Associations of dietary inflammatory and antioxidant indices with mortality in adults with cardiovascular–kidney–metabolic syndrome: A mediation analysis

Abstract

Cardiovascular–kidney–metabolic (CKM) syndrome carries high mortality risk. Whether A Body Shape Index (ABSI), Weight-Adjusted Waist Index (WWI), Systemic Inflammation Response Index (SIRI), and Systemic Immune-Inflammation Index (SII) mediate the associations of Dietary Inflammatory Index (DII) and Composite Dietary Antioxidant Index (CDAI) with mortality in this population remains unknown. We included 21,420 adults with CKM from National Health and Nutrition Examination Survey 1999 to 2018 with mortality follow-up through December 31, 2019. DII and CDAI were calculated from 24-hour dietary recalls. Survey-weighted Cox proportional hazards regression was used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for all-cause and cardiovascular mortality, modeling DII and CDAI as continuous variables and tertiles. Restricted cubic splines and 2-piecewise Cox models assessed nonlinearity. Mediation analyses evaluated indirect effects through ABSI, WWI, SIRI, and SII. During follow-up, 3532 all-cause and 1103 cardiovascular deaths occurred. In fully adjusted models, higher DII was associated with increased all-cause mortality (HR per 1-unit increase: 1.06, 95% CI: 1.04–1.09) and cardiovascular mortality (HR: 1.06, 95% CI: 1.02–1.10); the highest versus lowest tertile was associated with HRs of 1.27 (95% CI: 1.16–1.39) and 1.27 (95% CI: 1.08–1.49), respectively. Higher CDAI was associated with lower all-cause and cardiovascular mortality; the highest versus lowest tertile showed HRs of 0.84 (95% CI: 0.77–0.93) and 0.78 (95% CI: 0.65–0.93), respectively. Nonlinearity was observed between CDAI and all-cause mortality (P = .017). The percentage mediated by ABSI, WWI, SIRI, and SII ranged from approximately 4% to 9% for the association with all-cause mortality. In adults with CKM, DII and CDAI were independently associated with mortality outcomes, and ABSI, WWI, SIRI, and SII mediated a modest proportion of the association with all-cause mortality. Further studies are needed to confirm these findings.

1. Introduction

In recent times, there has been a growing acknowledgments of the cardiovascular–kidney–metabolic (CKM) syndrome as a significant public health concern globally, primarily due to its substantial impact on both disease burden and mortality levels. CKM syndrome denotes the intricate interactions among cardiovascular diseases (CVD), chronic kidney disease (CKD), and metabolic disorders such as obesity and diabetes mellitus (DM).^[1,2] These ailments frequently exist together, creating a complex network of pathophysiological processes that markedly elevate the risks of morbidity and mortality, especially from cardiovascular issues. Extensive studies have validated the complex relationships among the cardiovascular, renal, and metabolic systems.^[3] CKD is recognized as a major factor contributing to CVD, as a diminished estimated glomerular filtration rate is closely associated with an increased likelihood of CVD and death.^[4] Moreover, about 20% of patients experiencing heart failure also have DM, which is nearly 4 times the incidence found in individuals without heart failure, where prevalence is between 4% to 6%.^[5] Recent statistics indicate that nearly 40% of people with DM also suffer from CKD.^[6] In the past few decades, the occurrence of CKM syndrome has escalated significantly, largely due to demographic shifts such as an aging population and a rise in lifestyle-related risk factors, which include sedentary living, poor dietary habits, and escalating obesity rates.^[7]

The Dietary Inflammatory Index (DII) serves as a measure based on the nutrient consumption of the population. This index is employed to evaluate the inflammatory potential of diets^[8] and is predominantly utilized in epidemiological research to investigate the relationship between dietary inflammation potential and various diseases.^[9–12] The Composite Dietary Antioxidant Index (CDAI) was created to measure the antioxidant capacity of dietary intake.^[13,14] Recent studies have consistently indicated that a high DII score or a low CDAI score correlates with an increased risk of developing multiple health conditions, including heart disease, diabetes, chronic obstructive pulmonary disease, and mild to moderate CKD.^[15–18]

In parallel, increasing evidence supports the prognostic value of circulating inflammatory and nutritional biomarkers in CVDs. Among these biomarkers, serum albumin has attracted particular attention because it reflects not only nutritional status but also systemic inflammation, oxidative stress, endothelial function, and overall disease burden. Previous studies have shown that lower serum albumin levels are associated with adverse outcomes in patients with acute coronary syndrome, permanent pacemaker implantation, atrial fibrillation, and heart failure with reduced ejection fraction receiving implantable cardioverter defibrillators. These findings suggest that nutritional and inflammatory biomarkers may provide clinically relevant information for cardiovascular risk stratification beyond conventional risk factors.

Although these studies highlight the clinical importance of inflammatory and nutritional biomarkers in CVD, evidence on the associations of dietary inflammatory and antioxidant indices with mortality in CKM syndrome remains limited. Moreover, whether adiposity- and inflammation-related indices, including A Body Shape Index (ABSI), Weight-Adjusted Waist Index (WWI), Systemic Inflammation Response Index (SIRI), and Systemic Immune-Inflammation Index (SII), mediate these associations has not been examined. We therefore investigated the associations of DII and CDAI with all-cause and cardiovascular mortality and evaluated the potential mediating roles of these indices.

2. Methods

2.1. Study design and population

Data for this study were obtained from the National Health and Nutrition Examination Survey (NHANES), a program conducted by the Centers for Disease Control and Prevention to evaluate the health and nutritional status of the US civilian, non-institutionalized population. NHANES employs a complex, multistage probability sampling design to produce nationally representative data, collecting comprehensive information on demographics, dietary intake, medical history, physical examinations, and laboratory measurements. All survey protocols were approved by the National Center for Health Statistics Research Ethics Review Board, and written informed consent was obtained from all participants. Public-use data files are freely available at https://www.cdc.gov/nchs/nhanes/index.html.

For this analysis, we combined data from 10 consecutive NHANES cycles conducted between 1999 and 2018. A total of 101,316 participants were initially identified. We excluded individuals younger than 18 years or with missing data required to determine CKM stage, resulting in 25,223 eligible adults. Participants were further excluded if they lacked mortality follow-up data (n = 30) or had missing values for both the DII and the CDAI (n = 3773). Consequently, 21,420 participants were included in the final analyses (Fig. 1).

Figure 1.

Flowchart of participants’ selection from the NHANES 1999 to 2018.

2.2. Calculation of CDAI and DII

Dietary intake data were collected through two 24-hour dietary recall interviews conducted by trained personnel. The 1st recall was performed in person at the NHANES Mobile Examination Center, and the 2nd was conducted via telephone 3 to 10 days later. The average of the 2 recalls was used to estimate usual intake; if only 1 recall was available, that value was used. All nutrient intake data used for CDAI and DII calculations were derived from the NHANES dietary recall database.

The DII was calculated to quantify the overall inflammatory potential of the diet using the method developed by Shivappa et al. In our analysis, 28 of the 45 food parameters were available from NHANES and included in the calculation. Briefly, each dietary component was standardized to a global reference database, converted to a centered percentile, multiplied by its respective inflammatory effect score, and summed across all components to derive an individual DII score. Higher DII values indicate a more pro-inflammatory diet, whereas lower values suggest a more anti-inflammatory diet.

The CDAI was calculated to assess the overall antioxidant capacity of the diet, following the approach described by Wright et al. Specifically, intakes of 6 dietary antioxidants (vitamin A, vitamin C, vitamin E, zinc, selenium, and carotenoids) were standardized by subtracting the study population mean and dividing by the standard deviation. The CDAI score was then computed by summing these standardized values across the 6 nutrients, yielding:

$${\displaystyle \mathrm{C}\mathrm{D}\mathrm{A}\mathrm{I}=\underset{\mathrm{i}=1}{\sum ^6}\frac{{\mathrm{X}}_{\mathrm{i}}\text{-}{\mu }_{\mathrm{i}}}{{\mathrm{S}}_{\mathrm{i}}}}$$

Notably, dietary contributions from supplements, medications, and plain drinking water were excluded from both DII and CDAI calculations, aligning with standard practice in similar epidemiological studies.^[19]

2.3. Definition of CKM syndrome

CKM syndrome in this study was defined as the coexistence of clinical or subclinical CVD, CKD, and metabolic abnormalities, based on criteria reported in previous studies.^[18,19] Clinical CVD included heart failure, coronary heart disease, myocardial infarction, and stroke, while subclinical CVD was determined by a predicted 10-year CVD risk of ≥20%, estimated using a modified American Heart Association algorithm that incorporates age, sex, smoking status, blood pressure, cholesterol levels, diabetes, kidney function, and the use of antihypertensive or lipid-lowering medications.^[20]

CKD was classified following the kidney disease: Improving Global Outcomes (KDIGO) guidelines,^[21] using categories of estimated glomerular filtration rate (<30, 30–44, 45–59, and ≥60 mL/min/1.73 m²) and urinary albumin-to-creatinine ratio (<30, 30–299, and ≥300 mg/g). CKD was considered present if participants met criteria for moderate to high risk. Metabolic conditions comprised overweight or obesity, abdominal obesity, prediabetes, DM, hypertension, dyslipidemia, and metabolic syndrome. Participants were stratified into 4 CKM stages: stage 0 (absence of metabolic or kidney disorders), stage 1 (presence of obesity or prediabetes), stage 2 (one additional metabolic abnormality or CKD), stage 3 (subclinical CVD with metabolic disorders or CKD), and stage 4 (clinical CVD coexisting with metabolic disorders or CKD).^[20]

2.4. Definition of clinical outcomes

Mortality data, including dates and causes of death, were obtained from the National Death Index database through December 31, 2019. Causes of death were classified according to the International Classification of Diseases, 10th Revision (ICD-10). Participants were followed from their initial NHANES examination date until the date of death or the end of the follow-up period, whichever came first. The primary outcome of this study was all-cause mortality. Secondary outcomes included mortality attributable to cardiovascular causes as well as mortality from non-cardiovascular causes.^[21]

2.5. Covariates

Covariates were selected based on established relevance to cardiometabolic outcomes and included demographic (age, sex, race/ethnicity), socioeconomic (education level, poverty income ratio), lifestyle (alcohol use, physical activity, marital status), and clinical factors (hypertension, diabetes status, hyperlipidemia, CVDs). Physical activity was categorized by total metabolic equivalent (MET)-minutes per week (<600, ≥600), and alcohol use by standard frequency categories. Hypertension, diabetes, and CVDs were defined using self-reported physician diagnoses, medication use, or examination measures.

2.6. Statistical analysis

All analyses incorporated NHANES sampling weights, strata, and primary sampling units to account for the complex, multistage probability design. Continuous variables are presented as weighted means with standard deviations, and categorical variables as weighted percentages. Differences between survivors and non-survivors were evaluated using survey-weighted linear regression for continuous variables and Rao–Scott χ² tests for categorical variables. Univariate Cox proportional hazards models were first constructed to evaluate crude associations between individual covariates and all-cause mortality. Survey-weighted Cox proportional hazards regression models were then used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the associations of DII and CDAI with all-cause and cardiovascular mortality. DII and CDAI were modeled as continuous variables and tertiles. Model I adjusted for age, sex, ethnicity, poverty income ratio, and education level; Model II further adjusted for alcohol use, total physical activity (MET/week), and hypertension. Restricted cubic splines were incorporated within the Cox regression framework to explore potential nonlinear associations. Two-piecewise Cox regression models were applied to assess threshold effects, with inflection points determined by likelihood ratio tests comparing linear and segmented models. Sex-stratified spline analyses were additionally performed to examine whether nonlinear associations differed between males and females. Stratified analyses were conducted across prespecified subgroups to explore the consistency of associations across strata. Kaplan–Meier survival curves were generated to compare survival probabilities across tertiles of DII and CDAI, and differences were evaluated using survey-weighted log-rank tests. Mediation analyses were conducted to evaluate the indirect effects of ABSI, WWI, SIRI, and SII on the associations between DII/CDAI and all-cause mortality. Indirect effects and proportions mediated were estimated using regression-based mediation models adjusted for the same covariates as in the fully adjusted Cox models. All statistical analyses were performed using R version 4.2.0 and EmpowerStats version 4.2. A 2-sided P value < .05 was considered statistically significant.

3. Results

3.1. Characteristics of the study population

Baseline characteristics of participants stratified by survival status are presented in Table 1. A total of 21,420 participants were included in the analysis. Compared with survivors, non-survivors were significantly older, more likely to be male and non-Hispanic White, and exhibited lower levels of physical activity, higher prevalence of hypertension, and more advanced CKM stages (all P < .001). Additionally, non-survivors had higher DII scores, lower antioxidant intake, slightly lower body mass index, and were more frequently widowed or separated. Differences in alcohol consumption patterns and socioeconomic indicators, including education and income levels, were also observed between the 2 groups.

Table 1.

Baseline characteristics of adults with CKM stratified by survival status.

All-cause mortality	Total	Survival	Non-survival	P value
N	21,420	17,888	3532
Age (yr)	50.17 ± 18.09	46.46 ± 16.50	68.96 ± 13.49	<.001
Time (mo)	119.99 ± 67.63	125.62 ± 67.93	91.45 ± 58.26	<.001
DII	1.42 ± 1.78	1.37 ± 1.80	1.67 ± 1.67	<.001
CDAI	0.55 ± 3.97	0.70 ± 3.98	−0.30 ± 3.78	<.001
Body mass index (kg/m2)	29.01 ± 6.76	29.11 ± 6.80	28.51 ± 6.51	<.001
Sex				<.001
Male	10,149 (47.38%)	8195 (45.81%)	1954 (55.32%)
Female	11,271 (52.62%)	9693 (54.19%)	1578 (44.68%)
Ethnicity				<.001
Non-Hispanic White	9871 (46.08%)	7710 (43.10%)	2161 (61.18%)
Non-Hispanic Black	4392 (20.50%)	3726 (20.83%)	666 (18.86%)
Mexican American	3748 (17.50%)	3278 (18.33%)	470 (13.31%)
Other Hispanic	1716 (8.01%)	1578 (8.82%)	138 (3.91%)
Other Race	1693 (7.90%)	1596 (8.92%)	97 (2.75%)
Poverty income ratio				<.001
Poor	3789 (17.69%)	3152 (17.62%)	637 (18.04%)
Nearly poor	5221 (24.37%)	4131 (23.09%)	1090 (30.86%)
Middle income	5449 (25.44%)	4530 (25.32%)	919 (26.02%)
High income	5164 (24.11%)	4603 (25.73%)	561 (15.88%)
Missing	1797 (8.39%)	1472 (8.23%)	325 (9.20%)
Education level, n (%)				<.001
Below high school	2500 (11.67%)	1768 (9.88%)	732 (20.72%)
High school	8074 (37.69%)	6535 (36.53%)	1539 (43.57%)
Above high school	10,819 (50.51%)	9569 (53.49%)	1250 (35.39%)
Missing	27 (0.13%)	16 (0.09%)	11 (0.31%)
Alcohol use, n (%)				<.001
Never	2823 (13.18%)	2273 (12.71%)	550 (15.57%)
Former	3695 (17.25%)	2537 (14.18%)	1158 (32.79%)
Mild	6744 (31.48%)	5708 (31.91%)	1036 (29.33%)
Moderate	2905 (13.56%)	2621 (14.65%)	284 (8.04%)
Heavy	3707 (17.31%)	3369 (18.83%)	338 (9.57%)
Missing	1546 (7.22%)	1380 (7.71%)	166 (4.70%)
Total physical activity (MET/wk)				<.001
<600	5404 (25.23%)	4486 (25.08%)	918 (25.99%)
≥600	10,096 (47.13%)	9075 (50.73%)	1021 (28.91%)
Missing	5920 (27.64%)	4327 (24.19%)	1593 (45.10%)
Hypertension				<.001
No	12,156 (56.76%)	11,128 (62.22%)	1028 (29.11%)
Yes	9259 (43.24%)	6756 (37.78%)	2503 (70.89%)
CKM stages, (%)				<.001
0	1918 (8.95%)	1856 (10.38%)	62 (1.76%)
1	3892 (18.17%)	3727 (20.84%)	165 (4.67%)
2	11,634 (54.31%)	10,327 (57.73%)	1307 (37.00%)
3	1263 (5.90%)	534 (2.99%)	729 (20.64%)
4	2713 (12.67%)	1444 (8.07%)	1269 (35.93%)
Marital status, n (%)				<.001
Married/living with partner	13,019 (61.45%)	11,146 (62.90%)	1873 (54.07%)
Widowed/divorced/separated	4682 (22.10%)	3319 (18.73%)	1363 (39.35%)
Never married	3484 (16.45%)	3256 (18.37%)	228 (6.58%)
All-cause mortality				<.001
Survival	17,888 (83.51%)	17,888 (100.00%)	0 (0.00%)
Non-survival	3532 (16.49%)	0 (0.00%)	3532 (100.00%)
Cardiovascular mortality				<.001
Survival	20,317 (94.85%)	17,888 (100.00%)	2429 (68.77%)
Non-survival	1103 (5.15%)	0 (0.00%)	1103 (31.23%)
Diabetes mellitus				<.001
No	13,762 (66.20%)	12,021 (69.58%)	1741 (49.56%)
Yes	4065 (19.55%)	2865 (16.58%)	1200 (34.16%)
IFG (impaired fasting glycaemia)	1819 (8.75%)	1440 (8.33%)	379 (10.79%)
IGT (impaired glucose tolerance)	1144 (5.50%)	951 (5.50%)	193 (5.49%)
Hyperlipidemia				<.001
No	6073 (28.35%)	5311 (29.69%)	762 (21.57%)
Yes	15,345 (71.65%)	12,575 (70.31%)	2770 (78.43%)
Cardiovascular diseases				<.001
No	18,922 (88.34%)	16,518 (92.35%)	2404 (68.06%)
Yes	2497 (11.66%)	1369 (7.65%)	1128 (31.94%)

Values are presented as weighted mean ± standard deviation for continuous variables and number (weighted percentage) for categorical variables.

Differences between survivors and non-survivors were assessed using survey-weighted linear regression for continuous variables and Rao–Scott χ² tests for categorical variables.

BMI = body mass index, CDAI = Composite Dietary Antioxidant Index, CKM = cardiovascular–kidney–metabolic syndrome, DII = Dietary Inflammatory Index, MET = metabolic equivalent of task.

3.2. Association of DII and CDAI with all-cause and cardiovascular mortality in CKM

Using Cox proportional hazards models, higher DII scores were associated with higher hazards of all-cause and cardiovascular mortality in the fully adjusted model (Model II). Compared with the lowest DII tertile, the highest tertile was associated with increased hazards of all-cause mortality (HR: 1.27, 95% CI: 1.16–1.39; P < .0001) and cardiovascular mortality (HR: 1.27, 95% CI: 1.08–1.49; P = .0033). When modeled continuously, each 1-unit higher DII was associated with a higher hazard of all-cause mortality (HR: 1.06, 95% CI: 1.04–1.09; P < .0001) and cardiovascular mortality (HR: 1.06, 95% CI: 1.02–1.10; P = .0019) (Table 2).

Table 2.

Associations of DII with all-cause and cardiovascular mortality in CKM.

Exposure	Adjust I	Adjust II
All-cause mortality
DII	1.09 (1.07, 1.11) < 0.0001	1.06 (1.04, 1.09) < 0.0001
DII tertile
Low	1.0	1.0
Middle	1.15 (1.06, 1.26) 0.0011	1.12 (1.03, 1.22) 0.0097
High	1.39 (1.27, 1.51) < 0.0001	1.27 (1.16, 1.39) < 0.0001
DII tertile continuous	1.18 (1.13, 1.23) < 0.0001	1.13 (1.08, 1.18) < 0.0001
Cardiovascular mortality
DII	1.10 (1.06, 1.14) < 0.0001	1.06 (1.02, 1.10) 0.0019
DII tertile
Low	1.0	1.0
Middle	1.22 (1.04, 1.42) 0.0119	1.18 (1.01, 1.38) 0.0361
High	1.42 (1.21, 1.66) < 0.0001	1.27 (1.08, 1.49) 0.0033
DII tertile continuous	1.19 (1.10, 1.28) < 0.0001	1.12 (1.04, 1.21) 0.0037

Hazard ratios (HRs) and 95% confidence intervals (CIs) are presented for DII as a continuous variable and by tertiles.

Adjust I model adjust for: age (years), sex, ethnicity, poverty income ratio, education level; adjust II model adjust for:age (years), sex, ethnicity, poverty income ratio, education level, alcohol use, total physical activity (MET/week), hypertension.

CKM = cardiovascular–kidney–metabolic syndrome, DII = Dietary Inflammatory Index.

Conversely, higher CDAI values were associated with lower hazards of mortality. In Model II, the highest CDAI tertile was associated with lower hazards of all-cause mortality (HR: 0.84, 95% CI: 0.77–0.93; P = .0005) and cardiovascular mortality (HR: 0.78, 95% CI: 0.65–0.93; P = .0050) compared with the lowest tertile. In continuous analyses, each 1-unit higher CDAI was associated with lower hazards of all-cause mortality (HR: 0.99, 95% CI: 0.98–1.00; P = .0123) and cardiovascular mortality (HR: 0.98, 95% CI: 0.96–1.00; P = .0344) (Table 3).

Table 3.

Associations of CDAI with all-cause and cardiovascular mortality in CKM.

Exposure	Adjust I	Adjust II
All-cause mortality
CDAI	0.98 (0.97, 0.99) < 0.0001	0.99 (0.98, 1.00) 0.0123
CDAI tertile
Low	1.0	1.0
Middle	0.89 (0.81, 0.97) 0.0075	0.92 (0.84, 1.00) 0.0558
High	0.78 (0.71, 0.86) < 0.0001	0.84 (0.77, 0.93) 0.0005
CDAI tertile continuous	0.88 (0.84, 0.93) < 0.0001	0.92 (0.88, 0.96) 0.0005
Cardiovascular mortality
CDAI	0.97 (0.95, 0.99) 0.0016	0.98 (0.96, 1.00) 0.0344
CDAI tertile
Low	1.0	1.0
Middle	0.95 (0.81, 1.10) 0.4751	0.98 (0.84, 1.14) 0.7507
High	0.71 (0.60, 0.85) 0.0002	0.78 (0.65, 0.93) 0.0050
CDAI tertile continuous	0.85 (0.79, 0.93) 0.0003	0.89 (0.82, 0.97) 0.0080

Hazard ratios (HRs) and 95% confidence intervals (CIs) are presented for CDAI as a continuous variable and by tertiles.

Adjust I model adjust for: age (years), sex, ethnicity, poverty income ratio, education level; adjust II model adjust for: age (years), sex, ethnicity, poverty income ratio, education level, alcohol use, total physical activity (MET/week), hypertension.

CDAI = Composite Dietary Antioxidant Index, CKM = cardiovascular–kidney–metabolic syndrome, MET = metabolic equivalent.

3.3. Nonlinear associations of DII and CDAI with mortality risk in CKM

Restricted cubic spline analyses and 2-piecewise Cox models were used to assess potential nonlinearity (Fig. 2; Table 4).

Figure 2.

(A) Association between Dietary Inflammatory Index (DII) and all-cause mortality. (B) Association between DII and cardiovascular mortality. (C) Association between Composite Dietary Antioxidant Index (CDAI) and all-cause mortality. (D) Association between CDAI and cardiovascular mortality. Restricted cubic spline analyses were applied using Cox regression with adjustments for age (years), sex, ethnicity, poverty income ratio, education level, alcohol use, total physical activity (MET/week), and hypertension. CDAI = Composite Dietary Antioxidant Index; DII = Dietary Inflammatory Index.

Table 4.

Threshold effect analysis.

Variable (CDAI)	All-cause mortality	Cardiovascular mortality
Model I
One-line effect	0.99 (0.98, 1.00) 0.0123	0.98 (0.96, 1.00) 0.0344
Model II
Inflection point (K)	1.81	−3.36
< K segment effect	0.97 (0.95, 0.99) 0.0006	1.08 (0.95, 1.22) 0.2590
> K segment effect	1.01 (0.99, 1.02) 0.5358	0.97 (0.95, 0.99) 0.0118
P for log-likelihood ratio test	.017	.134
Variable (DII)	All-cause mortality	Cardiovascular mortality
Model I
One-line effect	1.06 (1.04, 1.09) < 0.0001	1.06 (1.02, 1.10) 0.0019
Model II
Inflection point (K)	−1.27	3.64
< K segment effect	1.16 (1.00, 1.34) 0.0479	1.08 (1.03, 1.12) 0.0004
> K segment effect	1.06 (1.03, 1.08) < 0.0001	0.65 (0.37, 1.13) 0.1266
P for log-likelihood ratio test	.242	.070

Model I: linear (one-line) Cox regression; Model II: 2-piecewise Cox regression with estimated inflection point (K).

All models were adjusted for age (years), sex, ethnicity, poverty income ratio, education level, alcohol use, total physical activity (MET/week), hypertension.

CIs = 95% confidence intervals; HRs = hazard ratios, MET = metabolic equivalent.

For CDAI, a nonlinear association with all-cause mortality was detected (P for log-likelihood ratio test = 0.017), with an estimated inflection point at 1.81. Below this point, higher CDAI was associated with lower all-cause mortality (HR: 0.97, 95% CI: 0.95–0.99; P = .0006), whereas no significant association was observed above the threshold (HR: 1.01, 95% CI: 0.99–1.02; P = .5358). No statistically significant nonlinearity was observed for cardiovascular mortality (P = .134). In the linear model, higher CDAI was associated with lower cardiovascular mortality (HR: 0.98, 95% CI: 0.96–1.00; P = .0344).

For DII, no statistically significant nonlinear association was observed for all-cause mortality (P = .242). In the linear model, higher DII was associated with increased all-cause mortality (HR: 1.06, 95% CI: 1.04–1.09; P < .0001). For cardiovascular mortality, the test for nonlinearity was not statistically significant (P = .070), and the linear model showed a positive association (HR: 1.06, 95% CI: 1.02–1.10; P = .0019).

3.4. Survival analysis of DII and CDAI in CKM

Kaplan–Meier curves were generated to compare survival across tertiles of DII and CDAI (Fig. 3). For DII, survival distributions differed significantly across tertiles for both all-cause and cardiovascular mortality (log-rank P < .0001 for both; Fig. 3A and B). For CDAI, survival distributions also differed significantly across tertiles for all-cause and cardiovascular mortality (log-rank P < .0001 for both; Fig. 3C and D).

Figure 3.

(A) DII tertiles and all-cause mortality; (B) DII tertiles and cardiovascular mortality; (C) CDAI tertiles and all-cause mortality; (D) CDAI tertiles and cardiovascular mortality. Log-rank tests were used to compare survival distributions across tertile groups. CDAI = Composite Dietary Antioxidant Index; CKM = cardio–kidney–metabolic; DII = Dietary Inflammatory Index.

3.5. Mediation analysis of the associations of DII and CDAI with all-cause mortality

The mediation analyses showed that the indirect effects through ABSI, WWI, SIRI, and SII accounted for 7.06%, 9.22%, 5.00%, and 6.93% of the association between DII and all-cause mortality, respectively (Fig. 4). For CDAI, the corresponding proportions of the association with all-cause mortality accounted for by the indirect effects through ABSI, WWI, SIRI, and SII were 8.31%, 9.42%, 4.58%, and 8.16%, respectively (Fig. 4).

Figure 4.

The percentage mediated represents the proportion of the total association explained by ABSI, WWI, SIRI, and SII. (A) Mediation effect of ABSI on the association between DII and all‐cause mortality. (B) Mediation effect of WWI on the association between DII and all‐cause mortality. (C) Mediation effect of SIRI on the association between DII and all‐cause mortality. (D) Mediation effect of SII on the association between DII and all‐cause mortality. (E) Mediation effect of ABSI on the association between CDAI and all‐cause mortality. (F) Mediation effect of WWI on the association between CDAI and all‐cause mortality. (G) Mediation effect of SIRI on the association between CDAI and all‐cause mortality. (H) Mediation effect of SII on the association between CDAI and all‐cause mortality. All models were adjusted for age, sex, ethnicity, poverty income ratio, education level, alcohol use, total physical activity (MET/wk), and hypertension. ABSI = A Body Shape Index; CDAI = Composite Dietary Antioxidant Index; DII = Dietary Inflammatory Index; SIRI = Systemic Inflammation Response Index; SII = Systemic Immune-Inflammation Index; WWI = Weight-Adjusted Waist Index.

3.6. Sex-stratified nonlinear associations of DII and CDAI with mortality in CKM

Sex-stratified smooth curve analyses are presented in Figure S1, Supplemental Digital Content 1. For DII, the direction of the association with all-cause and cardiovascular mortality was generally similar in males and females. For CDAI, inverse associations with both mortality outcomes were observed in sex-stratified analyses. The overall patterns were broadly comparable between males and females.

3.7. Stratified analysis

Stratified analyses are presented in Tables S1 and S2, Supplemental Digital Content 2. For DII, positive associations with all-cause and cardiovascular mortality were generally observed across subgroups. For CDAI, inverse associations with both mortality outcomes were generally observed across subgroups.

3.8. Univariate analysis

Univariate Cox regression results for all-cause mortality are presented in Table S3, Supplemental Digital Content 3. In single-variable models, older age, higher DII, hypertension, and more advanced CKM stages were associated with higher all-cause mortality, whereas higher CDAI and greater physical activity were associated with lower risk. Several demographic and socioeconomic variables were also associated with all-cause mortality in univariate analyses.

4. Discussion

This study examined the associations of DII and CDAI with mortality and evaluated the mediating roles of ABSI, WWI, SIRI, and SII among adults with CKM syndrome. Higher DII was associated with increased all-cause and cardiovascular mortality, whereas higher CDAI was associated with lower mortality. ABSI, WWI, SIRI, and SII statistically mediated approximately 4% to 9% of the association with all-cause mortality, while the direct associations remained significant. These findings indicate that adiposity- and inflammation-related indices may partially account for the observed associations between dietary profiles and mortality in this high-risk population.

These findings are in line with previous studies reporting that pro-inflammatory diets are linked to greater risks of cardiometabolic diseases and mortality,^[22] whereas diets rich in antioxidants may have protective effects.^[23] A higher DII reflects greater intake of foods that may promote systemic inflammation, which has been implicated in the progression of CVD, kidney dysfunction, and metabolic disorders.^[24–28] Conversely, a higher CDAI represents greater dietary intake of antioxidant nutrients, which may help mitigate oxidative stress and inflammation.^[29] Prior studies in other populations have demonstrated similar trends, but evidence specifically in individuals with CKM syndrome has been lacking. Our study extends this evidence by highlighting that dietary inflammatory and antioxidant profiles may be relevant to mortality outcomes in this high-risk group.

Our findings should also be interpreted within the broader literature on inflammatory and nutritional biomarkers in CVD. Serum albumin is a representative biomarker at the intersection of nutrition and inflammation. Low albumin levels may reflect malnutrition, chronic inflammation, impaired hepatic synthetic response, increased catabolism, and systemic disease severity. Mechanistically, albumin has antioxidant, anti-inflammatory, anticoagulant, and endothelial-protective properties; therefore, reduced albumin may contribute to oxidative stress, endothelial dysfunction, plaque instability, thrombosis, and adverse cardiovascular remodeling. Consistent with this concept, prior evidence has shown that serum albumin predicts mortality in acute coronary syndrome and in patients undergoing permanent pacemaker implantation. Albumin has also been identified as an independent predictor of early mortality in patients with heart failure with reduced ejection fraction receiving implantable cardioverter defibrillators. Moreover, studies in elderly patients with dual-chamber pacemakers suggest that albumin is associated with atrial fibrillation occurrence, further supporting its relevance to cardiovascular risk stratification.^[21] Together with our findings on DII, CDAI, ABSI, WWI, SIRI, and SII, these studies indicate that nutritional and inflammatory biomarkers may capture overlapping but complementary dimensions of cardiovascular risk in CKM syndrome. However, serum albumin was not evaluated as a mediator in the present analysis; future studies should determine whether albumin improves risk prediction when combined with dietary inflammatory and antioxidant indices.

Beyond circulating metabolic and inflammatory markers, early identification of target-organ damage is also important for refining cardiovascular risk stratification in patients with CKM syndrome. Advanced cardiovascular imaging may provide complementary information, particularly for detecting subclinical myocardial dysfunction before overt heart failure develops. Speckle-tracking echocardiography, through the assessment of left ventricular global longitudinal strain and related strain-rate parameters, can detect subtle abnormalities in myocardial mechanics even when conventional left ventricular ejection fraction remains preserved. A recent systematic review and meta-analysis of patients with metabolic dysfunction-associated steatotic liver disease, a cardiometabolic condition closely related to the CKM spectrum, reported significantly impaired left ventricular global longitudinal strain and strain-rate parameters compared with healthy controls, whereas the difference in left ventricular ejection fraction was not statistically significant. These findings suggest that myocardial strain imaging may capture early cardiac involvement that is not apparent on conventional echocardiographic assessment. Therefore, integrating dietary inflammatory and antioxidant risk profiles with sensitive imaging biomarkers may provide a more comprehensive approach to identifying CKM patients at particularly high cardiovascular risk and may support earlier lifestyle, dietary, and preventive therapeutic interventions. However, because speckle-tracking echocardiography-derived parameters were not available in NHANES, future prospective studies incorporating both nutritional indices and cardiac imaging are needed to determine whether this multimodal strategy improves risk prediction beyond conventional clinical models.

This study has several strengths, including the use of a large, nationally representative sample with rigorous adjustment for multiple potential confounders and the incorporation of survey weights to ensure generalizability. However, some limitations should be acknowledged. First, the cross-sectional design with linked mortality follow-up allows for association but cannot establish causality. Second, dietary intake was assessed using self-reported data, which may be subject to recall bias and measurement error. Third, residual confounding from unmeasured factors cannot be completely ruled out. Further prospective studies and interventional research are needed to confirm these associations and clarify the potential mechanisms linking dietary inflammation and antioxidant intake to mortality risk in CKM populations.

5. Conclusion

In this nationally representative cohort of U.S. adults with CKM syndrome, higher dietary inflammatory potential was associated with increased all-cause and cardiovascular mortality, whereas higher dietary antioxidant capacity was associated with lower mortality. ABSI, WWI, SIRI, and SII statistically mediated a modest proportion of the association with all-cause mortality.

Acknowledgments

We sincerely thank all the participants and staff of the NHANES project.

Author contributions

Conceptualization: Yuefeng Li.

Data curation: Yuefeng Li.

Formal analysis: Yuefeng Li.

Funding acquisition: Jing Kang.

Supervision: Jing Kang.

Writing – original draft: Yuefeng Li.

Writing – review & editing: Jing Kang.