortality remained significant among participants with early (NHANES: HR 0.78, 95 % CI 0.70-0.86; CHARLS: HR 0.79, 95 % CI 0.71-0.88) and advanced CKM stages (NHAENS: HR 0.88, 95 % CI 0.79-0.98; CHALRS: HR 0.91, 95 % CI 0.84-0.99), but not in participants with no CKM (NHANES: HR 0.69, 95 % CI 0.16-2.98; CHARLS: HR 0.53, 95 % CI 0.19-1.44) (Supplementary Tables 7 and 8). Similar results were observed for cardiovascular mortality in the NHANES (Supplementary Table 9). Furthermore, no significant interactions between eGDR and CKM stages on the risk of mortality were observed (P for interaction >0.05). Meanwhile, the associations between eGDR and mortality risks were also significant across different age groups, sex, smoking status, and drinking status (Supplementary Tables 7-9). In individuals with/without abdominal obesity, as well as individuals with/without elevated HbA1c and with/without hypertension, the significant effects of eGDR on all-cause and cardiovascular mortality were discovered. The results showed interaction effects among participants with different age groups (P for interaction <0.05), with no interactions observed in all other subgroups (P for interaction >0.05).

3.3. Sensitivity analyses

Several sensitivity analyses were performed to evaluate the stability of the findings. After excluding participants who died during the initial two-year follow-up period, the correlation between eGDR and mortality persisted (Supplementary Tables 10 and 11). Participants with elevated eGDR levels had reduced risks of all-cause and cardiovascular mortality than those with lower levels. The associations of eGDR with all-cause and cardiovascular mortality were also consistent after excluding participants with continuous values outside of the ranges set by the AHA PREVENT equations (Supplementary Tables 12 and 13) or after further adjusting for the uses of antihypertensive, antidiabetic, and lipid-lowering drugs (Supplementary Tables 14 and 15). The results were consistent with the main analyses after excluding participants with a history of cancer (Supplementary Tables 16 and 17) or conducting the competing risk analysis between non-cardiovascular and cardiovascular mortality (Supplementary Table 18). Furthermore, the association of the eGDR with mortality risks were similar to the main analyses when using data not imputed (Supplementary Tables 19 and 20).

4. Discussion

Drawing from extensive, nationally representative samples of adults from the United States and China, this study examined the associations of eGDR with all-cause and cardiovascular mortality risks within the framework of CKM syndrome. The findings indicated that adults with high levels of eGDR had lower risks of all-cause and cardiovascular mortality in both NHANES and CHARLS. There was an approximately linear dose-response association of increased eGDR with reduced risk of all-cause and cardiovascular mortality. Our findings lend support to the eGDR to have utility in the prediction of future all-cause and cardiovascular mortality among adults with CKM syndrome.

Insulin resistance is characterized by a diminished sensitivity or responsiveness to the metabolic actions of insulin, leading to the appearance of metabolic abnormalities [26]. The eGDR, recently developed as an indicator of insulin resistance, has garnered widespread use owing to its simplicity and precision [17,27]. A cohort study with 1862 Chinese community-dwelling elderly participants showed that a high eGDR level was associated with a lower risk of all-cause mortality during a median follow-up of 10.8 years [28]. Another study analyzing 17,787 adults from NHANES found that for each SD increment in eGDR, the risk of all-cause and cardiovascular mortality decreased by 10 % and 15 %, respectively, with the dose-response curve indicating a linear relationship [16]. Our study unveiled an inverse correlation between eGDR and all-cause and cardiovascular mortality among adults with CKM syndrome. Specifically, for every 1-SD rise in eGDR, 18-24 % of all-cause mortality and 40 % of cardiovascular mortality risks were reduced. More importantly, the dose-response curve further substantiated this negative linear association. These findings help elucidate the prognostic significance of eGDR in individuals with CKM syndrome, facilitating more accurate identification of high-risk individuals.

Our results suggested significant associations of eGDR with all-cause and cardiovascular mortality among participants with early and advanced CKM stages, while no association among those with no CKM syndrome. This difference may be explained by two potential reasons. One reason is that insulin resistance played a pathogenic role in promoting thrombosis and exacerbated atherosclerosis by inhibiting the release of nitric oxide and vasodilation [29,30]. Insulin resistance also induced the overproduction of reactive oxygen species, causing endothelial damage and inflammation [31]. Therefore, better insulin sensitivity, reflected by higher levels of eGDR, reduced the risk of all-cause and CVD deaths in the CKM population (CKM stages 1-4). However, no CKM individuals did not have significant metabolic or cardiovascular risk factors, making eGDR irrelevant as a risk marker. The other reason was the relatively low proportions of no CKM participants (NHANES: 11.2 %; CHARLS: 10.7 %) and much lower mortality risks among no CKM participants compared with those with higher CKM stages, resulting in not yet reaching a statistical association. Notably, our findings indicated that the interaction among different CKM syndrome stages was statistically significant, emphasizing the necessity for more extensive, larger prospective cohort studies to elucidate whether the association between eGDR and mortality differs across CKM syndrome stages. Additionally, subgroup analyses revealed that the impact of eGDR on mortality varied by age, indicating that younger individuals might benefit more from higher eGDR levels compared to older ones. Younger participants typically usually have better metabolic reserve capacity, so improving insulin sensitivity at a younger age will yield better survival. Moreover, the relationship between eGDR and mortality were consistent within other subgroups (gender, smoking status, drinking status, obesity status, HbA1c levels, and hypertension status) in both NHANES and CHARLS. The results of several sensitivity analyses further corroborated the robustness of our findings.

Our study has important clinical and public health implications. First, the AHA recommends screening for obesity, hypertension, diabetes, metabolic syndrome, and all key CKM risk factors, and suggests early management of these conditions to improve the clinical outcomes [6,7]. Our study introduced a novel indicator of eGDR to comprehensively evaluate the mortality risks of participant with CKM syndrome. In addition, by calculating WC, hypertension and HbA1c, eGDR can be easily obtained from the single sample, even in primary healthcare institutions, enhancing the practicality and generalizability of eGDR in clinical and epidemiological investigations. Furthermore, integrating eGDR assessment into routine clinical practice is necessary, especially in the CKM syndrome population. Participants with low levels of eGDR should be considered as the primary target to prevent mortality. Participants with high levels of eGDR also need to evaluate the components of eGDR so that at-risk individuals can be identified early and the corresponding prevention measures can be performed to delay or prevent the eGDR reduction. In addition, interventions targeting eGDR components (e.g. WC reduction and lowering HbA1c levels) might improve survivals of individuals with CKM syndrome. For example, in a cohort study of 58,961 women over 18.6 years of follow-up, intentional weight loss with WC reductions were associated with lower risk of all-cause (HR 0.85; 95 % CI 0.76-0.95) and cardiovascular mortality (HR 0.79; 95 % CI 0.72-0.87) [32]. Further studies, both from real-world data and clinical trials, are needed to explore the optimal intervention for improving and maintain high levels of eGDR and evaluate its efficacy and safety in clinical practice.

This study has several strengths. First, this study evaluated the associations of eGDR with all-cause and cardiovascular mortality among adults with CKM syndrome for the first time. In addition, our findings are derived from two large prospective cohorts of different ethnicities, with rigorous study designs and long-term follow-up. The consistency of our results in both cohorts implies the broad applicability of our findings. Finally, subgroup and sensitivity analyses affirmed the solidity of our findings.

There are also limitations in this study. Firstly, our analyses were confined to the clinical characteristics accessible within the NHANES and CHARLS datasets. The recommended data to define advanced CKM syndrome stages, including cardiac biomarkers, echocardiography, coronary angiography, cardiac computed tomography, atrial fibrillation, and peripheral artery disease, were unavailable, which may lead to underestimation of CKM stages 3 and 4. However, our findings suggested that the associations of eGDR with mortality were consistent among participants with early and advanced stages. Therefore, the underestimation of advanced CKM stages would not attenuate the observed associations. Secondly, despite adjusting for multiple confounders, residual confounding may persist, including genetic predisposition, physical activity, dietary patterns, and socioeconomic status. For example, unhealthy diet, inactive physical activity, and low socioeconomic status have been reported to be associated with increased risks for mortality [[33], [34], [35]]. Therefore, not adjusting for these unmeasured factors might lead to overestimating the impact of eGDR on mortality. Thirdly, selection bias might occur in the main analyses due to the exclusion of participants with missing data necessary for calculating eGDR, determining CKM stages, or those lost to follow-up. Fourthly, we only used measures of eGDR at a single visit, and we did not consider changes in eGDR over time, which might impact the accuracy of our observed results. Finally, the two cohorts represented distinct populations with differences in age distribution (US adults aged ≥20 years vs. Chinese adults aged ≥45 years), which may influence the findings of this study. In addition, the generalizability of our results may be limited by the differences in racial background, dietary habits, and living environment. Therefore, further research should incorporate more diverse population samples from various ethnic groups to validate the global applicability of our results.

5. Conclusions

The study unveiled a significant association of eGDR with reduced all-cause and cardiovascular mortality in adults with CKM syndrome. Importantly, this association manifested as a linear relationship, suggesting eGDR can serve as a promising predictor and therapeutic target for reducing mortality risks among adults with CKM syndrome in clinical practice.

Data Sharing Statement

Data from the National Health and Nutrition Examination Survey and the China Health and Retirement Longitudinal Study are publicly available online.

Funding

None.

CRediT authorship contribution statement

Yupeng Wei: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Jiangtao Li: Writing – review & editing, Supervision, Resources, Project administration, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.