Abstract
Background
Chronic diseases pose significant global health burdens, necessitating robust biomarkers for early detection. The remnant cholesterol inflammatory index (RCII), integrating lipid remnants and systemic inflammation, may predict chronic disease risk, but its longitudinal associations remain understudied across diverse populations.
Methods
We analyzed data from two prospective cohorts—the China Health and Retirement Longitudinal Study (CHARLS, N = 9,491) and the English Longitudinal Study of Ageing (ELSA, N = 6,054)—to evaluate associations between baseline RCII and new-onset chronic diseases. RCII was calculated as (remnant cholesterol × hsCRP)/10. Cox models estimated hazard ratios (HRs) for incident diseases, adjusting for sociodemographic, metabolic, and comorbidity covariates.
Results
In CHARLS, each unit increase in logarithm-transformed RCII (lnRCII) was associated with higher diabetes risk (HR = 1.09, 95% CI: 1.02–1.15). In ELSA, elevated lnRCII was associated with multiple outcomes, including diabetes (HR = 1.13, 95% CI: 1.01–1.27), stroke (HR = 1.26, 95% CI: 1.15–1.38), psychiatric disease (HR = 1.16, 95% CI: 1.03–1.30), asthma (HR = 1.27, 95% CI: 1.11–1.46), and COPD (HR = 1.36, 95% CI: 1.01–1.51). Associations for hypertension and heart disease were significant in unadjusted models but no longer significant after adjustment.
Conclusion
RCII shows a consistent association with incident diabetes across both Chinese and UK cohorts, whereas its relationships with other chronic diseases were observed only in the UK cohort. These findings suggest that RCII may serve as a robust marker for diabetes risk and a broader indicator of multisystem vulnerability in certain populations, warranting cautious interpretation and further validation.
Supplementary Information
The online version contains supplementary material available at 10.1186/s41043-026-01240-x.
Keywords: RCII, Chronic disease, ELSA, CHARLS
Introduction
Chronic diseases remain the leading cause of death and disability worldwide, imposing profound socioeconomic and healthcare burdens across diverse populations [1]. As aging accelerates globally, the incidence and prevalence of metabolic and cardiovascular disorders continue to rise, underscoring the urgent need for robust biomarkers to enable early detection and risk stratification [2].
Remnant cholesterol (RC), the cholesterol content within triglyceride-rich lipoproteins, has been implicated in atherogenesis and residual inflammatory risk beyond traditional lipid metrics [3]. Previous studies have demonstrated that elevated RC is associated with an increased long-term risk of all-cause and cardiovascular mortality, highlighting its potential as an early indicator of cardiometabolic disturbances [4]. Building on this, the remnant cholesterol inflammatory index (RCII) was developed as a composite measure integrating RC and systemic inflammation, aiming to capture synergistic pathophysiological mechanisms—including lipid dysmetabolism, endothelial dysfunction, and chronic low-grade inflammation—that contribute to the development of cardiovascular and metabolic disorders [5, 6]. Despite accumulating evidence linking RCII to adverse outcomes, its predictive capacity for new-onset chronic diseases, especially across heterogeneous populations and disease stages, remains inadequately characterized.
Clinical data from the China Health and Retirement Longitudinal Study (CHARLS) cohort indicate that elevated RCII levels, both at baseline and cumulatively, are significantly associated with an increased risk of stroke over a median follow-up period of 7 years [7]. Furthermore, higher RCII concentrations are robustly linked to elevated all-cause mortality risk among middle-aged and elderly populations in both CHARLS and the National Health and Nutrition Examination Survey (NHANES) datasets. In the U.S. population of NHANES dataset, RCII has also been implicated in heightened risks of cardiovascular and cancer-related mortality, underscoring its potential as a versatile prognostic biomarker across diverse chronic disease outcomes [5]. However, these studies often focus on advanced disease states or single-center cohorts, limiting generalizability and the understanding of RCII’s role in early disease phases or multi-morbidity progression.
The dynamic interplay between lipid remnants and inflammation is particularly relevant in the context of aging populations, where multimorbidity and subclinical organ impairment frequently precede overt chronic disease [8]. Beyond cardiometabolic disorders, chronic low-grade inflammation and metabolic dysregulation have been increasingly implicated in respiratory diseases, neuropsychiatric conditions, and other age-related chronic disorders, suggesting that RCII may capture a broader state of multisystem vulnerability rather than a single disease pathway. Leveraging large-scale, prospective cohorts from both China and the United Kingdom—the CHARLS and the English Longitudinal Study of Ageing (ELSA)—offers a unique opportunity to assess the prognostic utility of RCII across diverse ethnicities and healthcare contexts. This multicohort design enables comprehensive evaluation of RCII’s association with incident chronic diseases, including hypertension, diabetes, cardiovascular disease, and psychiatric diseases, while accounting for potential confounders in each cohort.
In this study, we systematically investigated the longitudinal relationship between baseline RCII levels and the risk of new-onset chronic diseases in middle-aged and older adults, with a specific focus on both cross-national applicability and disease-specific prediction. By integrating evidence from CHARLS and ELSA, we aim to determine which chronic diseases are independently associated with RCII in each cohort and to assess whether these associations are consistent across Chinese and UK populations. This approach allows us to validate RCII as a potential biomarker for early risk stratification and personalized preventive strategies, ultimately informing risk stratification and preventive strategies for aging-related multisystem health challenges, including but not limited to cardiometabolic diseases.
Materials and methods
Study design and participants
This study leveraged data from two large-scale, nationally representative cohorts: the CHARLS and ELSA. CHARLS is a longitudinal study of Chinese adults aged 45 years and older, designed to represent the national population (http://charls.pku.edu.cn/). The baseline sample was drawn from 450 communities across 150 counties or districts in 28 provinces, with follow-up waves conducted from 2011 to 2020 [9]. ELSA is a population-based cohort involving individuals aged 50 and above residing in private households across the United Kingdom. Baseline data were obtained in 2002–2003, with biennial follow-up assessments conducted via computer-assisted interviews and self-administered questionnaires. In addition, biological specimens and physical measurements were collected every four years during home visits by trained nurses [10].
For the CHARLS cohort, baseline data from 2011 were analyzed, with follow-up extending to 2018. The 2011 survey included 11,847 respondents with available RCII-related data; individuals without baseline information on remnant cholesterol, high-sensitivity C-reactive protein (hsCRP), chronic disease history, or with other missing covariates were excluded. For the ELSA cohort, data were drawn from wave 6, comprising 10,601 participants, and those lacking information on remnant cholesterol, hsCRP, chronic disease history, or other relevant covariates were removed. In both cohorts, analyses for each chronic disease were limited to participants who were free of the condition at baseline and had complete follow-up data. These exclusion criteria ensured data integrity and enhanced the robustness and validity of subsequent statistical analyses. The participant selection process for the CHARLS and ELSA studies is summarized in the flowchart shown in Fig. 1.
Fig. 1.
Flowchart of participant screening in CHARLS and ELSA
Definition of RCII
RC was calculated by subtracting the sum of HDL-C and low-density lipoprotein cholesterol (LDL-C) from total cholesterol (TC). Subsequently, the RCII was derived by multiplying RC by hsCRP (mg/L) and dividing by 10, as previously described by Chen et al. [7].
Chronic diseases assessment in both cohorts
Incident chronic diseases—including hypertension, diabetes mellitus, heart disease, stroke, cancer, chronic obstructive pulmonary disease (COPD), asthma, psychiatric disease, memory disease, and arthritis—were ascertained prospectively in both cohorts, with disease-specific definitions harmonized at the conceptual level and cohort-specific operational criteria explicitly described below. For all incident conditions, the estimated onset date was defined as the midpoint between the follow-up wave in which the disease was first reported and the immediately preceding follow-up assessment.
In both CHARLS and ELSA, diabetes mellitus was defined using consistent criteria. Incident diabetes was identified if participants met any of the following conditions during follow-up: fasting plasma glucose ≥ 126 mg/dL, hemoglobin A1c (HbA1c) ≥ 6.5%, self-reported physician diagnosis, or current use of antidiabetic medication [11]. Hypertension was defined similarly in both cohorts. Participants were classified as having hypertension if they reported a physician diagnosis, reported use of antihypertensive medication, or had measured systolic/diastolic blood pressure ≥ 140/90 mmHg at examination [12].
In CHARLS, incident heart disease and stroke were identified through self-reported physician diagnoses during follow-up interviews [13]. In ELSA, heart disease was ascertained by an affirmative response to the question, “Has a doctor ever told you that you have heart disease, including angina, heart attack, congestive heart failure, or other heart problems?”, while stroke was identified by a positive response to, “Has a doctor ever told you that you have had a stroke?” [14].
Cancer incidence was assessed in the ELSA cohort based on a positive response to the question, “Have you ever been told by a doctor or other health professional that you had cancer or any other kind of malignancy?” [15]. In CHARLS, cancer was identified through self-reported physician diagnosis during follow-up [13].
In CHARLS, chronic obstructive pulmonary disease (COPD) and asthma were identified through self-reported physician diagnoses during follow-up [13]. In ELSA, COPD was defined as an affirmative response to a physician diagnosis of a chronic lung condition (e.g., emphysema or chronic bronchitis), with asthma recorded as a separate outcome and therefore excluded from the COPD category; given its higher prevalence among older adults, most cases identified by this definition were attributable to COPD [16]. Incident asthma was determined at follow-up through the question, “Since your last interview, has a doctor ever told you that you have asthma?” [17].
Psychiatric disease was defined as a self-reported physician diagnosis of any mental health condition. In CHARLS, this included depression and other emotional or mental disorders reported during follow-up [13]. In ELSA, psychiatric disease encompassed physician-diagnosed depression, anxiety, emotional disorders, schizophrenia, psychosis, or bipolar disorder [18]. Memory disease was identified through self-reported physician diagnoses of dementia or Alzheimer’s disease and/or elevated scores on the Informant Questionnaire on Cognitive Decline in the Elderly [19]. Arthritis was defined at baseline as a positive response to the question, “Have you been diagnosed with arthritis (including osteoarthritis or rheumatism) by a doctor? [20]”.
Covariate
Based on prior literature and expert consensus, baseline covariates considered as potential confounders or effect modifiers included age, sex (male or female), alcohol consumption, smoking status, BMI and educational attainment (less than high school, high school, or college in CHARLS; college or non-college in ELSA). Laboratory assessments were used to measure clinical parameters such as LDL-C and HbA1c concentrations. Alcohol consumption and smoking status were each classified as “yes” or “no” [21, 22]. To reduce potential confounding from pre-existing conditions, the history of 10 chronic diseases was included as covariates in each disease-specific cohort across both studies, excluding the condition under investigation. This adjustment was intended to account for the influence of baseline comorbidities on subsequent health outcomes [13].
Statistical analysis
Baseline characteristics were summarized according to RCII quartiles. Missing baseline covariate data (e.g., HbA1c, education, smoking, alcohol consumption) in the ELSA cohort were addressed using multiple imputation with predictive mean matching, incorporating auxiliary predictors of non-response. Outcome variables, including incident chronic diseases, were not imputed. In the CHARLS cohort, individuals with missing data were excluded, as the resultant dataset remained sufficiently complete, and thus imputation was not required [23]. Normally distributed continuous variables were reported as means with standard errors, while skewed variables were presented as medians with interquartile ranges (IQRs). Categorical variables were expressed as frequencies and percentages. Group differences at baseline were assessed using the chi-square test for categorical variables, one-way analysis of variance (ANOVA) for normally distributed continuous variables, and the Kruskal–Wallis test for non-normally distributed continuous variables [24].
Given the skewed distribution of RCII, values were transformed using the natural logarithm (lgRCII) for all statistical analyses [5]. Restricted cubic spline (RCS) models with four knots positioned at the 5th, 35th, 65th, and 95th percentiles of lnRCII were used to explore potential nonlinear associations. Kaplan–Meier curves were constructed to estimate survival probabilities across lnRCII quartiles, and differences between groups were assessed using log-rank tests. Cox proportional hazards models were applied to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between RCII/lnRCII and incident chronic disease risk. To balance overadjustment avoidance with the inclusion of relevant covariates, two models were fitted: Model 0 refers to crude model without any adjustments; Model 1 adjusted for age, sex, education, alcohol consumption, smoking status, LDL-C, HbA1c, and BMI; Model 2 additionally adjusted for the history of 10 baseline chronic diseases, excluding the disease under investigation in each cohort. We also made sensitivity analyses by excluding hsCRP > 10 mg/L in supplementary Table 1 and 2. In addition, RCII and all covariates were measured at baseline and treated as fixed exposures in the primary analyses. Time-dependent covariates were not modeled, as repeated measurements of RCII and key confounders were not consistently available across follow-up waves in both cohorts. Therefore, the analyses were designed to evaluate longitudinal associations between baseline RCII and incident chronic diseases rather than to infer causal effects.
Table 1.
Baseline characteristics of participant in CHARLS database
| Overall(n = 9491 |
Q1 (n = 2354) |
Q2 (n = 2363) |
Q3 (n = 2353) |
Q4 (n = 2358) |
P value | |
|---|---|---|---|---|---|---|
| Age (years) | 59.17 ± 9.64 | 58.00 ± 9.69 | 58.82 ± 9.48 | 59.69 ± 9.64 | 60.15 ± 9.59 | < 0.0001 |
| Sex (Male %) | 4314 (45.76) | 1042 (44.27) | 1112 (47.06) | 1089 (46.28) | 1071 (45.42) | 0.25 |
| BMIa (kg/m2) | 23.52 ± 3.89 | 22.20 ± 3.32 | 23.03 ± 3.49 | 24.11 ± 3.84 | 24.76 ± 4.34 | < 0.0001 |
| HbA1c | 5.27 ± 0.81 | 5.14 ± 0.67 | 5.17 ± 0.65 | 5.30 ± 0.82 | 5.45 ± 1.01 | < 0.0001 |
| LDL-Cb (mg/dL) | 116.38 ± 35.07 | 115.57 ± 31.04 | 117.54 ± 33.36 | 119.54 ± 34.48 | 112.87 ± 40.38 | < 0.0001 |
| Education (%) | 0.60 | |||||
| Less Than High School | 8524 (90.41) | 2131 (90.53) | 2131 (90.18) | 2122 (90.18) | 2140 (90.75) | |
| College | 121 (1.28) | 23 (0.98) | 36 (1.52) | 28 (1.19) | 34 (1.44) | |
| High School | 783 (8.31) | 200 (8.50) | 196 (8.29) | 203 (8.63) | 184 (7.80) | |
| Smoke status (%) | 0.63 | |||||
| No | 6623 (70.25) | 1676 (71.20) | 1646 (69.66) | 1641 (69.74) | 1660 (70.40) | |
| Yes | 2805 (29.75) | 678 (28.80) | 717 (30.34) | 712 (30.26) | 698 (29.60) | |
| Alcohol drink (%) | 0.02 | |||||
| No | 6381 (67.68) | 1587 (67.42) | 1543 (65.30) | 1623 (68.98) | 1628 (69.04) | |
| Yes | 3047 (32.32) | 767 (32.58) | 820 (34.70) | 730 (31.02) | 730 (30.96) | |
| Cancer (%) | 76 (0.81) | 18 (0.76) | 17 (0.72) | 16 (0.68) | 25 (1.06) | 0.45 |
| COPDc (%) | 1006 (10.67) | 209 (8.88) | 242 (10.24) | 256 (10.88) | 299 (12.68) | < 0.001 |
| Heart disease (%) | 1078 (11.43) | 227 (9.64) | 246 (10.41) | 294 (12.49) | 311 (13.19) | < 0.001 |
| Hypertension (%) | 3845 (40.78) | 709 (30.12) | 844 (35.72) | 1080 (45.90) | 1212 (51.40) | < 0.0001 |
| Stroke (%) | 216 (2.29) | 27 (1.15) | 44 (1.86) | 62 (2.63) | 83 (3.52) | < 0.0001 |
| Psychiatric disease (%) | 102 (1.08) | 38 (1.61) | 20 (0.85) | 21 (0.89) | 23 (0.98) | 0.04 |
| Memory disease (%) | 124 (1.32) | 27 (1.15) | 31 (1.31) | 30 (1.27) | 36 (1.53) | 0.72 |
| Arthritis (%) | 3451 (36.60) | 811 (34.45) | 861 (36.44) | 863 (36.68) | 916 (38.85) | 0.02 |
| Asthma (%) | 397 (4.21) | 83 (3.53) | 74 (3.13) | 111 (4.72) | 129 (5.47) | < 0.001 |
| Diabetes Mellitus (%) | 1386 (14.70) | 197 (8.37) | 259 (10.96) | 380 (16.15) | 550 (23.32) | < 0.0001 |
RCII quartiles: Q1: –11.157–0.816; Q2: 0.816–2.084; Q3: 2.084–5.579; Q4: 5.579–3116.011
aBMI: body mass index
bLDL-C: low-density lipoprotein cholesterol
cCOPD: Chronic Obstructive Pulmonary Disease
All statistical analyses were performed in R (version 4.3.2). A two-sided P-value of < 0.05 was considered statistically significant [25].
Results
Study participants and baseline characteristics
This cross-national analysis included data from two population-based cohorts: 9,491 participants from the CHARLS and 6,054 participants from the ELSA (Tables 1–2). Table 1 presents the baseline characteristics of participants in the CHARLS cohort according to RCII quartiles. The distribution of participants across quartiles was approximately even: Q1 (RCII range: –11.157 to 0.816) included 2,354 participants (24.97%), Q2 (0.816 to 2.084) included 2,363 participants (25.06%), Q3 (2.084 to 5.579) included 2,353 participants (24.96%), and Q4 (5.579 to 3,116.011) included 2,358 participants (25.01%). Among the 9,491 participants, the mean age was 59.17 ± 9.64 years, with 45.76% being male. The average BMI was 23.52 ± 3.89 kg/m2, mean HbA1c was 5.27 ± 0.81, and mean LDL-C was 116.40 ± 35.20 mg/dL. Although education level, smoking status, and alcohol consumption did not differ significantly between quartiles, participants with higher RCII values had greater BMI and HbA1c, along with a higher prevalence and absolute number of participants with hypertension (n = 3,845), diabetes mellitus (n = 1,386), COPD (n = 1,006), heart disease (n = 1,078), stroke (n = 216), arthritis (n = 3,451), psychiatric disease (n = 102), and asthma (n = 397) (all P < 0.05). Cancer (n = 76) and memory disease (n = 124) did not differ significantly across quartiles.
Table 2.
Baseline characteristics of participant in ELSA database
|
Overall (n = 6054) |
Q1 (n = 1531) |
Q2 (n = 1527) |
Q3 (n = 1484) |
Q4 (n = 1512) |
P value | |
|---|---|---|---|---|---|---|
| Age (years) | 66.60 ± 9.14 | 65.39 ± 8.61 | 66.48 ± 8.90 | 67.06 ± 9.49 | 67.49 ± 9.44 | < 0.0001 |
| Sex (Male %) | 2714 (44.83) | 705 (46.05) | 698 (45.71) | 667 (44.95) | 644 (42.59) | 0.22 |
| BMI (kg/m2) | 28.06 ± 5.10 | 25.31 ± 3.75 | 27.37 ± 4.15 | 29.02 ± 4.65 | 30.59 ± 6.00 | < 0.0001 |
| HbA1c | 4.10 ± 0.83 | 3.92 ± 0.59 | 4.00 ± 0.63 | 4.13 ± 0.81 | 4.38 ± 1.11 | < 0.0001 |
| LDL-C (mg/dL) | 124.35 ± 40.28 | 119.97 ± 38.79 | 125.84 ± 39.55 | 125.46 ± 40.34 | 126.21 ± 42.10 | < 0.0001 |
| Education (%) | < 0.0001 | |||||
| College | 3338 (55.14) | 738 (48.20) | 782 (51.21) | 858 (57.82) | 960 (63.49) | |
| Non-college | 2716 (44.86) | 793 (51.80) | 745 (48.79) | 626 (42.18) | 552 (36.51) | |
| Smoke status (%) | < 0.0001 | |||||
| No | 5049 (83.40) | 1326 (86.61) | 1309 (85.72) | 1235 (83.22) | 1179 (77.98) | |
| Yes | 1005 (16.60) | 205 (13.39) | 218 (14.28) | 249 (16.78) | 333 (22.02) | |
| Alcohol drink (%) | < 0.0001 | |||||
| No | 5319 (87.86) | 1409 (92.03) | 1357 (88.87) | 1294 (87.20) | 1259 (83.27) | |
| Yes | 735 (12.14) | 122 (7.97) | 170 (11.13) | 190 (12.80) | 253 (16.73) | |
| Cancer (%) | 554 (9.15) | 130 (8.49) | 138 (9.04) | 129 (8.69) | 157 (10.38) | 0.27 |
| COPD (%) | 316 (5.22) | 45 (2.94) | 62 (4.06) | 80 (5.39) | 129 (8.53) | < 0.0001 |
| Heart disease (%) | 1161 (19.18) | 257 (16.79) | 272 (17.81) | 309 (20.82) | 323 (21.36) | < 0.01 |
| Hypertension (%) | 3241 (53.53) | 648 (42.33) | 784 (51.34) | 847 (57.08) | 962 (63.62) | < 0.0001 |
| Stroke (%) | 204 (3.37) | 42 (2.74) | 43 (2.82) | 47 (3.17) | 72 (4.76) | < 0.01 |
| Psychiatric disease (%) | 622 (10.27) | 139 (9.08) | 154 (10.09) | 146 (9.84) | 183 (12.10) | 0.04 |
| Memory disease (%) | 33 (0.55) | 7 (0.46) | 6 (0.39) | 7 (0.47) | 13 (0.86) | 0.29 |
| Arthritis (%) | 2214 (36.57) | 435 (28.41) | 511 (33.46) | 575 (38.75) | 693 (45.83) | < 0.0001 |
| Asthma (%) | 781 (12.90) | 156 (10.19) | 174 (11.39) | 207 (13.95) | 244 (16.14) | < 0.0001 |
| Diabetes Mellitus (%) | 617 (10.19) | 110 (7.18) | 116 (7.60) | 154 (10.38) | 237 (15.67) | < 0.0001 |
RCII quartiles: Q1: 0.077–1.624; Q2: 1.624–3.712; Q3: 3.712–8.817; Q4: 8.817–1822.285
Similarly, Table 2 presents the baseline characteristics of participants in the ELSA cohort according to RCII quartiles. The distribution of participants was roughly even across quartiles: Q1 (RCII range: 0.077–1.624) included 1,472 participants (24.31%), Q2 (1.624–3.712) included 1,529 participants (25.26%), Q3 (3.712–8.817) included 1,541 participants (25.45%), and Q4 (8.817–1,822.285) included 1,512 participants (24.98%). The mean age was 66.60 ± 9.14 years, with 44.83% male. The average BMI was 28.06 ± 5.11 kg/m2, mean HbA1c was 4.10 ± 0.83, and mean LDL-C was 124.35 ± 40.28 mg/dL. In this cohort, higher RCII quartiles were associated with older age, higher BMI, HbA1c, and LDL-C levels, lower educational attainment, and a greater proportion of current smokers. Participants in the highest RCII quartile also had a markedly higher prevalence and number of participants with cancer (n = 554), hypertension (n = 3,241), diabetes mellitus (n = 617), COPD (n = 316), heart disease (n = 1,161), stroke (n = 204), arthritis (n = 2,214), psychiatric disease (n = 622), and asthma (n = 781) (all P < 0.05), whereas memory disease (n = 33) prevalence did not differ significantly between quartiles.
Specifically, Supplementary Table 3 presents the baseline characteristics of participants prior to multiple imputation (n = 5,266), while Table 2 shows the corresponding characteristics after multiple imputation (n = 6,054). Overall, the distributions of key sociodemographic factors (age, sex, education), anthropometric measures (BMI), metabolic indicators (HbA1c, LDL-C), lifestyle behaviors (smoking and alcohol consumption), and major comorbidities were highly consistent between the two datasets. After multiple imputation, the total sample size increased as expected, but no material changes were observed in the direction or magnitude of group differences across RCII quartiles. Variables that showed statistically significant trends across quartiles before imputation (e.g., age, BMI, hypertension, diabetes mellitus, COPD, arthritis, and asthma) remained significant after imputation, while variables that were nonsignificant before imputation (e.g., sex, cancer, memory disease) also remained nonsignificant. In addition, the quartile cut-off values of RCII were only minimally altered after imputation, indicating good stability of the exposure classification.
Results of RCS analyses
In the CHARLS cohort (Fig. 2A–J), restricted cubic spline analyses were used to explore the potential dose–response relationships between lnRCII and incident chronic diseases. For stroke and diabetes mellitus, the tests for nonlinearity were not statistically significant (P for nonlinearity > 0.05), suggesting no strong evidence of nonlinear associations. For the remaining outcomes, spline curves did not demonstrate clear monotonic trends.
Fig. 2.
Association of lgRCII with new-onset chronic diseases of the CHARLS participants by RCS. The model adjusted for age, sex, education, alcohol drink, smoke, LDL, glucose, BMI, and the history of 9 chronic diseases at baseline (excluding the specific chronic disease under investigation in each cohort). A Hypertension; B Cancer; C COPD; D Heart disease; E Stroke; F Psychiatric disease; G Memory disease; H Arthritis; I Asthma; J Diabetes mellitus
In the ELSA cohort (Fig. 3A–J), a statistically significant nonlinear association was observed between lnRCII and incident cancer (P for nonlinearity = 0.0447), whereas approximately linear patterns were observed for COPD, stroke, psychiatric disease, and asthma (all P for nonlinearity > 0.05). RCS analyses were exploratory and intended to assess dose–response patterns rather than formal hypothesis testing.
Fig. 3.
Association of lgRCII with new-onset chronic diseases of the ELSA participants by RCS. The model adjusted for age, sex, education, alcohol drink, smoke, LDL, glucose, BMI, and the history of 9 chronic diseases at baseline (excluding the specific chronic disease under investigation in each cohort). A Hypertension; B Cancer; C COPD; D Heart disease; E Stroke; F Psychiatric disease; G Memory disease; H Arthritis; I Asthma; J Diabetes mellitus
Associations between lnRCII and new-onset chronic diseases in CHARLS and ELSA cohorts
Figures 4 and 5 depict Kaplan–Meier survival curves for the incidence of new-onset chronic diseases stratified by lnRCII quartiles. Diabetes mellitus was considered the primary outcome based on prior evidence, whereas analyses of other chronic diseases were exploratory. In the CHARLS cohort, significant differences were observed in the cumulative incidence of hypertension, heart disease, stroke, and diabetes over the follow-up period (Fig. 4A–J, log-rank P < 0.05). Similarly, in the ELSA cohort, the occurrence of new-onset hypertension, cancer, COPD, heart disease, stroke, psychiatric disorders, arthritis, asthma, and diabetes mellitus differed significantly across lnRCII quartiles during follow-up (Fig. 5A–J, log-rank P < 0.05).
Fig. 4.
Association of lgRCII with new-onset chronic diseases of the CHARLS participants by Kaplan–Meier curve. A Hypertension; B Cancer; C COPD; D Heart disease; E Stroke; F Psychiatric disease; G Memory disease; H Arthritis; I Asthma; J Diabetes mellitus
Fig. 5.
Association of lgRCII with new-onset chronic diseases of the ELSA participants by Kaplan–Meier curve. A Hypertension; B Cancer; C COPD; D Heart disease; E Stroke; F Psychiatric disease; G Memory disease; H Arthritis; I Asthma; J Diabetes mellitus
Multivariable Cox regression results of lnRCII for each new-onset chronic disease are summarized in Table 3 (CHARLS) and Table 4 (ELSA). In the CHARLS cohort, higher lnRCII was significantly associated with increased risk of incident hypertension, diabetes mellitus, heart disease and stroke in unadjusted models, with the associations for diabetes mellitus (HR = 1.09, 95% CI: 1.02–1.15) remaining significant after full adjustment. In the ELSA cohort, elevated lnRCII was positively associated with multiple outcomes, including hypertension, diabetes mellitus, heart disease, stroke, psychiatric disease, asthma, arthritis and COPD. Fully adjusted HRs for participants per unit increase in lnRCII were 1.13 (95% CI: 1.01–1.27) for diabetes mellitus, 1.26 (95% CI: 1.15–1.38) for stroke, 1.16 (95% CI: 1.03–1.30) for psychiatric disease, 1.27 (95% CI: 1.11–1.46) for asthma, and 1.36 (95% CI: 1.01, 1.51) for COPD (all P < 0.05). These findings indicate that elevated lnRCII is independently associated with incident diabetes in both cohorts, while associations with other chronic diseases were less consistent and should be interpreted as exploratory.
Table 3.
Risk classification of new-onset chronic diseases based on lnRCII by Multiple Cox Regression analysis in the CHARLS database
| Variables | Model 0 | Model 1a | Model 2b |
|---|---|---|---|
| Hypertension | 1.08 (1.04, 1.13)** | 1.04 (1.00, 1.08) | 1.04 (0.99, 1.08) |
| Cancer | 1.08 (0.94, 1.24) | 1.06 (0.91, 1.22) | 1.04 (0.90, 1.19) |
| Diabetes Mellitus | 1.21 (1.14, 1.28)*** | 1.11 (1.04, 1.17)** | 1.09 (1.02, 1.15)** |
| Heart disease | 1.10 (1.05, 1.16)* | 1.05 (0.99, 1.10) | 1.04 (0.98, 1.09) |
| Stroke | 1.17 (1.09, 1.25)*** | 1.10 (1.03, 1.18)* | 1.07 (1.00, 1.14) |
| Psychiatric disease | 0.98 (0.90, 1.08) | 0.97 (0.89, 1.06) | 0.96 (0.88, 1.04) |
| Asthma | 1.04 (0.96, 1.14) | 1.04 (0.95, 1.13) | 1.02 (0.93, 1.11) |
| Arthritis | 0.98 (0.95, 1.01) | 0.98 (0.95, 1.02) | 0.98 (0.95, 1.02) |
| Memory disease | 1.03 (0.95, 1.12) | 1.02 (0.94, 1.11) | 1.01 (0.93, 1.09) |
| COPD | 1.04 (0.98, 1.10) | 1.05 (0.99,1.12) | 1.05 (0.99, 1.12) |
aModel 1 adjusted for age, sex, education, alcohol drink, smoke, LDL, HbA1C and BMI
bModel 2 adjusted for age, sex, education, alcohol drink, smoke, LDL, HbA1C, BMI and the history of 9 chronic diseases at baseline (excluding the specific chronic disease under investigation in each cohort)
***P < 0.001, **P < 0.01, *P < 0.05
Table 4.
Risk classification of new-onset chronic diseases based on lnRCII by Multiple Cox Regression analysis in the ELSA database
| Variables | Model 0 | Model 1a | Model 2b |
|---|---|---|---|
| Hypertension | 1.19 (1.12, 1.27)** | 1.08 (1.00, 1.16) | 1.06 (0.98, 1.15) |
| Cancer | 1.10 (1.03, 1.08)* | 1.07 (0.99, 1.16) | 1.08 (0.99, 1.17) |
| Diabetes Mellitus | 1.58 (1.44, 1.74)*** | 1.20 (1.07, 1.34)** | 1.13 (1.01, 1.27)* |
| Heart disease | 1.16 (1.09, 1.23)*** | 1.08 (1.01, 1.16)* | 1.06 (0.99, 1.13) |
| Stroke | 1.36 (1.25, 1.47)*** | 1.01 (1.32, 1.44)*** | 1.26 (1.15, 1.38)*** |
| Psychiatric disease | 1.19 (1.08, 1.32)*** | 1.18 (1.05, 1.32)* | 1.16 (1.03, 1.30)* |
| Asthma | 1.41 (1.25, 1.59)*** | 1.31 (1.15, 1.50)*** | 1.27 (1.11, 1.46)** |
| Arthritis | 1.13 (1.07, 1.20)*** | 1.07 (1.01, 1.15)* | 1.06 (0.99, 1.13) |
| Memory disease | 1.04 (0.91, 1.18) | 0.96 (0.83, 1.12) | 0.95 (0.82, 1.11) |
| COPD | 1.43 (1.29, 1.58)*** | 1.39 (1.25, 1.55)*** | 1.36 (1.01, 1.51)*** |
aModel 1 adjusted for age, sex, education, alcohol drink, smoke, LDL, HbA1C and BMI
bModel 2 adjusted for age, sex, education, alcohol drink, smoke, LDL, HbA1C, BMI and the history of 9 chronic diseases at baseline (excluding the specific chronic disease under investigation in each cohort)
***P < 0.001, **P < 0.01, *P < 0.05
Sensitivity analyses
Sensitivity analyses were conducted to minimize the potential influence of acute inflammation by excluding participants with hsCRP levels > 10 mg/L (Supplementary Tables 1–2). The overall pattern of associations remained broadly consistent with the main findings. In the CHARLS cohort, higher lnRCII continued to show significant associations with an increased risk of incident diabetes mellitus and stroke in model 2, while the associations with hypertension and heart disease were slightly attenuated. In the ELSA cohort, elevated lnRCII remained positively associated with hypertension, cancer, diabetes mellitus, stroke, psychiatric disease, and COPD in model 2, suggesting that the observed relationships were robust and not driven by individuals with acute inflammatory responses.
Discussion
In two large population-based cohorts, higher baseline RCII was associated with greater risk of incident chronic disease, with the most consistent signal observed for new-onset diabetes; ELSA additionally showed associations with cerebrovascular and respiratory outcomes while CHARLS associations beyond diabetes were attenuated after full adjustment.
RCII is a pragmatic composite that captures the joint burden of remnant cholesterol and systemic low-grade inflammation. This construct rests on two complementary premises: (1) remnant particles are highly atherogenic and biologically active—readily infiltrating the arterial intima, activating vascular and immune cells, and contributing to plaque lipidation and inflammation [26]—and (2) chronic low-grade inflammation directly perturbs metabolic homeostasis, promoting insulin resistance (IR) and downstream organ injury [27]. Recent mechanistic and cohort studies establish RC as a driver of residual cardiometabolic risk beyond LDL-C and show that combining RC with hsCRP improves discrimination for vascular and systemic endpoints [28].
From a metabolic perspective, three interrelated, hypothesis-generating pathways may help contextualize the observed association between elevated RCII and subsequent diabetes risk and related chronic disorders. First, remnant-rich lipoproteins are associated with hepatic and peripheral insulin resistance [29]. Epidemiologic analyses and mediation studies indicate that higher RC correlates with surrogate and biochemical measures of IR, and IR partially mediates the RC–diabetes relationship [30]. Mechanistically, triglyceride-rich remnants are associated with alterations in hepatic lipid flux, insulin receptor signaling, and VLDL overproduction, which may contribute to hyperinsulinemia and glucose dysregulation [31]. Second, lipotoxic injury to pancreatic β-cells may reflect pathways linking between dyslipidemia and loss of insulin secretory capacity [32]. A body of experimental work shows that exposure of β-cells to cholesterol- and triglyceride-rich lipoproteins has been linked to endoplasmic reticulum stress, oxidative injury, and apoptosis, which may impair insulin secretion [33]. Clinical and translational studies further support the concept that perturbations in cholesterol handling within β-cells (including ABCA1/ABCG1 pathways) reduce β-cell function and accelerate progression from insulin resistance to frank type 2 diabetes [34]. These β-cell effects complement the systemic insulin resistance induced by RC and inflammation and help explain why RCII more strongly and reproducibly predicts diabetes onset than single markers alone. Third, inflammation functions both as consequence and amplifier of remnant-driven metabolic stress. Pro-inflammatory cytokines (TNF-α, IL-6 and others) impede insulin signalling through serine phosphorylation of insulin-receptor substrates, dysregulate adipose tissue lipolysis, and increase hepatic gluconeogenesis [35]. RC directly stimulates endothelial and innate immune cells to secrete such cytokines and to generate reactive oxygen species (ROS), creating a feed-forward loop in which remnant lipoproteins and inflammation may act synergistically in pathways associated with insulin resistance and organ dysfunction [36, 37]. Large reviews and mechanistic studies document these molecular links and the therapeutic rationale for interrupting inflammatory signaling in cardiometabolic disease. Notably, these mechanisms are primarily derived from experimental and translational studies and should be interpreted as indirect biological support rather than evidence of causality in the present observational analysis.
Although associations beyond diabetes were cohort-specific and exploratory, remnant–inflammation biology offers biologically plausible pathways that may link RCII to cerebrovascular and respiratory outcomes. Remnant particles are associated with endothelial dysfunction and procoagulant changes, which may underlie the observed associations with ischemic stroke [38]—while systemic inflammation contributes to adverse remodeling in the pulmonary compartment (airway inflammation, alveolar injury) and to susceptibility to COPD/asthma progression [39, 40]. Environmental exposures that increase systemic inflammation (for example long-term PM2.5 exposure) can operate along the same mechanistic axis, potentiating both metabolic and respiratory risk [41]. Animal and human data indicate that particulate pollution induces pulmonary oxidative stress and can provoke vascular insulin resistance and systemic inflammation, offering a biologically coherent pathway linking environment, RCII and multi-organ disease [42].
Epigenetic processes and the concept of “metabolic memory” provide a conceptual framework that may help contextualize persistent risk after transient metabolic insults [43]. Transient periods of dysmetabolism and oxidative stress—conditions that accompany elevated remnant load and inflammation—may leave durable DNA-methylation and histone marks that could help explain persistent risk after transient metabolic insults [44]. Such epigenetic imprinting offers a molecular rationale for why a baseline RCII signal may forecast long-term disease trajectories even if subsequent exposures change. However, direct evidence linking RCII to epigenetic modification is currently lacking and remains speculative.
The marked differences in association patterns between CHARLS and ELSA likely reflect both biological and methodological heterogeneity. Biologically, ELSA participants were older and had higher mean BMI and cardiometabolic burden than CHARLS (see cohort descriptions), conditions that amplify the clinical expression of an RCII signal and increase statistical power to detect multi-organ associations. In addition, outcome ascertainment (self-report vs clinic-verified events), follow-up intervals and competing risk structures differ across cohorts and can materially influence which associations achieve statistical significance. Together these factors plausibly explain why diabetes associations were consistent while other disease signals varied by cohort.
Clinically, RCII holds promise as a low-cost integrative biomarker to identify individuals at heightened multisystem risk. Importantly, the biological pathways implicated—triglyceride-rich lipoproteins and systemic inflammation—are actionable. Randomized evidence shows that targeting inflammation (e.g., canakinumab in CANTOS) or triglyceride-related pathways (e.g., icosapent ethyl in REDUCE-IT for selected high-risk patients) can lower cardiovascular events, highlighting a potential translational rationale for interventions targeting the RC–inflammation axis, though causal effects remain untested in prospective trials [45, 46]. However, whether lowering RCII per SE prevents diabetes, COPD, stroke or psychiatric outcomes remains untested and will require trials that combine lipid-lowering, triglyceride-targeted and anti-inflammatory strategies with metabolic endpoints.
This study presents several limitations. First, RCII was measured only at baseline, which precluded assessment of longitudinal trajectories or cumulative exposure effects. In CHARLS, only hypertension and diabetes were determined through objective measurements, whereas the remaining eight chronic diseases were self-reported. Such reliance on self-report increases the risk of misclassification, particularly for undiagnosed or subclinical conditions, and may affect the accuracy of outcome assessment. Residual confounding from unmeasured or imprecisely measured factors—such as diet, physical activity, air pollution exposure, and medication adherence—cannot be excluded despite comprehensive covariate adjustment. In addition, heterogeneity between cohorts in biomarker assays and healthcare context may complicate direct pooling and comparison of effect sizes. Second, although this investigation leveraged data from both CHARLS and ELSA, differences in demographic composition, baseline health profiles, and healthcare systems may limit direct comparability, and findings may not be fully generalizable to other populations. Validation in more ethnically and geographically diverse cohorts is warranted. Third, the diagnosis of diabetes did not differentiate between type 1 and type 2 diabetes. Given the age distribution of both cohorts and the high prevalence of type 2 diabetes in middle-aged and older adults, our findings most likely reflect associations with type 2 diabetes. Given the evaluation of multiple outcomes, some observed associations—particularly those with borderline significance—may be attributable to chance. Therefore, findings beyond diabetes mellitus should be interpreted as exploratory and require independent validation.
Conclusion
In this large prospective analysis of two national aging cohorts, elevated RCII was consistently associated with an increased risk of incident diabetes mellitus in both CHARLS and ELSA. In contrast, associations with other chronic diseases—including cerebrovascular, respiratory, and psychiatric outcomes—were observed primarily in the ELSA cohort and were attenuated after full adjustment in CHARLS.
Supplementary Information
Acknowledgements
None.
Author contributions
HZX conceived and designed the study, acquired the data and drafted the manuscript; TL analyzed the data; YJX contributed to the interpretation of the results and critical revision of the manuscript for important intellectual content; TL developed the software and provided technical support. QWF had the primary responsibility for final content. All authors have read and approved the final manuscript. The authors reported no conflicts of interest.
Funding
None.
Data availability
The datasets used and/or analyzed in this research are publicly accessible at http:/charls.pku.edu.cn/en and https://www.elsa-project.ac.uk/.
Declarations
Ethics approval and consent for publication
The CHARLS protocol was approved by the Institutional Review Board of Peking University (IRB00001052-11015), and written informed consent was obtained from all participants prior to enrolment. Ethical approval for ELSA wave 6 was granted by the NRES Committee South Central – Berkshire on 28 November 2012 (11/SC/0374), with all participants providing written informed consent before participation.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Hua-Zhao Xu and Tian Lv contributed equally.
References
- 1.Hacker K. The burden of chronic disease. Mayo Clinic Proceedings: Innovations, Quality & Outcomes. 2024;8(1):112–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Wang S, Jiang Y, Yang A, Meng F, Zhang J. The expanding burden of neurodegenerative diseases: an unmet medical and social need. Aging Dis. 2024. 10.14336/AD.2024.1071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Baratta F, Cocomello N, Coronati M, Ferro D, Pastori D, Angelico F, et al. Cholesterol remnants, triglyceride-rich lipoproteins and cardiovascular risk. Int J Mol Sci. 2023. 10.3390/ijms24054268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chen M, Chen Z, Ye H, Cheng Y, Jin Z, Cai S. Long-term association of remnant cholesterol with all-cause and cardiovascular disease mortality: a nationally representative cohort study. Front Cardiovasc Med. 2024;11:1286091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wang Y, Bi L, Li Q, Wang Q, Lv T, Zhang P. Remnant cholesterol inflammatory index and its association with all-cause and cause-specific mortality in middle-aged and elderly populations: evidence from US and Chinese national population surveys. Lipids Health Dis. 2025;24(1):155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Shao Q, Yang Z, Wang Y, Li Q, Han K, Liang J, et al. Elevated remnant cholesterol is associated with adverse cardiovascular outcomes in patients with acute coronary syndrome. J Atheroscler Thromb. 2022;29(12):1808–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chen J, Wu Q, Liu H, Hu W, Zhu J, Ji Z, et al. Predictive value of remnant cholesterol inflammatory index for stroke risk: Evidence from the China health and Retirement Longitudinal study. J Adv Res. 2024. 10.1016/j.jare.2024.12.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Liberale L, Badimon L, Montecucco F, Lüscher TF, Libby P, Camici GG. Inflammation, aging, and cardiovascular disease: JACC review topic of the week. J Am Coll Cardiol. 2022;79(8):837–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int J Epidemiol. 2014;43(1):61–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.He L, Yang J, Lin S, Shi K, Fang Y. Frailty detection with routine blood tests using data from the English longitudinal study of ageing (ELSA). Eur Geriatr Med. 2024;15(5):1347–55. [DOI] [PubMed] [Google Scholar]
- 11.Chen X, Li Z, Liu T, Li X, Li X, Duan R, et al. The relationship between cardiometabolic index and new-onset diabetes in adults aged over 45: a longitudinal analysis based on CHARLS. Int J Endocrinol. 2025;2025:2259853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Xiong YJ, Shao DM, Zhu XY, Lv T. Joint association of remnant cholesterol and body mass index with hypertension: a national cohort study in Chinese adults. J Multidiscip Healthc. 2025;18:1813–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zhuo L, Lai M, Wan L, Zhang X, Chen R. Cardiometabolic index and the risk of new-onset chronic diseases: results of a national prospective longitudinal study. Front Endocrinol (Lausanne). 2024;15:1446276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li X, Lu L, Chen Y, Liu B, Liu B, Tian H, et al. Association of atherogenic index of plasma trajectory with the incidence of cardiovascular disease over a 12-year follow-up: findings from the ELSA cohort study. Cardiovasc Diabetol. 2025;24(1):124. [DOI] [PMC free article] [PubMed]
- 15.Jackson SE, Williams K, Steptoe A, Wardle J. The impact of a cancer diagnosis on weight change: findings from prospective, population-based cohorts in the UK and the US. BMC Cancer. 2014;14(1):926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Loughran KJ, Tough D, Ryan CG, Wellburn S, Martin D, Dixon J, et al. The association of pain with incident falls in people with chronic obstructive pulmonary disease: evidence from the English longitudinal study of ageing. Int J Environ Res Public Health. 2023. 10.3390/ijerph20136236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Yu R, Chen Z, Zou L. The impact of sleep quality on asthma incidence in middle-aged and older adults: evidence from a prospective cohort study. Front Public Health. 2025. 10.3389/fpubh.2025.1646053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sun H, Carr H, Garcia-Argibay M, Cortese S, Solmi M, Golm D, et al. Large-scale evidence of a general disease (‘d’) factor accounting for both mental and physical health disorders in different age groups. Psychol Med. 2025;55:e78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Almeida-Meza P, Steptoe A, Cadar D. Markers of cognitive reserve and dementia incidence in the English Longitudinal Study of Ageing. Br J Psychiatry. 2021;218(5):243–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ma C, He S, Luo J, Zhang Z, Zhang H, Xiao W, et al. Social isolation and risk of mortality in middle-aged and older adults with arthritis: a prospective cohort study of four cohorts. Sci Rep. 2025;15(1):8073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Chen J, Yan L, Chu J, Wang X, Xu Z. Pain characteristics and progression to sarcopenia in Chinese middle-aged and older adults: a 4-year longitudinal study. J Gerontol A Biol Sci Med Sci. 2024. 10.1093/gerona/glae080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Yan J, Zhang MZ, He QQ. Association of changes and cumulative measures of triglyceride-glucose index-body mass index with hypertension risk: a prospective cohort study. BMC Public Health. 2024;24(1):2652. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kontto J, Paalanen L, Sund R, Sainio P, Koskinen S, Demakakos P, et al. Using multiple imputation and intervention-based scenarios to project the mobility of older adults. BMC Geriatr. 2022;22(1):311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhai L, Huo RR, Zuo YL. Atherogenic index of plasma and obesity-related risk of stroke in middle-aged and older Chinese adults: a national prospective cohort study. Diabetol Metab Syndr. 2024;16(1):245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Huo RR, Liao Q, Zhai L, You XM, Zuo YL. Interacting and joint effects of triglyceride-glucose index (TyG) and body mass index on stroke risk and the mediating role of TyG in middle-aged and older Chinese adults: a nationwide prospective cohort study. Cardiovasc Diabetol. 2024;23(1):30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Heo JH, Jo SH. Triglyceride-rich lipoproteins and remnant cholesterol in cardiovascular disease. J Korean Med Sci. 2023;38(38):e295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Che B, Zhong C, Zhang R, Pu L, Zhao T, Zhang Y, et al. Triglyceride-glucose index and triglyceride to high-density lipoprotein cholesterol ratio as potential cardiovascular disease risk factors: an analysis of UK biobank data. Cardiovasc Diabetol. 2023;22(1):34. [DOI] [PMC free article] [PubMed]
- 28.Quispe R, Martin SS, Michos ED, Lamba I, Blumenthal RS, Saeed A, et al. Remnant cholesterol predicts cardiovascular disease beyond LDL and ApoB: a primary prevention study. Eur Heart J. 2021;42(42):4324–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wang S, Zhang Q, Qin B. Association between remnant cholesterol and insulin resistance levels in patients with metabolic-associated fatty liver disease. Sci Rep. 2024;14(1):4596. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Li Y, Zeng Q, Peng D, Hu P, Luo J, Zheng K, et al. Association of remnant cholesterol with insulin resistance and type 2 diabetes: mediation analyses from NHANES 1999–2020. Lipids Health Dis. 2024;23(1):404. [DOI] [PMC free article] [PubMed]
- 31.Vargas-Vázquez A, Fermín-Martínez CA, Antonio-Villa NE, Fernández-Chirino L, Ramírez-García D, Dávila-López G, et al. Insulin resistance potentiates the effect of remnant cholesterol on cardiovascular mortality in individuals without diabetes. Atherosclerosis. 2024;395:117508. [DOI] [PubMed] [Google Scholar]
- 32.Al-Mrabeh A. β-cell dysfunction, hepatic lipid metabolism, and cardiovascular health in type 2 diabetes: new directions of research and novel therapeutic strategies. Biomedicines. 2021. 10.3390/biomedicines9020226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Roehrich ME, Mooser V, Lenain V, Herz J, Nimpf J, Azhar S, et al. Insulin-secreting beta-cell dysfunction induced by human lipoproteins. J Biol Chem. 2003;278(20):18368–75. [DOI] [PubMed] [Google Scholar]
- 34.Kruit JK, Wijesekara N, Westwell-Roper C, Vanmierlo T, de Haan W, Bhattacharjee A, et al. Loss of both ABCA1 and ABCG1 results in increased disturbances in islet sterol homeostasis, inflammation, and impaired β-cell function. Diabetes. 2012;61(3):659–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Chen L, Chen R, Wang H, Liang F. Mechanisms linking inflammation to insulin resistance. Int J Endocrinol. 2015;2015:508409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Tall AR, Yvan-Charvet L. Cholesterol, inflammation and innate immunity. Nat Rev Immunol. 2015;15(2):104–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Araujo G, Valencia LM, Martin-Ozimek A, Soto Y, Proctor SD. Atherosclerosis: from lipid-lowering and anti-inflammatory therapies to targeting arterial retention of ApoB-containing lipoproteins. Front Immunol. 2025;16:1485801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Miceli G, Basso MG, Pintus C, Pennacchio AR, Cocciola E, Cuffaro M, et al. Molecular pathways of vulnerable carotid plaques at risk of ischemic stroke: a narrative review. Int J Mol Sci. 2024. 10.3390/ijms25084351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Listyoko AS, Okazaki R, Harada T, Inui G, Yamasaki A. Impact of obesity on airway remodeling in asthma: pathophysiological insights and clinical implications. Front Allergy. 2024;5:1365801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Xu J, Zeng Q, Li S, Su Q, Fan H. Inflammation mechanism and research progress of COPD. Front Immunol. 2024;15:1404615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Xing YF, Xu YH, Shi MH, Lian YX. The impact of PM2.5 on the human respiratory system. J Thorac Dis. 2016;8(1):E69-74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Haberzettl P, O’Toole TE, Bhatnagar A, Conklin DJ. Exposure to fine particulate air pollution causes vascular insulin resistance by inducing pulmonary oxidative stress. Environ Health Perspect. 2016;124(12):1830–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Chen Z, Natarajan R. Epigenetic modifications in metabolic memory: what are the memories, and can we erase them? Am J Physiol Cell Physiol. 2022;323(2):C570-82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Zheng X, Sawalha AH. The role of oxidative stress in epigenetic changes underlying autoimmunity. Antioxid Redox Signal. 2022;36(7–9):423–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Ridker PM, Everett BM, Thuren T, MacFadyen JG, Chang WH, Ballantyne C, et al. Antiinflammatory therapy with canakinumab for atherosclerotic disease. N Engl J Med. 2017;377(12):1119–31. [DOI] [PubMed] [Google Scholar]
- 46.Bhatt DL, Steg PG, Miller M, Brinton EA, Jacobson TA, Ketchum SB, et al. Cardiovascular risk reduction with Icosapent Ethyl for hypertriglyceridemia. N Engl J Med. 2019;380(1):11–22. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analyzed in this research are publicly accessible at http:/charls.pku.edu.cn/en and https://www.elsa-project.ac.uk/.