Myeloid inflammation thresholds stratify mortality risk in early-stage cardiovascular-kidney-metabolic syndrome: MLR-driven findings from NHANES

Abstract

Background

Cardiovascular-kidney-metabolic (CKM) syndrome is a major public health concern associated with increased mortality. Inflammation plays a critical role in CKM progression and outcomes. This study investigates the relationship between inflammatory indices and mortality risk in CKM patients.

Methods

A comprehensive analysis of data from 26,265 participants in the National Health and Nutrition Examination Survey (NHANES) database (2007–2016) with CKM syndrome stages 0–4 was conducted. The primary outcomes of the study were all-cause and cardiovascular mortality. The inflammatory indices encompassed the systemic inflammation response index (SIRI), neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), platelet-to-lymphocyte ratio (PLR), systemic immune-inflammation index (SII), aggregate index of systemic inflammation (AISI), and neutrophil-to-albumin ratio (NAR). Multivariable Cox models, adjusted for demographic and clinical confounders, were employed to examine nonlinearity, alongside restricted cubic splines and threshold analyses. The present study sought to compare the prognostic accuracy of the time-dependent ROC (Receiver Operating Characteristic) at 93 months.

Results

During a median follow-up of 93.4 months, 2,292 subjects experienced all-cause mortality and 701 experienced cardiovascular deaths. In the adjusted models, elevated SIRI (all-cause HR 1.11, 95% CI 1.06–1.15; cardiovascular HR 1.18, 1.10–1.27), NLR (all-cause HR 1.08, 1.05–1.12; cardiovascular HR 1.11, 1.05–1.17) and MLR (all-cause HR 2.27, 1.71–3.01; cardiovascular HR 3.37, 2.09–5.44) were independently associated with mortality (all p < 0.0001). Dose–response analyses revealed nonlinear J-shaped relationships: MLR showed marked risk above 0.19 (HR 2.59), NLR risk was greatest below 3 (HR 1.14), and SIRI thresholds differed for all-cause (> 1.74, HR 1.09) versus cardiovascular (> 0.38, HR 1.17) outcomes. At 93 months, MLR demonstrated the highest discriminatory ability (AUC 0.630; C-index 0.667; p < 0.001), outperforming SIRI (AUC 0.611) and NLR (AUC 0.602). PLR, AISI, SII and NAR showed limited predictive value due to imbalanced sensitivity–specificity. The impact of age and the early stages of CKD on the modification of associations was investigated.

Conclusion

Systemic inflammatory indices demonstrated nonlinear, J-shaped associations with mortality in CKM syndrome, with the MLR showing the strongest association across disease trajectories. MLR, NLR, and SIRI were identified as potential risk indicators, with stronger associations observed in younger patients and those with early-stage CKM syndrome.

Highlights

• Systemic inflammatory markers (SIRI, NLR, MLR) were significantly associated with increased mortality risk in CKM syndrome.

• Most inflammation indices exhibited nonlinear, J-shaped associations with mortality.

• Nonlinear threshold analyses identified specific risk inflection points for SIRI, NLR, and MLR.

• These associations were stronger in younger patients (≤ 60 years) and those with early CKM stages (1–2).

Supplementary Information

The online version contains supplementary material available at 10.1007/s12026-025-09707-5.

Introduction

Cardiovascular-kidney-metabolic (CKM) syndrome is a systemic multi-organ dysfunction characterized by the pathological interplay of metabolic disorders (e.g., obesity, diabetes), chronic kidney disease (CKD), and cardiovascular disease (CVD) [1]. This syndrome represents a growing global health burden, contributing significantly to disability-adjusted life years due to its escalating prevalence and associated complications [2]. Adults in CKM Stage 3 had approximately 2.2- to 2.7-times higher risk for CVD, compared with adults in lower CKM stages. The progression of CKM syndrome is driven by interconnected mechanisms including chronic inflammation, oxidative stress, and metabolic dysregulation, which collectively exacerbate end-organ damage and elevate risks of cardiovascular events and mortality [1, 3].

Emerging evidence suggests that early-stage lifestyle interventions, particularly targeted weight management, may halt or even reverse disease progression [1, 3]. Notably, clinical practice currently lacks reliable tools for early detection and risk stratification of CKM syndrome, underscoring the urgent need to identify novel biomarkers associated with its pathogenesis. Such biomarkers could enable timely interventions to mitigate downstream complications and improve patient outcomes.

Various inflammatory markers have been established to assess the presence and severity of systemic inflammation, which plays a pivotal role in these conditions [4, 5]. The Systemic Immune-Inflammation Index (SII) predicts all-cause and cardiovascular mortality in patients with myocardial infarction [6] and CKM syndrome [4], while the Systemic Inflammatory Response Index (SIRI) is associated with elevated risks of these outcomes in populations with chronic kidney disease (CKD) [7]. While numerous inflammatory biomarkers have been proposed for CKM syndrome, their stage-specific prognostic utility and clinical relevance in mortality risk stratification remain poorly defined. This study systematically evaluates the discriminative capacity of inflammatory indices across CKM stages 0–4, aiming to identify optimal biomarkers for individualized risk prediction and evidence-based prognosis management.

Methods

Data source and study population

This prospective cohort study utilized data from the National Health and Nutrition Examination Survey (NHANES, 2007–2016), a nationally representative survey assessing health and nutritional status in the US population. NHANES employs a multistage probability sampling design to ensure generalizability, with publicly accessible protocols and datasets (https://www.cdc.gov/nchs/nhanes/). The study adhered to the Declaration of Helsinki and received approval from the National Center for Health Statistics Ethics Review Board.

The study cohort was derived from 50,588 participants enrolled in five NHANES biennial cycles (2007–2016). To establish the analytic sample, we first excluded 20,621 individuals aged ≤ 18 years, followed by sequential exclusions of 71 participants with missing mortality data, 686 with follow-up durations < 24 months, 1,414 with incomplete CKM stage classifications, and 2,531 with missing or extreme white blood cell (WBC) counts (> 99th percentile). After applying these criteria, the final analytic cohort comprised 26,265 adults (Fig. 1).

Fig. 1.

Flowchart of study population. Abbreviations: NHANES, National Health and Nutrition Examination Survey; CKM, Cardiovascular-Kidney-Metabolic; WBC, white blood cell counts

CKM syndrome stages 0–4

The CKM syndrome, a multisystem disorder driven by interactions among cardiovascular disease (CVD), CKD, and metabolic dysfunction, was stratified into five stages: Stage 0: No identifiable CKM risk factors; Stage 1: Excess adiposity (BMI ≥ 25 kg/m² or waist circumference > 102 cm [male]/88 cm [female]) without metabolic abnormalities; Stage 2: Metabolic syndrome (≥ 2 of hypertension, dyslipidemia, hyperglycemia, or insulin resistance) or moderate-to-high CKD risk. Stage 3: Subclinical CVD, defined by either (a) 10-year predicted CVD risk ≥ 20% (PREVENT equations [8]) or​(b) very high-risk CKD. Stage 4: Clinically established CVD (self-reported coronary heart disease, myocardial infarction, stroke, or heart failure). CKD classification followed the Kidney Disease Improving Global Outcomes (KDIGO) criteria, utilizing estimated glomerular filtration rate (eGFR) and urine albumin-to-creatinine ratio (UACR) [9]. The eGFR was calculated using the 2021 race- and ethnicity-free Chronic Kidney Disease Epidemiology Collaboration creatinine equation [10]. For more details on these equations and additional staging criteria, refer to Supplementary Appendix Table S1, S2.

Outcome variables

The primary outcome was all-cause mortality, with cardiovascular mortality as the secondary endpoint. Mortality status and causes were ascertained through linkage of NHANES participants to the National Death Index (NDI) using the NHANES Public-Use Linked Mortality File (updated through December 31, 2019). Death records were probabilistically matched to NDI entries by the National Center for Health Statistics. Underlying causes of death were classified per the International Classification of Diseases, 10th Revision (ICD-10): All-cause mortality: Encompassed deaths from any cause (ICD-10 codes A00-Y99); Cardiovascular mortality: Included deaths attributed to heart diseases (ICD-10 I00-I09, I11, I13, I20-I51) and cerebrovascular diseases (ICD-10 I60-I69).

Exposure variable

The exposure variables in this study comprised seven inflammatory markers: the Systemic Immune-Inflammation Index (SII, calculated as [neutrophil count × platelet count]/lymphocyte count), Systemic Inflammation Response Index (SIRI, [neutrophil count × monocyte count]/lymphocyte count), Aggregate Index of Systemic Inflammation (AISI, [neutrophil count × platelet count × monocyte count]/lymphocyte count), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), and neutrophil-to-albumin ratio (NAR) (Table S1). All cellular parameters (neutrophils, lymphocytes, monocytes, platelets) were derived from complete blood count (CBC) analyses, while albumin levels were measured via standardized biochemical assays (Supplementary Appendix Table S3).

Data collection

We collected comprehensive data across five domains: (A) Demographics: age, sex, race, education level, marital status, and poverty-to-income ratio; (B) Physical Examinations: Anthropometrics: Body mass index (BMI), waist circumference, standing height; (C) Blood pressure: systolic (SBP) and diastolic (DBP) blood pressure, calculated as the mean of three consecutive measurements; (D) Laboratory Tests: renal function: Estimated glomerular filtration rate (eGFR; calculated via CKD-EPI 2021 equation); (E) Metabolic markers: fasting blood glucose (FBG), glycated hemoglobin (HbA1c), triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C); (F) Lifestyle & Medical History: behaviors: smoking status, alcohol consumption, physical activity (≥ 150 min/week moderate-intensity or ≥ 75 min/week vigorous-intensity). (G) Medications: Use of antihypertensive, antihyperglycemic, antihyperlipidemic therapies; (H) Diagnoses: Hypertension (SBP ≥ 130 mmHg, DBP ≥ 80 mmHg, self-reported diagnosis, or antihypertensive use[11]), diabetes (FBG ≥ 126 mg/dL, HbA1c ≥ 6.5%, diagnosis, or glucose-lowering therapy), liver disease, cancer. Metabolic syndrome (MeTS): Presence of ≥ 3 criteria: abdominal obesity (waist circumference ≥ 102 cm [male] or 88 cm [female]), HDL-C < 40 mg/dL [male] or < 50 mg/dL [female], TG ≥ 150 mg/dL, elevated blood pressure (as defined above), FBG ≥ 100 mg/dL.

Statistical methods

Continuous variables are presented as mean ± standard deviation (SD), and categorical variables as counts (percentages). Group comparisons between survivors and non-survivors (all-cause/cardiovascular mortality) were performed using one-way ANOVA for continuous variables and chi-square tests for categorical variables.

To investigate non-linear relationships between inflammatory markers and mortality, we applied generalized additive models (GAMs) with penalized splines, generating smoothed curves to visualize potential threshold effects. For multivariable-adjusted analyses, Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for mortality outcomes. Missing covariate data were addressed through missing-indicator methods (dummy variables)[12].

Given the issue of collinearity, we checked the variance inflation factor of the covariates so that only covariates with VIF < 10 were included in the model. Variables altering the exposure-outcome association by > 10% upon inclusion/exclusion were retained[13]. We adjusted for the following covariates, taking into account their clinical significance: Demographics: age, gender, race, education level, marital status, poverty income ratio; Behaviors: smoking status, alcohol consumption (drinks/year); Physical Examinations: SBP, DBP; Laboratory Tests: Renal function: eGFR, CR, URCA; Metabolic markers: HBA1c, FBS,TG, LDL-C, albumin; Hematologic markers: HB, RDW; Comorbidities and Diagnoses: anemia, stroke, cancer, hyperlipidemia, hypertension, diabetes, MeTS; Medications: use of antihypertensive agents, antihyperglycemic agents; Syndrome Staging/Risk: CKD Risk, CKM stage.

To account for multiple comparisons across the seven inflammatory indices and two primary outcomes (all-cause and cardiovascular mortality), a Bonferroni correction was applied. This set the significance threshold for the primary analyses at a two-sided p < 0.00357 (0.05/14 tests).

Threshold effect analysis

First, to identify the inflection points (breakpoints) in these relationships, we applied a two-piecewise linear regression model. The optimal breakpoint was determined by moving a trial turning point across a predefined range and selecting the value that maximized the model likelihood. The significance of the nonlinearity was tested using a log-likelihood ratio test, comparing the two-piecewise model against a simple linear model.

Second, to visualize survival differences, we constructed Kaplan–Meier curves after stratifying participants into groups based on tertiles of MLR, NLR, and SIRI. The log-rank test was used to assess the statistical significance of differences between survival curves.

Third, to establish clinically relevant prognostic cut-offs, we performed time-dependent receiver operating characteristic (ROC) analysis at the 93-month follow-up mark. The optimal cut-off values were derived via bootstrap resampling (500 iterations), defined by Youden's index (maximizing the sum of sensitivity and specificity)[14].

Sensitivity analyses

To ensure the robustness of findings, sensitivity analyses were performed. Missing covariate data (> 5%) were managed with dummy variables and appropriate imputation methods to minimize bias, and methods to handle missing data were used to minimize potential bias. E-values were also calculated to evaluate the potential impact of unmeasured confounding. The E-value indicated the strength of an unmeasured confounder needed to negate the observed association between vitamin D levels and mortality. In this study, a two-tailed p-value of less than 0.05 was considered statistically significant. All analyses were conducted using EmpowerStats (www.empowerstats.com, X&Y Solutions, Inc., Boston, MA) and R software (version 4.4.3, http://www.r-project.org).

Results

Baseline characteristics

The study included 26,265 adults with CKM syndrome stages 0–4 from the NHANES database (2007–2016), with a median follow-up of 93.4 months. Overall, 8.8% (2,292) experienced all-cause mortality and 2.7% (701) cardiovascular mortality. The cohort had a mean age of 48.6 ± 17.9 years, with significantly older individuals in mortality groups (all-cause: 69.4 ± 12.5; cardiovascular: 71.4 ± 11.4; p < 0.001). Male (48.5% overall) and Non-Hispanic White individuals (41.6%) were disproportionately represented in mortality outcomes (both p < 0.001). Key comorbidities, including hypertension (64.8% vs. 32.2%), diabetes (27.9% vs. 11.0%), cardiovascular disease (27.5% vs. 6.0%), and advanced CKM stages (stage 3: 48.1% and stage 4: 14.4% all-cause deaths), were significantly more prevalent in non-survivors (all p < 0.001). Biochemical profiles showed elevated fasting glucose, HbA1c, and inflammatory markers (e.g., neutrophil-to-lymphocyte ratio) alongside reduced eGFR in mortality groups (all p < 0.001) (Table 1, S4). Table S5 showed the baseline characteristics stratified by MLR tertile. In the dataset of 26,265 samples, LDLC (53.69%) and FBG (51.87%) exceeded the 5% missing threshold and were treated with dummy variables. Other variables with missing values included: waist circumference (4.16%), poverty income ratio (4.06%), SBP/DBP (3.44% each), TG (1.6%), creatinine (1.54%), eGFR (1.54%), albumin (1.53%), smoker status (1.54%), URCA (1.18%), BMI (0.6%), standing height (0.46%), and HBA1C (0.26%). All variables below 5% missingness were retained without imputation to minimize bias.​

Table 1.

Baseline characteristics of participants with CKM syndrome by mortality status: NHANES 2007–2016

		All-cause mortality				Cardiovascular mortality
		Survivor	Non-survivor	p-value	Survivor		Non-survivor	p-value
N	Mean + SD	23973	2292		25564		701
Age, years	48.6 ± 17.9	46.6 ± 17.0	69.4 ± 12.5	< 0.001	47.9 ± 17.6		71.4 ± 11.4	< 0.001
Gender, n (%)				< 0.001				< 0.001
Male	12739 (48.5%)	11447 (47.7%)	1292 (56.4%)		12337 (48.3%)		402 (57.3%)
Female	13526 (51.5%)	12526 (52.3%)	1000 (43.6%)		13227 (51.7%)		299 (42.7%)
Race, n (%)				< 0.001				< 0.001
Mexican American	4187 (15.9%)	3998 (16.7%)	189 (8.2%)		4140 (16.2%)		47 (6.7%)
Other Hispanic	2884 (11.0%)	2725 (11.4%)	159 (6.9%)		2834 (11.1%)		50 (7.1%)
Non-Hispanic White	10938 (41.6%)	9539 (39.8%)	1399 (61.0%)		10504 (41.1%)		434 (61.9%)
Non-Hispanic Black	5382 (20.5%)	4946 (20.6%)	436 (19.0%)		5238 (20.5%)		144 (20.5%)
Other Race – Including Multi-Racial	1018 (3.9%)	948 (4.0%)	70 (3.1%)		1000 (3.9%)		18 (2.6%)
Non-Hispanic Asian	1856 (7.1%)	1817 (7.6%)	39 (1.7%)		1848 (7.2%)		8 (1.1%)
Education Level, n (%)				< 0.001				< 0.001
Less Than High School Grad	6516 (24.8%)	5706 (23.8%)	810 (35.3%)		6257 (24.5%)		259 (36.9%)
High School Grad/GED or Equivalent	5787 (22.0%)	5206 (21.7%)	581 (25.3%)		5605 (21.9%)		182 (26.0%)
Some College or above	13336 (50.8%)	12442 (51.9%)	894 (39.0%)		13077 (51.2%)		259 (36.9%)
Marital Status, n (%)				< 0.001				< 0.001
Married or living with partner	15392 (58.6%)	14239 (59.4%)	1153 (50.3%)		15057 (58.9%)		335 (47.8%)
Separated or never married	10260 (39.1%)	9125 (38.1%)	1135 (49.5%)		9894 (38.7%)		366 (52.2%)
Poverty income ratio, n (%)				< 0.001				< 0.001
≤ 1.30	9005 (34.3%)	8174 (34.1%)	831 (36.3%)		8772 (34.3%)		233 (33.2%)
1.3 ~ 1.85	3603 (13.7%)	3204 (13.4%)	399 (17.4%)		3466 (13.6%)		137 (19.5%)
> 1.85	12033 (45.8%)	11113 (46.4%)	920 (40.1%)		11741 (45.9%)		292 (41.7%)
Smoking Status, n (%)				< 0.001				< 0.001
Never	14491 (55.2%)	13585 (56.7%)	906 (39.5%)		14168 (55.4%)		323 (46.1%)
Current smoker	5281 (20.1%)	4825 (20.1%)	456 (19.9%)		5169 (20.2%)		112 (16.0%)
Ever smoker	6088 (23.2%)	5163 (21.5%)	925 (40.4%)		5822 (22.8%)		266 (37.9%)
Alcohol Consumption, drinks/year	24.6 ± 311.2	22.9 ± 203.1	42.1 ± 823.7	0.005	24.4 ± 306.0		30.4 ± 462.6	0.617
SBP, mmHg	123.6 ± 18.1	122.5 ± 17.3	134.5 ± 22.6	< 0.001	123.2 ± 17.9		136.2 ± 22.4	< 0.001
DBP, mmHg	69.8 ± 12.8	70.1 ± 12.4	66.0 ± 15.4	< 0.001	69.9 ± 12.6		64.9 ± 16.0	< 0.001
TG, mg/dL	154.8 ± 125.4	154.6 ± 127.2	157.2 ± 104.8	0.343	154.8 ± 126.1		154.9 ± 96.6	0.975
LDL-C, mg/dL	113.6 ± 35.4	114.3 ± 35.1	106.2 ± 37.7	< 0.001	113.8 ± 35.3		105.5 ± 38.8	< 0.001
FBS, mg/dL	109.3 ± 36.6	107.9 ± 34.7	123.3 ± 50.6	< 0.001	108.9 ± 36.2		123.5 ± 46.0	< 0.001
Albumin, g/L	42.6 ± 3.4	42.8 ± 3.3	41.2 ± 3.4	< 0.001	42.7 ± 3.4		41.1 ± 3.4	< 0.001
CR, mg/dL	0.9 ± 0.4	0.9 ± 0.4	1.1 ± 0.7	< 0.001	0.9 ± 0.4		1.2 ± 0.9	< 0.001
UACR, mg/g	33.3 ± 298.7	28.4 ± 275.3	86.5 ± 482.1	< 0.001	31.5 ± 291.0		99.5 ± 509.2	< 0.001
HBA1C, %	5.8 ± 1.1	5.7 ± 1.1	6.2 ± 1.3	< 0.001	5.7 ± 1.1		6.2 ± 1.4	< 0.001
eGFR, ml/min/1.73m2	95.7 ± 22.7	97.8 ± 21.4	73.1 ± 23.7	< 0.001	96.4 ± 22.2		69.1 ± 22.8	< 0.001
HB, g/L	14.0 ± 1.5	14.1 ± 1.5	13.7 ± 1.7	< 0.001	14.1 ± 1.5		13.6 ± 1.6	< 0.001
RDW, %	13.3 ± 1.3	13.2 ± 1.3	13.7 ± 1.6	< 0.001	13.3 ± 1.3		13.7 ± 1.5	< 0.001
MLR	0.28 ± 0.12	0.27 ± 0.11	0.35 ± 0.17	< 0.001	0.27 ± 0.12		0.37 ± 0.19	< 0.001
SIRI	1.21 ± 0.84	1.17 ± 0.80	1.61 ± 1.13	< 0.001	1.20 ± 0.82		1.73 ± 1.21	< 0.001
NLR	2.15 ± 1.09	2.10 ± 1.03	2.68 ± 1.50	< 0.001	2.13 ± 1.07		2.80 ± 1.47	< 0.001
PLR	123.51 ± 48.09	122.31 ± 45.98	136.04 ± 64.98	< 0.001	123.09 ± 47.50		138.67 ± 64.34	< 0.001
AISI	298.35 ± 231.80	291.19 ± 222.43	373.21 ± 303.51	< 0.001	295.48 ± 227.50		403.12 ± 338.48	< 0.001
NAR	0.10 ± 0.04	0.10 ± 0.04	0.11 ± 0.04	< 0.001	0.10 ± 0.04		0.11 ± 0.04	< 0.001
SII	524.08 ± 303.72	515.55 ± 292.11	613.25 ± 394.87	< 0.001	520.87 ± 299.91		641.08 ± 402.81	< 0.001
Stroke, n (%)	904 (3.4%)	611 (2.5%)	293 (12.8%)	< 0.001	801 (3.1%)		103 (14.7%)	< 0.001
Cancer, n (%)	2317 (8.8%)	1789 (7.5%)	528 (23.0%)	< 0.001	2180 (8.5%)		137 (19.5%)	< 0.001
Anemia, n (%)	1098 (4.2%)	901 (3.8%)	197 (8.6%)	< 0.001	1030 (4.0%)		68 (9.7%)	< 0.001
Hyperlipidemia, n (%)	8547 (32.5%)	7447 (31.1%)	1100 (48.0%)	< 0.001	8182 (32.0%)		365 (52.1%)	< 0.001
Cardiovascular disease, n (%)	2070 (7.9%)	1439 (6.0%)	631 (27.5%)	< 0.001	1815 (7.1%)		255 (36.4%)	< 0.001
Hypertension, n (%)	9214 (35.1%)	7728 (32.2%)	1486 (64.8%)	< 0.001	8727 (34.1%)		487 (69.5%)	< 0.001
Diabetes, n (%)	3286 (12.5%)	2647 (11.0%)	639 (27.9%)	< 0.001	3067 (12.0%)		219 (31.2%)	< 0.001
MeTS, n (%)	10107 (38.5%)	8939 (37.3%)	1168 (51.0%)	< 0.001	9722 (38.0%)		385 (54.9%)	< 0.001
CKD Risk, n (%)				< 0.001				< 0.001
Low-risk	22774 (86.7%)	21306 (88.9%)	1468 (64.0%)		22364 (87.5%)		410 (58.5%)
Moderate to high-risk	3101 (11.8%)	2438 (10.2%)	663 (28.9%)		2860 (11.2%)		241 (34.4%)
Very high-risk	390 (1.5%)	229 (1.0%)	161 (7.0%)		340 (1.3%)		50 (7.1%)
Antihypertensive agents, n (%)	7988 (30.4%)	6576 (27.4%)	1412 (61.6%)	< 0.001	7518 (29.4%)		470 (67.0%)	< 0.001
Antihyperlipidemic agents, n (%)	6178 (23.5%)	5209 (21.7%)	969 (42.3%)	< 0.001	5858 (22.9%)		320 (45.6%)	< 0.001
Antihyperglycemic agents, n (%)	2391 (72.8%)	1952 (73.7%)	439 (68.7%)	0.010	2233 (72.8%)		158 (72.1%)	0.832
CKM Stage, n (%)				< 0.001				< 0.001
Stage 0	5181 (19.7%)	5031 (21.0%)	150 (6.5%)		5162 (20.2%)		19 (2.7%)
Stage 1	8514 (32.4%)	8255 (34.4%)	259 (11.3%)		8452 (33.1%)		62 (8.8%)
Stage 2	9223 (35.1%)	8773 (36.6%)	450 (19.6%)		9113 (35.6%)		110 (15.7%)
Stage 3	1159 (4.4%)	57 (0.2%)	1102 (48.1%)		760 (3.0%)		399 (56.9%)
Stage 4	2188 (8.3%)	1857 (7.7%)	331 (14.4%)		2077 (8.1%)		111 (15.8%)

Abbreviations: HR, hazard ratio; CI, confidence interval; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; FBG, fasting blood glucose; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride; UACR, urinary albumin creatinine ratio; HB, hemoglobin; RDW, red cell distribution width; CKD, Chronic kidney disease; MeTS, metabolic syndrome; CKM, Cardiovascular-Kidney-Metabolic Syndrome. Frequencies are expressed as absolute numbers and percentages (%); values are means (standard deviation). In the dataset of 26,265 samples, LDLC (53.69%) and FBG (51.87%) exceeded the 5% missing threshold and were treated with dummy variables. Other variables with missing values included: waist circumference (4.16%), poverty income ratio (4.06%), SBP/DBP (3.44% each), TG (1.6%), creatinine (1.54%), eGFR (1.54%), albumin (1.53%), smoker status (1.54%), URCA (1.18%), BMI (0.6%), standing height (0.46%), and HBA1C (0.26%)

Association inflammatory indices and mortality

A nonlinear, J-shaped relationship was observed between inflammatory indices and both all-cause (Fig. 2) and cardiovascular mortality (Fig. 3). In the Cox regression models for cardiovascular-kidney-metabolic (CKM) syndrome, SIRI (all-cause mortality HR = 1.11, 95% CI = 1.06, 1.15, p < 0.0001; cardiovascular mortality HR = 1.18, 95% CI = 1.10–1.27, p < 0.0001), NLR (all-cause mortality HR = 1.08, 95% CI = 1.05–1.12, p < 0.0001; cardiovascular mortality HR = 1.11, 95% CI = 1.05–1.17, p < 0.0001), and MLR (all-cause mortality HR = 2.27, 95% CI = 1.71–3.01, p < 0.0001; cardiovascular mortality HR = 3.37, 95% CI = 2.09–5.44, p < 0.0001) showed significant associations with mortality in the ADJUST II model. In the tertile analysis, high tertiles of SIRI (all-cause HR = 1.20, p = 0.0047; cardiovascular HR = 1.35, p = 0.0138), NLR (all-cause HR = 1.23, p = 0.0007; cardiovascular HR = 1.42, p = 0.0021), and MLR (all-cause HR = 1.25, p = 0.0006; cardiovascular HR = 1.24, p = 0.0660) were significantly associated with increased mortality risk. Other indices like PLR, AISI, NAR, and SII did not show significant associations with mortality in the ADJUST II model or its tertile analysis (Table 2).

Fig. 2.

Smooth curve fitting of the association between Inflammatory Indices and all-cause mortality in patients with CKM syndrome stage 0–4 (A, MLR; B, SIRI; C, NLR; E, SII; F, AISI; G, NAR)

Fig. 3.

Smooth curve fitting of the association between Inflammatory Indices and cardiovascular mortality in patients with CKM syndrome stage 0–4(A, MLR; B, SIRI; C, NLR; E, SII; F, AISI; G, NAR)

Table 2.

Association of inflammatory indices with all-cause and cardiovascular mortality in patients with CKM syndrome: Multivariable cox regression analysis

Exposure	Non-adjusted	Adjust I	Adjust II
All-Cause Mortality
SIRI	1.44 (1.40, 1.48) < 0.0001	1.27 (1.23, 1.31) < 0.0001	1.11 (1.06, 1.15) < 0.0001 *
SIRI tertile
Low	Reference	Reference	Reference
Middle	1.38 (1.23, 1.56) < 0.0001	1.15 (1.02, 1.30) 0.0241	1.06 (0.93, 1.21) 0.3512
High	2.85 (2.56, 3.17) < 0.0001	1.71 (1.52, 1.92) < 0.0001	1.20 (1.06, 1.36) 0.0047
MLR	14.84 (12.85, 17.13) < 0.0001	4.28 (3.38, 5.42) < 0.0001	2.27 (1.71, 3.01) < 0.0001 *
MLR tertile
Low	Reference	Reference	Reference
Middle	1.30 (1.15, 1.47) < 0.0001	0.97 (0.86, 1.10) 0.6509	1.02 (0.89, 1.16) 0.8054
High	3.25 (2.92, 3.63) < 0.0001	1.44 (1.28, 1.62) < 0.0001	1.25 (1.10, 1.42) 0.0006 *
NLR	1.36 (1.31, 1.40) < 0.0001	1.23 (1.17, 1.28) < 0.0001	1.08(1.05,1.12) < 0.0001 *
NLR tertile
Low	Reference	Reference	Reference
Middle	1.26 (1.12, 1.42) 0.0001	1.19 (1.05, 1.34) 0.0048	1.15 (1.01, 1.30) 0.0323
High	2.51 (2.26, 2.79) < 0.0001	1.64 (1.47, 1.83) < 0.0001	1.23 (1.09, 1.38) 0.0007 *
PLR	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) 0.0132
PLR tertile
Low	Reference	Reference	Reference
Middle	0.86 (0.77, 0.95) 0.0045	0.89 (0.80, 0.99) 0.0362	0.98 (0.87, 1.10) 0.7216
High	1.22 (1.11, 1.35) < 0.0001	1.02 (0.93, 1.13) 0.6313	1.02 (0.91, 1.13) 0.7585
AISI	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) 0.0061
AISI tertile
Low	Reference	Reference	Reference
Middle	1.20 (1.08, 1.35) 0.0012	1.09 (0.97, 1.22) 0.1409	1.02 (0.90, 1.15) 0.7575
High	1.91 (1.73, 2.12) < 0.0001	1.49 (1.34, 1.66) < 0.0001	1.09 (0.97, 1.22) 0.1386
NAR	93.33 (40.86, 213.16) < 0.0001	409.40 (162.23, 1033.16) < 0.0001	2.16 (0.69, 6.74) 0.1862
NAR tertile
Low	Reference	Reference	Reference
Middle	1.26 (1.13, 1.41) < 0.0001	1.09 (0.98, 1.22) 0.1163	0.91 (0.81, 1.03) 0.1382
High	1.78 (1.61, 1.98) < 0.0001	1.69 (1.52, 1.88) < 0.0001	1.08 (0.96, 1.22) 0.2046
SII	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) 0.0089
SII tertile
Low	Reference	Reference	Reference
Middle	1.02 (0.91, 1.14) 0.7128	1.04 (0.93, 1.16) 0.4596	0.95 (0.85, 1.07) 0.4054
High	1.53 (1.38, 1.69) < 0.0001	1.39 (1.25, 1.54) < 0.0001	1.04 (0.94, 1.17) 0.4397
Cardiovascular Mortality
SIRI	1.50 (1.44, 1.56) < 0.0001	1.33 (1.26, 1.41) < 0.0001	1.18 (1.10, 1.27) < 0.0001 *
SIRI tertile
Low	Reference	Reference	Reference
Middle	1.54 (1.22, 1.93) 0.0002	1.27 (1.01, 1.61) 0.0430	1.09 (0.85, 1.40) 0.4963
High	3.55 (2.90, 4.35) < 0.0001	2.05 (1.65, 2.55) < 0.0001	1.35 (1.06, 1.71) 0.0138
MLR	18.14 (14.43, 22.79) < 0.0001	5.93 (4.05, 8.67) < 0.0001	3.37 (2.09, 5.44) < 0.0001 *
MLR tertile
Low	Reference	Reference	Reference
Middle	1.12 (0.88, 1.43) 0.3406	0.82 (0.64, 1.04) 0.0987	0.82 (0.63, 1.06) 0.1310
High	3.72 (3.05, 4.53) < 0.0001	1.50 (1.22, 1.86) 0.0002	1.24 (0.99, 1.57) 0.0660
NLR	1.36 (1.31, 1.40) < 0.0001	1.23 (1.17, 1.28) < 0.0001	1.11 (1.05, 1.17) 0.0003 *
NLR tertile
Low	Reference	Reference	Reference
Middle	1.40 (1.12, 1.76) 0.0037	1.34 (1.06, 1.68) 0.0140	1.19 (0.93, 1.52) 0.1683
High	3.24 (2.65, 3.96) < 0.0001	2.05 (1.66, 2.53) < 0.0001	1.42 (1.14, 1.78) 0.0021
PLR	1.00 (1.00, 1.01) < 0.0001	1.00 (1.00, 1.00) 0.0012	1.00 (1.00, 1.00) 0.0478
PLR tertile
Low	Reference	Reference	Reference
Middle	0.91 (0.75, 1.11) 0.3670	0.96 (0.78, 1.17) 0.6584	1.01 (0.82, 1.26) 0.8964
High	1.37 (1.14, 1.63) 0.0006	1.13 (0.94, 1.35) 0.1971	1.11 (0.92, 1.36) 0.2820
AISI	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) < 0.0001 *
AISI tertile
Low	Reference	Reference	Reference
Middle	1.36 (1.10, 1.68) 0.0040	1.24 (1.01, 1.54) 0.0440	1.03 (0.82, 1.30) 0.7754
High	2.32 (1.92, 2.81) < 0.0001	1.79 (1.46, 2.18) < 0.0001	1.23 (1.00, 1.53) 0.0529
NAR	297.19 (73.22, 1206.29) < 0.0001	2063.92 (418.62, 10,175.70) < 0.0001	13.86 (1.90, 101.04) 0.0095
NAR tertile
Low	Reference	Reference	Reference
Middle	1.38 (1.12, 1.69) 0.0026	1.18 (0.95, 1.45) 0.1267	0.90 (0.72, 1.13) 0.3600
High	2.07 (1.71, 2.51) < 0.0001	1.97 (1.62, 2.39) < 0.0001	1.21 (0.97, 1.51) 0.0858
SII	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) < 0.0001	1.00 (1.00, 1.00) 0.0030 *
SII tertile
Low	Reference	Reference	Reference
Middle	1.11 (0.91, 1.36) 0.3059	1.16 (0.94, 1.42) 0.1644	1.04 (0.84, 1.28) 0.7381
High	1.78 (1.48, 2.14) < 0.0001	1.62 (1.34, 1.95) < 0.0001	1.18 (0.97, 1.43) 0.1072

Data were presented as HR (95%Cl), p value; *p values in boldface indicate statistical significance after Bonferroni correction for 14 tests (p < 0.00357)

Non-adjusted model adjust for: None

Adjust I model adjust for: age; gender; race

Adjust II model adjust for: age, gender, race, education level, marital status, poverty income ratio; smoking status; alcohol consumption; SBP; DBP; eGFR; CR; URCA; HBA1c; FBS;TG; LDL-C; albumin; HB; RDW; anemia; stroke; cancer; hyperlipidemia; hypertension; diabetes; MeTS; antihypertensive agents; antihyperglycemic agents; CKD Risk; CKM stage

Restrict cubic spline smoothing only applies for continuous variables

Abbreviations: HR, hazard ratio; CI, confidence interval; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; CR, creatinine; eGFR, estimated glomerular filtration rate; FBG, fasting blood glucose; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride; UACR, urinary albumin creatinine ratio; HB, hemoglobin; RDW, red cell distribution width; CKD, Chronic kidney disease; DM, diabetes mellitus; MeTS, metabolic syndrome; CKM, Cardiovascular-Kidney-Metabolic Syndrome

Threshold effect

In the threshold effect analysis of inflammatory indices in CKM syndrome, NLR showed a breakpoint at 3 (95% CI 2.57–3.77), with a significant effect below this value (HR = 1.14, p = 0.0002) and a weaker effect above (HR = 1.07, p = 0.0037). MLR had a breakpoint at 0.19 (95% CI 0.17–0.21), showing a significant effect above this value (HR = 2.59, p < 0.0001). SIRI demonstrated breakpoints at 1.74 (95% CI 1.45–2.30) for all-cause mortality and 0.38 (95% CI 0.37–0.54) for cardiovascular mortality, with significant effects above these thresholds (HR = 1.09, p = 0.0020 for all-cause; HR = 1.17, p < 0.0001 for cardiovascular) (Table 3).

Table 3.

Threshold effect analysis of inflammatory indices in CKM syndrome

NLR Outcome:	All-cause mortality	Cardiovascular mortality
Model I
One line effect	1.09 (1.06, 1.12) < 0.0001	1.11 (1.05, 1.17) < 0.0001
Model II
Turning point (K)	3	2.94
< K	1.14 (1.06, 1.22) 0.0002	1.28 (1.12, 1.46) 0.0002
≥ K	1.07 (1.02, 1.12) 0.0037	1.04 (0.96, 1.12) 0.3336
P value for LRT test	0.168	0.017
95% CI for turning point	2.57, 3.77	2.56, 3.67
MLR Outcome:	All-cause mortality	Cardiovascular mortality
Model I
One line effect	2.41 (1.85, 3.14) < 0.0001	3.34 (2.13, 5.23) < 0.0001
Model II
Turning point (K)	0.19	0.21
< K	0.27 (0.02, 3.19) 0.2994	0.15 (0.00, 4.93) 0.2841
≥ K	2.59 (1.97, 3.40) < 0.0001	3.81 (2.40, 6.06) < 0.0001
P value for LRT test	0.089	0.091
95% CI for turning point	0.17, 0.21	0.19, 0.23
SIRI Outcome:	All-cause mortality	Cardiovascular mortality
Model I
One line effect	1.12 (1.07, 1.16) < 0.0001	1.18 (1.10, 1.26) < 0.0001
Model II
Turning point (K)	1.74	0.38
< K	1.19 (1.07, 1.32) 0.0018	364.33 (0.72, 183,547.95) 0.0632
≥ K	1.09 (1.03, 1.15) 0.0020	1.17 (1.09, 1.25) < 0.0001
P value for LRT test	0.231	0.030
95% CI for turning point	1.45, 2.3	0.37, 0.54

Data were presented as HR (95%Cl), p value; Model I, linear analysis; Model ll, non-linear analysis. Adjust for: age, gender, race, education level, marital status, poverty income ratio; smoking status; alcohol consumption; SBP; DBP; eGFR; CR; URCA; HBA1c; FBS;TG; LDL-C; albumin; HB; RDW; anemia, stroke, cancer, hyperlipidemia, hypertension, diabetes, MeTS; antihypertensive agents; antihyperglycemic agents; CKD Risk; CKM stage

Abbreviations: HR, hazard ratio; Cl, confidence interval; LRT, logarithm likelihood ratio test; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; CR, creatinine; eGFR, estimated glomerular filtration rate; FBG, fasting blood glucose; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride; UACR, urinary albumin creatinine ratio; HB, hemoglobin; RDW, red cell distribution width; CKD, Chronic kidney disease; DM, diabetes mellitus; MeTS, metabolic syndrome; CKM, Cardiovascular-Kidney-Metabolic Syndrome

Kaplan–Meier curves were utilized to evaluate the survival differences in terms of all-cause mortality. For MLR tertile, the survival curves showed no significant difference between the Low and Middle groups (p = 0.5687), but the survival of the High group was significantly lower than the Low group (p < 0.0001). For NLR tertile, the survival of the High group (p < 0.0001) and the Middle group (p = 0.034) were significantly lower than the Low group. For SIRI tertile, the survival curves showed no significant difference the between Low and Middle groups (p = 0.3390), but the survival of the High group was significantly lower than the Low group (p = 0.0002) (Fig. 4). Detailed analyses of cardiovascular mortality (Fig. S1) and all—cause mortality with threshold points (Fig. S2) are provided in the supplementary materials.

Fig. 4.

Kaplan–Meier survival curves for participants with inflammatory indices and All-cause mortality (A, MLR; B, NLR; C, SIRI). Abbreviations: MLR, Monocyte-to-lymphocyte ratio; NLR, Neutrophil-to-lymphocyte Ratio; SIRI, Systemic Inflammation Response Index

Time-dependent ROC and harrell’s C-index analysis

ROC curves of Inflammatory Indices were used to determine the optimal cut-off values for prediction of events during follow-up by using a bootstrap resampling (times = 500) showed an area under curve (Fig. 5). Time-dependent ROC curve analysis evaluated inflammatory indices for predicting all-cause mortality at 93 months. The MLR demonstrated the highest discriminative performance, with an AUC of 0.630 and Harrell’s C-index of 0.6665 (P < 0.001), achieving balanced sensitivity (55.0%) and specificity (66.6%) at its optimal cutoff (0.29268). SIRI (AUC = 0.611, C-index = 0.6434) and NLR (AUC = 0.602, C-index = 0.6326) followed, with SIRI showing higher sensitivity (52.6% vs. 46.6%) but lower specificity (65.9% vs. 70.4%) compared to NLR. Other indices exhibited weaker predictive capacity: PLR (AUC = 0.521, C-index = 0.5394), SII (AUC = 0.547, C-index = 0.5644), and NAR (AUC = 0.576, C-index = 0.5771), primarily due to low sensitivity (e.g., PLR: 28.1%) or specificity (e.g., NAR: 52.0%). Optimal cutoffs were determined by maximizing sensitivity–specificity sums, with detailed values provided in ROC output files (Table 4).

Fig. 5.

ROC curves for Inflammatory Indices, using Bootstrap resampling (times = 500) for the association of all-cause mortality

Table 4.

Time-Dependent ROC Curve Analysis for Inflammatory Indices Predicting All-Cause Mortality

Test	Cut.value	Sensitivity	Specificity	AUC	C index
SIRI	1.24	0.52603	0.65872	0.61126	0.6434
NLR	2.375	0.46647	0.70424	0.60237	0.6326
PLR	148.66667	0.2805	0.77506	0.52058	0.5394
MLR	0.29268	0.55025	0.6655	0.62975	0.6665
AISI	281.4	0.50842	0.61186	0.56567	0.5877
NAR	0.09362	0.60582	0.52013	0.57596	0.5771
SII	622.85714	0.35277	0.7365	0.54655	0.5644

Data derived from time-dependent ROC analysis with 93-month follow-up. AUC (Area Under the Curve); C index: Harrell’s Concordance Index (0.5 = random; 1.0 = perfect prediction)

Stratified analysis

In subgroup analyses, age significantly modified the associations of MLR and SIRI with all-cause mortality (MLR: HR = 3.75 vs 2.42 for ≤ 60 vs > 60 years, interaction p = 0.0115; SIRI: HR = 1.11 vs 1.12, interaction p = 0.0310), with significant interaction p-values (< 0.05). For cardiovascular mortality, age interacted with SIRI and NLR (SIRI: HR = 1.34 vs 1.16, interaction p = 0.0084; NLR: HR = 1.22 vs 1.10, interaction p = 0.0380), both showing significant interactions (p < 0.05). Inflammatory markers demonstrated significant interaction effects with CKM stages on all-cause mortality (MLR: interaction p = 0.0009; SIRI: p = 0.0112; NLR: p = 0.0101), whereas no interactions were observed for cardiovascular mortality (MLR: p = 0.7466; SIRI: p = 0.6324; NLR: p = 0.4273). Stratified analyses indicated no significant interaction effects for most covariates, including gender, smoking status, and heart disease history, hypertension, MeTS, (all interaction P > 0.05) (Fig. 6 and 7).

Fig. 6.

Subgroup Analyses of Inflammatory indices concerning all-cause mortality in a CKM syndrome stage 0–4 population (A, MLR; B, NLR; C, SIRI). Abbreviations: MLR, Monocyte-to-lymphocyte ratio; NLR, Neutrophil-to-lymphocyte Ratio; SIRI, Systemic Inflammation Response Index

Fig. 7.

Subgroup Analyses of Inflammatory indices concerning cardiovascular mortality in a CKM syndrome stage 0–4 population (A, MLR; B, NLR; C, SIRI). Abbreviations: MLR, Monocyte-to-lymphocyte ratio; NLR, Neutrophil-to-lymphocyte Ratio; SIRI, Systemic Inflammation Response Index

Discussion

Our study investigated the relationship between inflammatory indices and mortality in a cohort of 26,265 participants from the NHANES database (2007–2016) with CKM syndrome stages 0–4. This study first discussed the relationships between inflammatory indices and mortality in patients with CKM syndrome, and identified systemic inflammatory markers SIRI, NLR, and MLR as significant predictors of mortality in CKM syndrome. Most inflammation indices showed a nonlinear, J- shaped relationship with mortality. Threshold effects showed that mortality risk increased significantly above SIRI > 1.74 (all-cause), NLR > 3.0, and MLR > 0.19. Stronger predictive power observed in younger patients (≤ 60 years) and early CKM stages (1–2)(Fig. 8).

Fig. 8.

Graphic abstract

Inflammation has been defined as an evolutionarily conserved defense mechanism that protects against pathogens and promotes tissue repair [15, 16].

However, systemic chronic inflammation fundamentally differs from acute inflammation. While the latter represents a transient, localized response to immediate threats such as infection or injury, the former constitutes a maladaptive, low-grade state driven by persistent social, environmental, and metabolic stressors. This state is characterized by non-resolving immune activation and homeostatic disruption[17]. The present study hypothesizes that this chronic inflammatory process, devoid of overt infectious triggers, contributes to tissue damage and is implicated in various chronic diseases and premature mortality, including CVD [18, 19], CKD [20, 21], obese and DM [20, 22], and premature mortality[16].

The present study contributes to the advancement of this field by conducting a comprehensive, multi-index comparative analysis. A panel of seven inflammatory indices was evaluated, which can be broadly categorized into those reflecting the myeloid-lymphoid axis (SIRI, MLR, NLR) and those incorporating the thrombotic-nutritional axis (SII, AISI, PLR, NAR). The consistent outperformance of the myeloid-lymphoid indices in predicting CKM mortality provides a novel hematological perspective on the syndrome's drivers. This discrepancy is likely attributable to the fact that platelet-based markers (SII, AISI, PLR), while valuable in acute thrombotic or oncologic settings, may be less specific in chronic CKM syndrome. This is due to the fact that platelet counts are influenced by numerous confounders, such as medication, which potentially dilutes their signal.

The superior prognostic performance of the myeloid-lymphoid indices can be attributed to their direct reflection of the core inflammatory drivers in CKM pathophysiology. CKM syndrome is fundamentally characterized by a state of persistent, low-grade immune activation, which in turn promotes end-organ damage[1, 3]. The myeloid-lymphoid indices have been shown to precisely quantify this dysregulated crosstalk: monocytes and neutrophils are primary sources of pro-inflammatory cytokines (e.g., IL-6, TNF-α) and key effectors in the pathogenesis of vascular endothelial dysfunction, renal fibrosis, and adipose tissue inflammation [1, 18]. Concurrently, lymphopenia, reflected in these ratios, signifies a state of chronic immune stimulation and potential immunosuppression. The robust association of MLR, SIRI, and NLR with mortality, which is particularly pronounced in younger individuals and early CKM stages, highlights their significant clinical potential. These readily available biomarkers could serve as sensitive tools for early risk stratification, identifying a subgroup of patients who might derive the greatest benefit from aggressive management of inflammation and CKM risk factors, potentially altering the disease trajectory before the establishment of advanced, irreversible complications. It is important to note that the associations remained consistent across other major subgroups defined by sex, diabetes status, and hypertension history, as no significant interactions were found.

Clinical validations highlight these differential associations. In CKD cohorts, Gu et al. (n = 3,262, NHANES) demonstrated superiority of the SIRI over the NLR and PLR as an independent predictor of cardiovascular and all-cause mortality, exhibiting a nonlinear dose–response relationship with cardiovascular mortality (threshold effect at SIRI > 1.2)[7]. This finding was corroborated by Huang et al. (n = 40,937), which established significant associations of SIRI and systemic immune-inflammation index (SII) with CKD prevalence and mortality[7, 21].

In the context of coronary artery disease, Vakhshoori et al. identified the monocyte-to-lymphocyte ratio (MLR ≥ 0.34) as a robust predictor of major adverse cardiovascular events (HR = 1.87), a finding that was particularly evident in acute coronary syndrome subgroups [23]. Parallel findings emerged in metabolic disorders. Elevated MLR independently predicted 90-day all-cause mortality in type 2 diabetes mellitus (T2DM) patients with CKD[24]. Jing et al. showed that the MLR was significantly higher in patients with diabetic retinopathy[25]. And, in a cohort study involving 29,459 patients (≥ 20 years), Cao et al. emphasized the clinical relevance of integrating CKM staging with SIRI for mortality prediction[26]. Notably, inflammatory indices retain predictive value even in ostensibly healthy populations. Yang et al. 's analysis of 35,813 NHANES general population (1999–2014) further showed that individuals in the highest MLR tertile faced significantly increased risks of all-cause and CVD mortality [27]. Imtiaz et al. demonstrated NLR's association with hypertension and diabetes mellitus risk in ostensibly healthy individuals (n = 1,070)[28]. The findings of this study suggest that, while SII, AISI, PLR, and NAR are useful inflammatory markers, MLR, SIRI, and NLR demonstrate a stronger correlation with all-cause and cardiovascular mortality in CKM patients. It has been hypothesized that MLR, SIRI and NLR may be more directly capable of capturing leukocyte-driven immune dysregulation, which is central to the pathogenesis of CKM's metabolic, cardiac and renal complications, whereas platelet-related indices may be more indicative of thrombotic risk than chronic immune activation.

The present study contributes to the advancement of this understanding by establishing specific, clinically applicable thresholds for these indices within the CKM population. The MLR breakpoint of 0.19 aligns with the upper limit of normal monocyte-lymphocyte homeostasis, suggesting that even a subtle shift towards a pro-inflammatory state within the normal range may be sufficient to elevate mortality risk in the susceptible CKM population [1, 3]. The NLR threshold of 3.0 is a well-documented critical value in cardiometabolic literature, marking a transition to a state of significant neutrophilia and relative lymphopenia [18, 28].

For SIRI, the breakpoint for all-cause mortality (1.74) is consistent with thresholds reported in prior studies of chronic conditions and constitutes our primary, stable finding [7, 21]. However, the breakpoint for cardiovascular mortality (0.38) exhibited a wider confidence interval, a phenomenon that is likely attributable to the reduced number of cause-specific events and the consequent reduction in estimation precision. Consequently, while the precise Siri threshold for cardiovascular mortality should be interpreted with caution, the consistent J-shaped relationship across indices provides robust confirmation of the existence of a risk transition point.

The consistent J-shaped associations observed for MLR, SIRI, and NLR suggest a dual risk profile tied to immune imbalance. At the high end, risk is likely driven by excessive inflammation, where elevated monocytes and neutrophils promote end-organ damage through pro-inflammatory cytokines [1, 18]. Conversely, the increased risk observed at low index levels may be indicative of an impaired immune competence, with very low lymphocyte counts potentially signifying immune senescence or vulnerability to infections and other comorbidities [15, 16]. Consequently, the J-shaped relationship may indicate that individuals aged ≤ 60 years with both hyperactive and weakened immune states are associated with an elevated mortality risk in CKM syndrome. The stronger association of inflammatory indices with mortality in younger patients (≤ 60 years), which may appear counterintuitive in light of the concept of "inflammaging", can be explained by the presence of competing risks. In older populations, mortality is driven by a high burden of comorbidities and age-related frailty, which may serve to dilute the specific signal of inflammation-driven risk. Conversely, in younger individuals with CKM syndrome, elevated systemic inflammation is likely to be a primary and dominant driver of premature mortality, thereby rendering its relative effect more pronounced. It is important to note that the associations remained consistent across other major subgroups defined by sex, diabetes status, and hypertension history, as no significant interactions were found. This homogeneity enhances the general applicability of these inflammatory markers for risk assessment.

The underlying mechanisms by which these associations between systemic inflammation and mortality arise require further elucidation. SIRI, NLR, and MLR have been demonstrated to reflect distinct inflammatory pathways. Monocyte-lymphocyte ratio (MLR) is a key indicator of monocyte-driven chronic inflammation and immune activation through tissue fibrosis and pro-inflammatory cytokine production (e.g., IL-6, TNF-α). This has significant prognostic value in a number of contexts, including diabetic nephropathy [24], retinopathy[25], and general population CVD mortality[27]. NLR has been shown to quantify systemic inflammatory burden via neutrophil (acute-phase response) to lymphocyte (immune regulation) ratios, correlating with metabolic dysfunction and cardiovascular pathology[28]. SIRI has been demonstrated to integrate neutrophils, monocytes (which represent the innate immune system) and lymphocytes (which represent the adaptive immune system) through its formula (neutrophils x monocytes/lymphocytes), thus capturing both acute and chronic inflammatory components[7]. This may account for its superior predictive performance in CKM stages.

The present study makes a significant contribution to the advancement of this field by conducting the first multi-index comparative analysis, thereby establishing stage-specific thresholds for SIRI, NLR, and MLR in CKM syndrome populations. Stage-specific diagnostic thresholds were established for SIRI (> 1.2), NLR (> 2.8), and MLR (> 0.34), and a novel inflammatory framework for risk stratification was proposed.

The present study contributes to the advancement of this field by conducting the first multi-index comparative analysis, thereby establishing stage-specific thresholds for SIRI, NLR, and MLR in CKM syndrome populations. Stage-specific diagnostic thresholds were established for SIRI (> 1.2), NLR (> 2.8), and MLR (> 0.34), and a novel inflammatory framework for risk stratification was proposed.

It has been demonstrated by preceding studies that staging exerts a significant impact on all-cause mortality among patients diagnosed with CKM syndrome. Ji et al.'s analysis of 33,868 NHANES participants (1988–2018) established cardiovascular-kidney-metabolic (CKM) staging as a mortality determinant, revealing advanced stages (3–4) accounted for 62.5% of deaths with notable sex disparities (female: HR = 1.24–3.33 vs. male: 0.85–2.60) [29]. In the present study, Advanced CKM stages (stage 3: 48.1% and stage 4: 14.4% all-cause deaths) contributed to the majority of all-cause (stage 3 and 4: 62.5%) and cardiovascular deaths (stage 3 and 4: 72.7%). Furthermore, CKM stages demonstrated significant interactions with all inflammatory markers (MLR/SIRI/NLR), specifically with regard to all-cause mortality (all interaction p-values < 0.05). In contrast to the focus of Ji et al. on sex, we identified NLR, SIRI, and MLR as stage-dependent predictors, thereby demonstrating enhanced prognostic accuracy in early-stage patients and those ≤ 60 years.

While MLR demonstrated the optimal discriminatory performance among the indices evaluated, it is acknowledged that its predictive accuracy, as reflected by an AUC of 0.630 and a C-index of 0.667, remains modest for standalone risk prediction at the individual level. This level of accuracy is, however, comparable to that of many other novel biomarkers in complex chronic diseases, thereby underscoring the multifactorial nature of mortality in CKM syndrome, which cannot be fully captured by a single inflammatory marker. Consequently, the primary clinical utility of MLR may not lie in replacing established risk models, but in complementing them. It is proposed that future research should investigate the integration of easily obtainable inflammatory indices, such as the modified Glasgow ratio (MLR), into comprehensive coronary artery disease (CAD) risk assessment tools. The PREVENT equations are one such tool that could be utilized for this purpose [8]. The combination of conventional risk elements (e.g., age, blood pressure, cholesterol) with a biomarker of ongoing systemic inflammation has the potential to establish a more comprehensive risk classification framework. This framework can identify high-risk patients who would benefit from enhanced management of risk factors and the implementation of anti-inflammatory strategies.

In the present study, the following limitations were observed: Firstly, the observational design precludes causal inference, with residual confounding likely to persist despite adjustments for major cardiometabolic risk factors. Secondly, the thresholds for inflammatory indices derived from this predominantly U.S.-based cohort may have limited generalizability when directly applied to non-U.S. populations. The influence of disparities in genetic background, dietary patterns, lifestyle, and healthcare systems across regions on baseline levels of inflammatory markers and their relationship with mortality is a subject that merits further investigation. Consequently, the validation of our proposed cut-off values in diverse ethnic and geographical populations is essential prior to their widespread clinical application. Thirdly, inflammatory indices are dependent on single-time point measurements, which may underestimate intraindividual biological variability. Fourthly, while the stage-specific thresholds employed are statistically robust, they lack histological validation against end-organ damage markers.

Conclusion

In this large observational study, it was demonstrated that systemic inflammatory indices exhibit nonlinear, J-shaped associations with mortality risk in individuals with CKM syndrome. The monocyte-to-lymphocyte ratio (MLR) was identified as the index most strongly associated with mortality across all disease trajectories. The findings of this study indicate that MLR, NLR, and SIRI demonstrate significant associations with all-cause mortality in CKM syndrome, suggesting their potential as useful prognostic indicators. It is of particular significance that these associations were more pronounced in younger patients (aged ≤ 60 years) and those in early CKM stages (1–2). This finding underscores the potential value of these indices for risk stratification in these subpopulations.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

MLR

Monocyte-to-lymphocyte ratio

NLR

Neutrophil-to-lymphocyte Ratio

SIRI

Systemic Inflammation Response Index

PLR

Platelet-to-Lymphocyte Ratio

AISI

Aggregate Index of Systemic Inflammatio

SII

Systemic Inflammation Index

NAR

Neutrophil-to-albumin ratio

CKM syndrome

Cardiovascular-kidney-metabolic syndrome

CKD

Chronic kidney disease

NHANES

National Health and Nutrition Examination Survey

PREVENT equation

Predicting risk of CVD events equation

KDIGO

Kidney Disease Improving Global Outcomes

BMI

Body mass index

CI

Confidence interval

HR

Hazard ratio

CVD

Cardiovascular disease

DM

Diabetes mellitus

eGFR

Estimated glomerular filtration rate

FBG

Fasting blood glucose

HbA1c

Hemoglobin A1c

HDL-C

High-density lipoprotein cholesterol

LDL-C

Low-density lipoprotein cholesterol

MetS

Metabolic syndrome

TC

Total cholesterol

TG

Triglyceride

CR

Creatinine

HB

Hemoglobin

RDW

Red cell distribution width

URCA

Urinary albumin creatinine ratio

Author contributions

Conceptualization, Yannv Qu and Yansun Sun; Funding acquisition, Yannv Qu; Investigation, Li Liu and Ling Wang; Methodology, Yannv Qu and Ling Wang; Writing – original draft, Yannv Qu and Ling Wang; Writing – review & editing, Li Liu and Yansun Sun. All authors warrant that they have reviewed, provided their consent, and approved the manuscript prior to submission.

Funding

Shenzhen Science and technology program: JCYJ20220531094006013, Guangdong Basic and Applied Basic Research Foundation: 2023A1515012394, Peking University Shenzhen Hospital—Ye Chenghai Charity Foundation. The manuscript editing and polishing was supported by the fundings.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval

Ethical approval for the NHANES was obtained from the National Center for Health Statistics Ethics Review Board. All participants provided written informed consent prior to enrollment. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki.

Consent to participate

Informed consent was obtained from all individual participants in the original NHANES study.

Clinical trial number

Not applicable.

Consent to publish

Written informed consent was obtained from the patient for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.

Conflicts of interest

The authors declare no competing interests.