Abstract
Background
Residual cardiovascular risk remains a significant clinical challenge despite the intensive management of conventional risk factors. This study aimed to evaluate the impact of the Remnant Cholesterol-Inflammation Index (RCII)—a novel metric integrating dyslipidemia and systemic inflammation—on the incidence of hypertension and its subsequent cardiovascular sequelae.
Methods
We leveraged data from two large-scale prospective cohorts. The China Health and Retirement Longitudinal Study (CHARLS) included a cross-sectional cohort for prevalent hypertension (N = 8,650) and a longitudinal incident-hypertension cohort (N = 5,022). The UK Biobank (UKB) included 273,122 participants with hypertension at baseline. Multivariable logistic regression and Cox proportional hazards models were used to assess incident hypertension and long-term adverse outcomes (all-cause mortality, major adverse cardiovascular events [MACE], and cardiovascular death), respectively.
Results
In the CHARLS cohort, high RCII was associated with a higher risk of incident hypertension in the fully adjusted model (OR, 1.346; 95% CI, 1.192–1.519). In quartile analysis, the highest RCII quartile was also associated with incident hypertension compared with the lowest quartile (OR, 1.364; 95% CI, 1.150–1.618; P for trend < 0.001). In the UKB cohort, compared with participants in the lowest RCII quartile, those in the highest quartile had higher risks of MACE (HR, 1.273; 95% CI, 1.232–1.315), all-cause mortality (HR, 1.214; 95% CI, 1.169–1.261), and cardiovascular death (HR, 1.284; 95% CI, 1.205–1.368) in fully adjusted complete-case models. Joint-effect analysis showed that concurrent elevation of remnant cholesterol and systemic inflammation was associated with the highest risk of MACE in UKB (HR, 1.16; 95% CI, 1.12–1.20). Exploratory mediation analyses suggested possible indirect pathways involving insulin resistance and oxidative stress.
Conclusions
RCII was associated with both incident hypertension and adverse prognosis among hypertensive participants. These findings support RCII as a complementary lipid-inflammatory residual-risk marker; however, external validation and decision-utility studies are needed before clinical implementation.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12872-026-06322-6.
Keywords: Remnant cholesterol, Inflammation, Hypertension, Cardiovascular prognosis, Major adverse cardiovascular events
Novelty and relevance
What is new?
A novel metric, RCII, integrates remnant cholesterol and inflammation to evaluate incident hypertension and adverse cardiovascular outcomes.
Insulin resistance and oxidative stress were evaluated as possible mediating pathways linking RCII to clinical outcomes.
What is relevant?
High RCII was associated with higher risk of developing hypertension and adverse cardiovascular events.
The findings suggest that remnant lipid accumulation and chronic inflammation may jointly contribute to vascular risk.
Clinical/pathophysiological implications?
Calculating RCII from routine blood tests may help identify individuals with higher residual risk.
Further validation is needed to determine whether RCII-guided risk stratification can improve prevention strategies.
Introduction
Hypertension is the chronic noncommunicable disease with the heaviest global disease burden and the leading risk factor for cardiovascular and cerebrovascular mortality. It is estimated that approximately 1.3 billion adults worldwide will be affected by 2025, and over half of these patients fail to achieve effective blood pressure control [1, 2]. Notably, even when blood pressure is controlled to target levels through treatment, patients’ cardiovascular disease risk is not fully eliminated, and residual risk persists due to diurnal blood pressure variability and early-onset target organ damage. Chronic low-grade inflammation and lipid deposition are closely associated with atherosclerotic plaque formation and adverse events such as coronary heart disease [3]. Among these, remnant cholesterol (RC)—characterized by higher cholesterol content and smaller particle size compared with low-density lipoprotein cholesterol (LDL-C)—is more readily taken up by subendothelial macrophages, thereby accelerating foam cell formation and atherogenesis [4, 5].
Emerging evidence suggests that lipids and inflammation are not merely parallel risk factors but act synergistically; inflammatory cytokines can impair triglyceride‑rich lipoprotein clearance, while RC can activate the NLRP3 inflammasome [6], creating a deleterious pro‑atherogenic feedback loop [7].
However, current cardiovascular risk assessment tools widely used in hypertensive populations focus predominantly on LDL‑C as a single conventional biomarker and do not sufficiently incorporate the synergistic and additive effects of remnant cholesterol and systemic inflammatory markers.
To synthesize these interrelated pathways, the Remnant Cholesterol‑Inflammation Index (RCII) has been proposed as a novel, integrated metric. While recent studies have separately linked RC and CRP to arterial stiffness and metabolic syndrome [8, 9], there is a paucity of longitudinal evidence characterizing the impact of their combined index, RCII, on the full continuum of hypertension—from its initial development to its terminal complications. Furthermore, whether these associations remain consistent across diverse ethnic backgrounds and through which metabolic or oxidative conduits they operate remains largely unexplored.
We leveraged data from two large-scale prospective cohorts: the China Health and Retirement Longitudinal Study (CHARLS) and the UK Biobank (UKB). By integrating these distinct datasets, this study aimed to: (1) evaluate the prospective association between RCII and hypertension incidence in a Chinese population; (2) determine the long-term prognostic value of RCII for all-cause and cardiovascular mortality among hypertensive participants in a UK cohort; and (3) explore possible mediating pathways involving insulin resistance and oxidative stress.
We hypothesized that RCII may serve as a complementary risk stratification marker across the hypertension continuum, supporting early identification of individuals at higher risk of hypertension onset and adverse prognosis among those with established hypertension.
Methods
Study design and participants
This study employed a dual-cohort, prospective observational design to systematically evaluate the association of RCII with both incident hypertension and long-term cardiovascular prognosis in hypertensive patients. The study populations were derived from the China Health and Retirement Longitudinal Study (CHARLS) and the UK Biobank (UKB). CHARLS was used to assess hypertension onset among participants without hypertension at baseline, whereas UK Biobank was used to assess prognosis among participants with established hypertension. In CHARLS, 8,650 participants with available baseline information were included in the cross-sectional analysis of prevalent hypertension. For the longitudinal incidence analysis, participants with hypertension at baseline were excluded. After further excluding 141 participants with missing follow-up hypertension status, 5,022 participants were included in the primary incident-hypertension cohort (Fig. 1A). In UK Biobank, 273,122 participants with hypertension at baseline were retained for long-term prognostic analyses (Fig. 1B). Fully adjusted Cox models were performed using complete-case analysis and included 270,861 participants. The two cohorts were analyzed separately because they differed in ethnicity, age structure, measurement platforms, follow-up design, and outcome definitions.
Fig. 1.
Flowchart of participant selection in the CHARLS and UK Biobank cohorts. Panel A shows the selection of CHARLS participants for the baseline cross-sectional analysis of prevalent hypertension and the longitudinal analysis of incident hypertension. The incident-hypertension cohort excluded participants with hypertension at baseline and further excluded participants with missing follow-up hypertension status, yielding 5,022 participants for the primary incident-hypertension analysis. Panel B shows the selection of UK Biobank participants with hypertension at baseline for prognostic analyses
Exposure and outcome
Definitions Exposure Variable: The primary exposure of interest was the Remnant Cholesterol-Inflammation Index (RCII). RC was defined as total cholesterol minus high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C). RCII was calculated as [total cholesterol − (HDL-C + LDL-C)] × hs-CRP/10. In the analyses, RCII was evaluated as a continuous variable and categorized into quartiles (Q1 to Q4), with the lowest quartile (Q1) serving as the reference group. Median-based high versus low RCII groups were also used for the CHARLS incident-hypertension analysis. To investigate the joint effect of RC and C-reactive protein (CRP), participants were stratified into four mutually exclusive joint exposure categories: Low RC/Low CRP, High RC/Low CRP, Low RC/High CRP, and High RC/High CRP. In CHARLS, high RC and high CRP were defined using cohort-specific medians. In UK Biobank, high RC was defined using the cohort-specific median, whereas high CRP was defined as hs-CRP ≥ 2 mg/L. LDL-C values used for RC calculation were obtained from cohort-provided biomarker measurements rather than calculated by the study investigators using the Friedewald formula. In UK Biobank, LDL-C was based on the “LDL direct” biomarker field measured by enzymatic protective selection analysis. In CHARLS, LDL-C was measured as part of the blood-based lipid biomarker panel using an enzymatic colorimetric assay.
Clinical outcomes
In CHARLS, the primary incidence outcome was incident hypertension during follow-up among participants without hypertension at baseline. Hypertension status was determined from follow-up wave assessments. In UK Biobank, the primary prognostic outcomes were major adverse cardiovascular events (MACE), all-cause mortality, and cardiovascular death. MACE was defined as the first occurrence of non-fatal myocardial infarction, non-fatal stroke, or cardiovascular death. Follow-up time was calculated from the baseline assessment date to the first occurrence of the outcome of interest or the end of follow-up, whichever came first.
Covariates assessment
To minimize potential confounding biases, comprehensive covariates were incorporated into the multivariable models. For the CHARLS cohort, adjusted covariates included demographic characteristics, socioeconomic status, lifestyle factors, body mass index, baseline comorbidities, and a comprehensive panel of baseline biochemical markers. For the UKB cohort, adjusted covariates comprised age, sex, BMI, Townsend deprivation index, education level, smoking and drinking status, history of lipid-lowering medication use, history of diabetes, serum urate, estimated glomerular filtration rate (eGFR), and glycated hemoglobin (HbA1c).
Statistical analysis
Baseline characteristics were summarized according to RCII quartiles. Continuous variables were expressed as means (standard deviations, SD) or medians (interquartile ranges, IQR), while categorical variables were presented as frequencies (percentages). The P for trend was calculated to evaluate linear trends across quartiles. In the UKB cohort, Kaplan–Meier survival curves were constructed to estimate the cumulative incidence of adverse events, and the log-rank test was used to compare survival distributions across RCII quartiles. Multivariable Cox proportional hazards regression models were used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for MACE, all-cause mortality, and cardiovascular death. In the CHARLS cohort, incident hypertension was analyzed as a follow-up binary outcome using multivariable logistic regression, yielding odds ratios (ORs) and 95% CIs. Three models were constructed: Model 1 was unadjusted; Model 2 was adjusted for age and sex; and Model 3 was fully adjusted for prespecified covariates. Restricted cubic spline analyses were used to examine dose–response associations between continuous RCII and clinical outcomes, with P values estimated for overall association and non-linearity. Subgroup analyses were used to evaluate consistency of associations across demographic and clinical strata, and P for interaction was calculated to assess heterogeneity. To assess effect modification by lipid-lowering medication, UK Biobank analyses were stratified by baseline lipid-lowering medication use and RCII-by-medication interaction terms were fitted. Sensitivity analyses were performed to evaluate the robustness of the primary findings by excluding participants or events potentially susceptible to baseline comorbidity bias or reverse causation. To evaluate incremental predictive performance, RCII was added to traditional, Framingham-like, and ASCVD-like risk-factor models constructed from available baseline variables. Model discrimination was assessed with Harrell’s C-index, and incremental discrimination was quantified as the change in C-index after adding RCII. Exploratory mediation analyses were conducted to estimate indirect associations through selected metabolic and oxidative biomarkers. All statistical analyses were performed using R software (version 4.5.0, R Foundation for Statistical Computing). A two-sided P-value of < 0.05 was considered statistically significant.
Results
Baseline characteristics of the study populations
The stepwise participant selection process for both cohorts is detailed in Fig. 1. The baseline characteristics of the study populations, stratified by RCII quartiles, are summarized in Table 1. In the CHARLS cohort, higher RCII levels were associated with older age, a higher proportion of current smokers, greater adiposity, and a less favorable metabolic and inflammatory profile. Specifically, participants in the highest RCII quartile (Q4) had higher body mass index (BMI), waist circumference, triglycerides (TG), fasting plasma glucose (FPG), and C-reactive protein (CRP), and lower HDL-C, than those in the lowest quartile (Q1). In the UK Biobank cohort, increasing RCII quartiles were also associated with higher BMI, systolic blood pressure, triglycerides, HbA1c, and CRP, together with lower HDL-C.
Table 1.
Baseline characteristics of participants in CHARLS and UK Biobank by RCII quartiles
| Characteristic | Total | Q1 | Q2 | Q3 | Q4 | P for trend |
|---|---|---|---|---|---|---|
| Panel A: CHARLS Cohort | (N = 5,022) | (N = 1,256) | (N = 1,256) | (N = 1,255) | (N = 1,255) | |
| Demographics | ||||||
| Age, years, mean (SD) | 57.38 (8.71) | 56.62 (8.70) | 57.18 (8.51) | 57.73 (8.73) | 58.00 (8.86) | < 0.001 |
| Sex, No. (%) | 0.208 | |||||
| Male | 2377 (47.3) | 565 (45.0) | 605 (48.2) | 615 (49.0) | 592 (47.2) | |
| Female | 2645 (52.7) | 691 (55.0) | 651 (51.8) | 640 (51.0) | 663 (52.8) | |
| Education level, No. (%) | 0.894 | |||||
| Primary school or below | 4510 (89.8) | 1131 (90.0) | 1125 (89.6) | 1120 (89.2) | 1134 (90.4) | |
| Middle/High school | 460 (9.2) | 115 (9.2) | 115 (9.2) | 121 (9.6) | 109 (8.7) | |
| College or above | 52 (1.0) | 10 (0.8) | 16 (1.3) | 14 (1.1) | 12 (1.0) | |
| Marital status, No. (%) | 0.775 | |||||
| Married | 4313 (85.9) | 1080 (86.0) | 1080 (86.0) | 1089 (86.8) | 1064 (84.8) | |
| Separated/Divorced | 254 (5.1) | 70 (5.6) | 61 (4.9) | 57 (4.5) | 66 (5.3) | |
| Widowed | 34 (0.7) | 7 (0.6) | 9 (0.7) | 7 (0.6) | 11 (0.9) | |
| Other | 421 (8.4) | 99 (7.9) | 106 (8.4) | 102 (8.1) | 114 (9.1) | |
| Residence, No. (%) | < 0.001 | |||||
| Urban | 1648 (32.8) | 341 (27.1) | 391 (31.1) | 452 (36.0) | 464 (37.0) | |
| Rural | 3374 (67.2) | 915 (72.9) | 865 (68.9) | 803 (64.0) | 791 (63.0) | |
| Lifestyle | ||||||
| Drinking status (Current), No. (%) | 1950 (38.8) | 451 (35.9) | 502 (40.0) | 497 (39.6) | 500 (39.8) | 0.109 |
| Smoking status (Current), No. (%) | 1989 (39.6) | 449 (35.7) | 495 (39.4) | 529 (42.2) | 516 (41.1) | 0.006 |
| Clinical Characteristics | ||||||
| BMI, kg/m2, mean (SD) | 23.59 (3.51) | 21.82 (3.07) | 22.84 (3.29) | 25.98 (3.69) | 23.73 (3.98) | 0.033 |
| Waist circumference, cm, mean (SD) | 82.17 (11.66) | 79.01 (10.46) | 81.32 (10.78) | 83.11 (11.70) | 85.24 (12.69) | < 0.001 |
| Incident hypertension during follow-up, No. (%) | 1640 (32.7) | 365 (29.1) | 362 (28.8) | 445 (35.5) | 468 (37.3) | < 0.001 |
| Diabetes, No. (%) | 181 (3.6) | 33 (2.6) | 32 (2.5) | 47 (3.7) | 69 (5.5) | < 0.001 |
| Heart disease, No. (%) | 383 (7.6) | 83 (6.6) | 79 (6.3) | 113 (9.0) | 108 (8.6) | 0.018 |
| Stroke, No. (%) | 62 (1.2) | 10 (0.8) | 21 (1.7) | 13 (1.0) | 18 (1.4) | 0.19 |
| Laboratory Measures | ||||||
| Fasting plasma glucose, mg/dL | 100.98 | 98.46 | 99.72 | 101.34 | 104.76 | < 0.001 |
| median [IQR] | [93.42–110.52] | [92.16–106.02] | [92.88–108.90] | [93.60–111.42] | [95.58–117.27] | |
| HbA1c, %, median [IQR] | 5.10 [4.90–5.40] | 5.10 [4.80–5.30] | 5.10 [4.90–5.40] | 5.10 [4.90–5.40] | 5.20 [4.90–5.50] | < 0.001 |
| Total cholesterol, mg/dL, mean (SD) | 191.16 (37.49) | 185.69 (34.38) | 188.38 (34.45) | 193.23 (37.58) | 197.34 (41.99) | < 0.001 |
| Triglycerides, mg/dL, median [IQR] | 99.12 [72.57–143.37] | 72.57 [58.41–90.27] | 96.46 [74.12–123.01] | 116.82 [84.07–157.53] | 143.37 [97.79–213.29] | < 0.001 |
| HDL-C, mg/dL, mean (SD) | 52.17 (15.13) | 59.74 (14.31) | 53.63 (13.97) | 50.32 (13.87) | 44.97 (14.47) | < 0.001 |
| LDL-C, mg/dL, mean (SD) | 115.20 (33.68) | 115.88 (30.50) | 115.84 (31.41) | 116.82 (34.86) | 112.24 (37.37) | 0.004 |
| Blood urea nitrogen, mg/dL | 15.15 [12.55–18.12] | 15.18 [12.55–18.21] | 15.29 [12.63–18.24] | 15.24 [12.83–18.19] | 14.87 [12.30–17.72] | 0.034 |
| Creatinine, mg/dL, median [IQR] | 0.75 [0.64–0.87] | 0.72 [0.63–0.85] | 0.75 [0.64–0.87] | 0.76 [0.64–0.88] | 0.76 [0.66–0.88] | < 0.001 |
| Serum uric acid, mmol/L | 4.18 [3.50–4.99] | 3.90 [3.26–4.60] | 4.10 [3.43–4.85] | 4.31 [3.62–5.12] | 4.52 [3.75–5.33] | < 0.001 |
| Cystatin C, mg/L, median [IQR] | 0.96 [0.85–1.10] | 0.94 [0.84–1.06] | 0.96 [0.85–1.09] | 0.98 [0.86–1.11] | 0.98 [0.83–1.13] | 0.001 |
| C-reactive protein, mg/L, median [IQR] | 0.89 [0.50–1.88] | 0.43 [0.30–0.59] | 0.65 [0.47–0.93] | 1.14 [0.79–1.70] | 3.14 [1.91–6.06] | < 0.001 |
| Panel B: UK Biobank Cohort | (N = 273,122) | (N = 68,281) | (N = 68,281) | (N = 68,280) | (N = 68,280) | |
| Demographics | ||||||
| Age, years, mean (SD) | 58.18 (7.54) | 57.64 (7.77) | 58.43 (7.47) | 58.40 (7.46) | 58.25 (7.42) | < 0.001 |
| Sex, No. (%) | < 0.001 | |||||
| Male | 134,749 (49.3) | 35,068 (51.4) | 35,882 (52.6) | 33,646 (49.3) | 30,153 (44.2) | |
| Female | 138,373 (50.7) | 33,213 (48.6) | 32,399 (47.4) | 34,634 (50.7) | 38,127 (55.8) | |
| Race, No. (%) | < 0.001 | |||||
| White | 257,345 (94.2) | 63,997 (93.7) | 64,444 (94.4) | 64,223 (94.1) | 64,681 (94.7) | |
| Non-White | 15,777 (5.8) | 4,284 (6.3) | 3,837 (5.6) | 4,057 (5.9) | 3,599 (5.3) | |
| Education level, No. (%) | < 0.001 | |||||
| University or above | 28,464 (10.4) | 8,884 (13.0) | 7,351 (10.8) | 6,573 (9.6) | 5,656 (8.3) | |
| Secondary | 81,824 (30.0) | 19,290 (28.3) | 20,480 (30.0) | 20,390 (29.9) | 21,664 (31.7) | |
| Primary or below | 162,834 (59.6) | 40,107 (58.7) | 40,450 (59.2) | 41,317 (60.5) | 40,960 (60.0) | |
| Townsend index, median [IQR] | −2.2 [−3.7–0.47] | −2.3 [−3.8–0.10] | −2.3 [−3.7– 0.21] | −2.1 [−3.6–0.59] | −1.9 [−3.5–0.96] | < 0.001 |
| Lifestyle | ||||||
| Smoking status (Current), No. (%) | 98,492 (36.1) | 23,461 (34.4) | 25,046 (36.7) | 24,602 (36.0) | 25,383 (37.2) | < 0.001 |
| Drinking status (Current), No. (%) | 9,974 (3.7) | 1,910 (2.8) | 2,319 (3.4) | 2,569 (3.8) | 3,176 (4.7) | < 0.001 |
| Physical activity, Moderate, mean (SD) | 1.42 (0.77) | 1.45 (0.79) | 1.41 (0.75) | 1.42 (0.76) | 1.41 (0.76) | < 0.001 |
| Medical History | ||||||
| Diabetes mellitus, No. (%) | 18,963 (6.9) | 4,918 (7.2) | 4,543 (6.7) | 4,670 (6.8) | 4,832 (7.1) | < 0.001 |
| Atrial fibrillation, No. (%) | 3,243 (1.2) | 822 (1.2) | 892 (1.3) | 731 (1.1) | 798 (1.2) | 0.001 |
| Lipid-lowering medication, No. (%) | 21,301 (7.8) | 6,334 (9.3) | 5,636 (8.3) | 5,014 (7.3) | 4,317 (6.3) | < 0.001 |
| Clinical & Laboratory Measures | ||||||
| BMI, kg/m2, mean (SD) | 28.30 (4.89) | 25.78 (3.74) | 27.84 (4.32) | 28.92 (4.76) | 30.68 (5.28) | < 0.001 |
| Systolic BP, mmHg, mean (SD) | 150.19 (16.64) | 149.13 (16.18) | 150.22 (16.42) | 150.54 (16.72) | 150.88 (17.16) | < 0.001 |
| Diastolic BP, mmHg, mean (SD) | 87.01 (9.73) | 85.70 (9.58) | 86.90 (9.67) | 87.37 (9.70) | 88.08 (9.81) | < 0.001 |
| Pulse pressure, mmHg, median [IQR] | 62.0 [54.0–71.0] | 62.0 [54.0–71.0] | 62.0 [54.0–72.0] | 62.0 [53.8–71.0] | 62.0 [53.0–71.0] | < 0.001 |
| Total Cholesterol, mmol/L, mean (SD) | 5.8 (1.1) | 5.4 (1.1) | 5.7 (1.1) | 5.9 (0.95) | 6.2 (1.2) | < 0.001 |
| LDL-C, mmol/L, mean (SD) | 3.62 (0.85) | 3.28 (0.80) | 3.57 (0.83) | 3.71 (0.73) | 3.93 (0.88) | < 0.001 |
| HDL-C, mmol/L, mean (SD) | 1.42 (0.35) | 1.56 (0.40) | 1.44 (0.36) | 1.38 (0.29) | 1.32 (0.31) | < 0.001 |
| Triglycerides, mmol/L, median [IQR] | 1.60 [1.17–2.23] | 1.19 [0.89–1.62] | 1.54 [1.14–2.11] | 1.60 [1.50–2.19] | 2.08 [1.51–2.92] | < 0.001 |
| HbA1c, mmol/mol, mean (SD) | 36 (5.0) | 35 (5.0) | 36 (4.9) | 36 (4.8) | 37 (5.2) | < 0.001 |
| Creatinine, mmol/L, mean (SD) | 72 (14) | 72 (14) | 72 (14) | 71 (14) | 72 (14) | < 0.001 |
| C-reactive protein, mg/L, median [IQR] | 1.50 [0.82–2.73] | 0.52 [0.36–0.74] | 1.19 [0.90–1.50] | 1.65 [1.50–2.31] | 4.35 [3.05–6.51] | < 0.001 |
Continuous variables are presented as Mean ± SD or Median (IQR) for skewed distributions. Categorical variables are presented as n (%). P-values were calculated using ANOVA, Kruskal–Wallis, or Chi-square tests
In the CHARLS incident cohort, hypertension denotes incident hypertension during follow-up among participants without hypertension at baseline
Abbreviations: BMI body mass index, SBP systolic blood pressure, FPG fasting plasma glucose, TC total cholesterol, TG triglyceride, RC remnant cholesterol, SUA serum uric acid
Association between RCII and incident hypertension
In the prospective analysis of the CHARLS cohort, elevated RCII was significantly associated with an increased risk of incident hypertension. Among 5,022 CHARLS participants without hypertension at baseline and with available follow-up hypertension status, 1,640 developed incident hypertension during follow-up. In the fully adjusted complete-case model (Model 3), high RCII was associated with a higher risk of incident hypertension compared with low RCII (OR, 1.346; 95% CI, 1.192–1.519; P < 0.001; n = 4,972; events = 1,622) (Table 2). In quartile analysis, participants in the highest RCII quartile had a higher risk of incident hypertension than those in the lowest quartile (OR, 1.364; 95% CI, 1.150–1.618; P < 0.001; P for trend < 0.001). Restricted cubic spline (RCS) analysis demonstrated a positive dose–response association between continuous RCII and the incidence of hypertension (Fig. 3D).
Table 2.
Association of RCII with hypertension prevalence and incidence in the CHARLS cohort
| Analysis and RCII Categories | Model 1 (Crude) | Model 2 (Age & Sex) | Model 3 (Fully Adjusted) |
|---|---|---|---|
| OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Panel A: Prevalence of Hypertension (n = 8,650) | |||
| Quartile 1 (Lowest) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Quartile 2 | 1.43 (1.23–1.66) | 1.40 (1.21–1.63) | 1.29 (1.08–1.54) |
| Quartile 3 | 2.01 (1.74–2.32) | 1.94 (1.68–2.25) | 1.59 (1.33–1.89) |
| Quartile 4 (Highest) | 2.48 (2.15–2.86) | 2.36 (2.05–2.73) | 1.86 (1.56–2.22) |
| P for trend | < 0.001 | < 0.001 | < 0.001 |
| Panel B: Incidence of Hypertension (n = 5,022) | |||
| Low RCII | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| High RCII | 1.407 (1.250–1.584) | 1.369 (1.214–1.544) | 1.346 (1.192–1.519) |
| P value | < 0.001 | < 0.001 | < 0.001 |
| Panel C: Incidence of Hypertension by RCII Quartiles (n = 5,022) | |||
| Quartile 1 (Lowest) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Quartile 2 | 0.988 (0.832–1.175) | 0.969 (0.814–1.154) | 0.960 (0.806–1.144) |
| Quartile 3 | 1.341 (1.134–1.586) | 1.297 (1.094–1.537) | 1.267 (1.067–1.504) |
| Quartile 4 (Highest) | 1.452 (1.228–1.716) | 1.393 (1.176–1.649) | 1.364 (1.150–1.618) |
| P for trend | < 0.001 | < 0.001 | < 0.001 |
Abbreviations: RCII Remnant Cholesterol-Inflammation Index, OR odds ratio, CI confidence interval
Panel A (Prevalence): Assessed the cross-sectional association between RCII quartiles and hypertension prevalence at baseline
Panel B (Incidence): Evaluated the prospective association between RCII levels (categorized as High vs. Low) and incident hypertension during follow-up among participants without hypertension at baseline
Model 1: Unadjusted model
Model 2: Adjusted for age and sex
Model 3: Adjusted for age, sex, BMI, smoking status, drinking status, and baseline comorbidities (diabetes, heart disease, and stroke). For CHARLS incident hypertension, the primary analytic cohort included 5,022 participants; the fully adjusted Model 3 included 4,972 participants and 1,622 events because of complete-case covariate adjustment
Fig. 3.
Dose–response relationships between RCII and clinical outcomes. Restricted cubic spline curves illustrate associations of continuous RCII with MACE, all-cause mortality, and cardiovascular death in UK Biobank and incident hypertension in CHARLS. Panels A–C were based on fully adjusted Cox models, and Panel D was based on a fully adjusted logistic regression model
Prognostic value of RCII in hypertensive patients
For hypertensive patients in the UKB cohort, Kaplan–Meier survival curves revealed that individuals in the highest RCII quartile had a significantly higher cumulative incidence of major adverse cardiovascular events (MACE), all-cause mortality, and cardiovascular death over the follow-up period (log-rank P < 0.001 for all outcomes, Fig. 2). Multivariable Cox proportional hazards regression confirmed these prognostic associations (Table 3). In fully adjusted complete-case Cox models including 270,861 participants, compared with the lowest RCII quartile, the highest quartile was associated with higher risks of MACE (HR, 1.273; 95% CI, 1.232–1.315; P < 0.001), all-cause mortality (HR, 1.214; 95% CI, 1.169–1.261; P < 0.001), and cardiovascular death (HR, 1.284; 95% CI, 1.205–1.368; P < 0.001). RCS analyses also supported dose–response associations between continuous RCII and the risk of these long-term adverse outcomes (Fig. 3A-C). Adding RCII to traditional, Framingham-like, and ASCVD-like risk-factor models produced small but consistent improvements in Harrell’s C-index for MACE and all-cause mortality (Table S3 and Table S4).
Fig. 2.
Kaplan–Meier survival estimates for long-term clinical outcomes stratified by RCII quartiles in participants with hypertension. Curves display the cumulative incidence of major adverse cardiovascular events (MACE), all-cause mortality, and cardiovascular death in UK Biobank across RCII quartiles
Table 3.
Prognostic association of RCII quartiles with long-term clinical outcomes in hypertensive participants (UK Biobank)
| Clinical Outcomes & RCII Quartiles | Model 1 (Crude) | Model 2 (Age & Sex) | Model 3 (Fully Adjusted) |
|---|---|---|---|
| HR (95% CI) | HR (95% CI) | HR (95% CI) | |
| All-cause Mortality | |||
| Quartile 1 (Lowest) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Quartile 2 | 1.163 (1.121–1.205) | 1.105 (1.066–1.146) | 1.038 (1.000–1.077) |
| Quartile 3 | 1.281 (1.236–1.327) | 1.243 (1.199–1.288) | 1.097 (1.057–1.138) |
| Quartile 4 (Highest) | 1.455 (1.406–1.506) | 1.476 (1.426–1.528) | 1.214 (1.169–1.261) |
| P for trend | < 0.001 | < 0.001 | < 0.001 |
| Major Adverse Cardiovascular Events (MACE) | |||
| Quartile 1 (Lowest) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Quartile 2 | 1.176 (1.139–1.214) | 1.130 (1.095–1.166) | 1.053 (1.020–1.087) |
| Quartile 3 | 1.309 (1.269–1.350) | 1.292 (1.253–1.332) | 1.140 (1.104–1.177) |
| Quartile 4 (Highest) | 1.500 (1.456–1.546) | 1.558 (1.512–1.606) | 1.273 (1.232–1.315) |
| P for trend | < 0.001 | < 0.001 | < 0.001 |
| Cardiovascular (CVD) Mortality | |||
| Quartile 1 (Lowest) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Quartile 2 | 1.222 (1.148–1.300) | 1.160 (1.090–1.235) | 1.040 (0.976–1.108) |
| Quartile 3 | 1.416 (1.333–1.504) | 1.383 (1.302–1.470) | 1.138 (1.069–1.212) |
| Quartile 4 (Highest) | 1.692 (1.596–1.793) | 1.755 (1.655–1.860) | 1.284 (1.205–1.368) |
| P for trend | < 0.001 | < 0.001 | < 0.001 |
Model 1 was unadjusted. Model 2 was adjusted for age and sex. Model 3 was fully adjusted for age, sex, BMI, Townsend deprivation index, education level, smoking status, drinking status, lipid-lowering medication use, history of diabetes, serum urate, eGFR, and HbA1c. CVD death was adjudicated using ICD-10 codes I00–I99. MACE was a composite of non-fatal myocardial infarction, non-fatal stroke, and CVD death
Abbreviations: RCII Remnant Cholesterol-Inflammation Index, HR hazard ratio, CI confidence interval, MACE major adverse cardiovascular events, CVD cardiovascular disease
Joint effect of remnant cholesterol and systemic inflammation
To further evaluate the joint impact of lipid and inflammatory pathways, participants were stratified into four mutually exclusive groups based on RC and CRP levels. A stepwise increase in risk was observed with the concurrent elevation of both markers. In the CHARLS cohort, individuals with both high RC and high CRP exhibited the highest risk of incident hypertension (OR, 2.05; 95% CI, 1.73–2.42) compared to the low-risk reference group (Low RC/Low CRP) (Table 4). A similar pattern was also observed in the baseline cross-sectional analysis for prevalent hypertension (Table S2). In the UKB cohort, the High RC/High CRP phenotype was associated with the highest risk of MACE (HR, 1.16; 95% CI, 1.12–1.20) in fully adjusted models (Table 4, Fig. S1).
Table 4.
Joint association of remnant cholesterol and C-reactive protein levels with hypertension incidence and MACE risk
| Analysis and Joint Categories | Model 1 (Crude) | P-value | Model 2 (Age & Sex) | P-value | Model 3 (Fully Adjusted) | P-value |
|---|---|---|---|---|---|---|
| OR/HR (95% CI) | OR/HR (95% CI) | OR/HR (95% CI) | ||||
| Panel A: Incident Hypertension (CHARLS) | ||||||
| Low RC/Low CRP | 1.00 (Reference) | — | 1.00 (Reference) | — | 1.00 (Reference) | — |
| High RC/Low CRP | 1.46 (1.27–1.67) | < 0.001 | 1.47 (1.28–1.68) | < 0.001 | 1.43 (1.25–1.64) | < 0.001 |
| Low RC/High CRP | 1.50 (1.25–1.80) | < 0.001 | 1.44 (1.20–1.73) | < 0.001 | 1.39 (1.16–1.67) | < 0.001 |
| High RC/High CRP | 2.29 (1.95–2.69) | < 0.001 | 2.21 (1.88–2.60) | < 0.001 | 2.05 (1.73–2.42) | < 0.001 |
| Panel B: Major Adverse Cardiovascular Events (UKB) | ||||||
| Low RC/Low CRP | 1.00 (Reference) | — | 1.00 (Reference) | — | 1.00 (Reference) | — |
| High RC/Low CRP | 1.09 (1.06–1.12) | < 0.001 | 1.10 (1.07–1.13) | < 0.001 | 1.07 (1.04–1.10) | < 0.001 |
| Low RC/High CRP | 1.32 (1.28–1.37) | < 0.001 | 1.25 (1.21–1.29) | < 0.001 | 1.12 (1.08–1.16) | < 0.001 |
| High RC/High CRP | 1.36 (1.32–1.40) | < 0.001 | 1.33 (1.29–1.37) | < 0.001 | 1.16 (1.12–1.20) | < 0.001 |
Abbreviations: RC remnant cholesterol, CRP C-reactive protein, OR odds ratio, HR hazard ratio, CI confidence interval, MACE major adverse cardiovascular events, BMI body mass index, SUA serum uric acid, HbA1c glycated hemoglobin, eGFR estimated glomerular filtration rate
Model Definitions: Model 1: Crude model without any adjustment
Model 2: Adjusted for demographic factors (age and sex)
Model 3 (UKB): Adjusted for age, sex, BMI, Townsend deprivation index, education level, smoking status, drinking status, lipid-lowering medication use, history of diabetes, serum urate, eGFR, and HbA1c
Model 3 (CHARLS): Adjusted for age, sex, education, hukou, residence, history of diabetes, heart disease and stroke, smoking status, drinking status, BMI, and biochemical markers (blood urea nitrogen, fasting plasma glucose, creatinine, HbA1c, SUA, and Cystatin C)
Operational Definitions: In CHARLS, high RC and high CRP were defined using cohort-specific medians. In UK Biobank, high RC was defined using the cohort-specific median, whereas high CRP was defined as hs-CRP ≥ 2 mg/L
Subgroup and robustness analyses
Subgroup analyses demonstrated that the associations between RCII and clinical outcomes were broadly consistent across various demographic and clinical strata. Specifically, the association with incident hypertension in CHARLS remained robust across prespecified subgroups (Fig. S2). In the UKB cohort, subgroup analyses for the primary prognostic outcomes are presented in Fig. 4. Although some interaction tests reached statistical significance, the associations were generally directionally consistent across subgroups defined by age, sex, BMI, and smoking status. In analyses stratified by baseline lipid-lowering medication use, the association between RCII and MACE was stronger among non-users than users, with a significant interaction, whereas interactions for all-cause mortality and cardiovascular death were not statistically significant (Table S5). Sensitivity analyses are summarized in Table S1. In the CHARLS cohort, the association with incident hypertension remained directionally consistent after sequentially excluding participants with a history of cardiovascular disease, diabetes, or obesity. In the UKB cohort, analyses excluding outcome events that occurred within the first 2 years of follow-up were conducted to mitigate potential reverse causation, and the associations remained materially unchanged. Additional cross-sectional associations and dose–response relationships for prevalent hypertension in CHARLS are provided in the Supplementary Material (Table S2, Fig. S3, Fig. S4).
Fig. 4.
Subgroup analyses of associations between RCII and primary prognostic outcomes in UK Biobank. Forest plots show fully adjusted hazard ratios comparing the highest with the lowest RCII quartile across prespecified subgroups
Mediation analysis
Exploratory mediation analyses were conducted to evaluate whether selected metabolic and oxidative biomarkers statistically accounted for part of the associations between RCII and clinical outcomes. In the CHARLS cohort (Fig. 5A), the TyG index, a surrogate marker of insulin resistance, and serum uric acid accounted for 35.2% and 13.9% of the association between baseline RCII and incident hypertension, respectively (both P < 0.05). In the UK Biobank cohort (Fig. 5B), HbA1c and serum urate accounted for 6.2% and 5.1% of the association between RCII and major adverse cardiovascular events (MACE), respectively. By contrast, eGFR showed little evidence of an indirect association, with a mediated proportion of −0.3%. Overall, these findings suggest that metabolic and oxidative biomarkers may represent modest indirect pathways linking RCII to hypertension onset and cardiovascular prognosis.
Fig. 5.
Exploratory mediation analyses of potential pathways linking RCII to clinical outcomes. Panel A illustrates possible indirect pathways for incident hypertension in CHARLS, and Panel B illustrates possible indirect pathways for MACE in UK Biobank. These analyses were exploratory and hypothesis-generating and do not establish causal mechanisms
Discussion
In this large-scale, dual-cohort prospective investigation, we evaluated the predictive value of the Remnant Cholesterol-Inflammation Index (RCII) across the hypertension continuum. Our principal findings showed that an elevated RCII was associated with both incident hypertension in CHARLS and adverse long-term cardiovascular outcomes among hypertensive participants in UK Biobank. By integrating lipid metabolic derangements and systemic low-grade inflammation, RCII may capture a dual residual-risk phenotype. Exploratory mediation analyses further suggested possible indirect pathways involving insulin resistance and oxidative stress.
Despite the widespread clinical application of low-density lipoprotein cholesterol (LDL-C)-lowering therapies, a substantial residual cardiovascular risk persists [10]. Accumulating evidence has attributed this residual risk to triglyceride-rich lipoproteins and chronic low-grade inflammation. However, previous studies have largely evaluated these two domains in isolation. Prior evidence suggests that the concurrent elevation of RC and C-reactive protein (CRP) is associated with cardiovascular and mortality risks [11]. Our joint association analysis extends this observation: while isolated elevations in either RC or CRP conferred moderate risk increases, concurrent elevation of both markers (the High RC/High CRP phenotype) was associated with higher risks of incident hypertension and major adverse cardiovascular events (MACE). As a composite metric, RCII quantifies this lipid-inflammatory phenotype and may complement previously established composite indices [12, 13].
The biological plausibility underlying the predictive value of RCII is rooted in the interplay between remnant lipoproteins and the vascular endothelium. Remnant cholesterol (RC), comprising the cholesterol content of triglyceride-rich lipoproteins [5], has emerged as an atherogenic marker associated with incident hypertension [14]. Compared with traditional lipids, RC particles carry a higher cholesterol burden and reflect the atherogenic remnant lipoprotein load [15]. They can penetrate and deposit within the arterial intima and may contribute to endothelial dysfunction. During this process, the localized release of free fatty acids from RC degradation may promote oxidative stress and alter vasomotor regulation [16, 17]. These changes may reduce nitric oxide (NO) bioavailability and interact with the renin–angiotensin–aldosterone system (RAAS), thereby contributing to vascular dysfunction and blood pressure elevation [18, 19]. The dose–response relationship between RCII and incident hypertension observed in our study is consistent with this biological framework.
RC-induced lipotoxicity is closely linked with systemic low-grade inflammation, a key process in cardiovascular pathogenesis [20]. Cholesterol crystals formed after macrophage uptake of remnant particles can activate the NLRP3 inflammasome and promote the release of pro-inflammatory cytokines such as IL-1β and IL-18 [21, 22]. This inflammatory milieu may contribute to vascular smooth muscle cell proliferation, intimal hyperplasia, and endothelial nitric oxide synthase dysfunction [23–25]. Inflammatory cytokines may also interact with RAAS components and vasoconstrictive pathways [26]. Our exploratory mediation analyses suggested a stage-specific pattern. During early hypertension development, systemic metabolic perturbations, including insulin resistance and oxidative stress, appeared to contribute to the association between RCII and hypertension. In the later prognostic setting, the mediated proportions were more modest, suggesting that multiple direct and indirect pathways may be involved.
The findings of this study may have clinical relevance because RCII can be derived from routine and inexpensive laboratory tests. In analyses of predictive performance, adding RCII to traditional, Framingham-like, and ASCVD-like risk-factor models produced small but consistent improvements in Harrell’s C-index, suggesting that RCII may provide complementary residual-risk information beyond established risk-factor frameworks. Recent machine-learning and bioinformatics studies have highlighted the importance of data-driven biomarker discovery and cardiovascular risk prediction [27–29]. Compared with more complex modeling frameworks, RCII represents a simple biomarker-based index that may be easier to implement, although its clinical decision utility requires further external validation, calibration assessment, and decision-curve analysis.
The major strengths of this study include the large dual-cohort design spanning two continents, multivariable adjustment, and sensitivity analyses, which support the robustness of the findings. However, several limitations warrant acknowledgment. First, given the observational nature of the study, causality between RCII and outcomes cannot be definitively established, and the potential for residual confounding remains. Second, RCII was calculated based on a single baseline measurement, precluding the capture of dynamic longitudinal trajectories of lipid and inflammatory profiles. Third, exact hypertension onset dates were unavailable in CHARLS, limiting precise time-to-event modeling of incident hypertension. Fourth, the available UK Biobank analytic file included only a binary indicator of baseline lipid-lowering medication use; information on medication class, statin intensity, PCSK9 inhibitor use, dosage, adherence, duration, and time-updated treatment was unavailable. Fifth, although LDL-C was not calculated by the present investigators using the Friedewald formula, RCII may still be affected by assay variability and the lack of direct remnant-lipoprotein particle measurements. Finally, the predictive efficacy and clinical utility of RCII require further evaluation in additional multi-ethnic cohorts and clinical settings.
Conclusion
The present study showed that RCII was associated with both the onset of hypertension and subsequent cardiovascular complications. By integrating prospective incidence data from a Chinese population with long-term prognostic data from a UK population, our findings suggest that concurrent remnant lipid accumulation and systemic inflammation identify a dual residual-risk phenotype. Exploratory mediation analyses suggested possible indirect pathways involving insulin resistance and oxidative stress.
Incorporating RCII into risk assessment may provide a readily accessible approach to refine risk stratification beyond traditional cardiovascular metrics. However, RCII should be interpreted as a complementary residual-risk marker rather than a stand-alone clinical decision tool, and further validation is required before routine clinical implementation.
Perspectives
In this large, binational cohort study, the Remnant Cholesterol-Inflammation Index (RCII) was associated with both hypertension development and long-term cardiovascular complications. Elevated RCII was consistently associated with increased hypertension incidence and demonstrated dose–response associations with higher risks of all-cause mortality, MACE, and cardiovascular death. The concurrent elevation of remnant cholesterol and systemic inflammation was associated with higher MACE risk. These findings support the potential utility of RCII in identifying dual residual risk and highlight the need for further validation of lipid-inflammatory risk stratification.
Supplementary Information
Abbreviations
- RCII
Remnant Cholesterol-Inflammation Index
- CHARLS
China Health and Retirement Longitudinal Study
- UKB
UK Biobank
- RC
Remnant Cholesterol
- MACE
Major Adverse Cardiovascular Events
- TyG
Triglyceride-Glucose Index
- RCS
Restricted Cubic Spline
- ASCVD
Atherosclerotic cardiovascular disease
- PCE
Pooled Cohort Equations
Authors’ contributions
Lerui Wang and Siyuan Wen contributed equally to this work. Boda Zhou had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: Lerui Wang, Zhiyi Ma, Boda Zhou. Acquisition, analysis, or interpretation of data: All authors. Drafting of the manuscript: Lerui Wang, Siyuan Wen, Zhiyi Ma, Boda Zhou. Critical review of the manuscript for important intellectual content: All authors. Statistical analysis: Lerui Wang. Administrative, technical, or material support: Zhiyi Ma, Boda Zhou. Supervision: Zhiyi Ma, Boda Zhou.
Funding
This work was supported by Tsinghua University (ZY01_07 and 2611T0812), and Beijing Municipal Health Commission (2024-3-034).
Data availability
The datasets analyzed during the current study are available in the CHARLS repository for researchers who meet the criteria for access to confidential data. The datasets analyzed during the current study are available in the CHARLS repository (http://charls.pku.edu.cn/index/en.html) and the UK Biobank resource (https://www.ukbiobank.ac.uk/) for researchers who meet the criteria for access to confidential data.
Declarations
Ethics approval and consent to participate
The CHARLS and UKB studies were conducted in accordance with the Declaration of Helsinki. The CHARLS protocol was approved by the Ethical Review Committee of Peking University (IRB00001052-11015). The UKB study received ethical approval from the North West Multi-centre Research Ethics Committee (11/NW/0382). Written informed consent was obtained from all participants prior to data collection.
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.
Lerui Wang and Siyuan Wen contributed equally to this work.
Contributor Information
Zhiyi Ma, Email: mzya03330@btch.edu.cn.
Boda Zhou, Email: zhouboda@tsinghua.edu.cn.
References
- 1.Martin SS, Aday AW, Allen NB, Almarzooq ZI, Anderson CAM, Arora P, et al. 2025 heart disease and stroke statistics: A report of us and global data from the american heart association. Circulation. 2025;151:e41–660. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Cheraghi Z, Azmi-Naei B, Cheraghi P, Doosti-Irani A. The global prevalence of uncontrolled hypertension: a systematic review and meta-analysis. BMC Public Health. 2025;25:4259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Stone PH, Libby P, Boden WE. Fundamental pathobiology of coronary atherosclerosis and clinical implications for chronic ischemic heart disease management-the plaque hypothesis: a narrative review. JAMA Cardiol. 2023;8:192–201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rosenson RS, Shaik A, Song W. New therapies for lowering triglyceride-rich lipoproteins: Jacc focus seminar 3/4. J Am Coll Cardiol. 2021;78:1817–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Burnett JR, Hooper AJ, Hegele RA. Remnant cholesterol and atherosclerotic cardiovascular disease risk. J Am Coll Cardiol. 2020;76:2736–9. [DOI] [PubMed] [Google Scholar]
- 6.Chen Y, Ye X, Escames G, Lei W, Zhang X, Li M, et al. The nlrp3 inflammasome: contributions to inflammation-related diseases. Cell Mol Biol Lett. 2023;28:51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kong P, Cui ZY, Huang XF, Zhang DD, Guo RJ, Han M. Inflammation and atherosclerosis: Signaling pathways and therapeutic intervention. Signal Transduct Target Ther. 2022;7:131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chen H, Wu JY, Yan H, Gao J, Li C, Xie JH, et al. Association of remnant cholesterol inflammatory index with stroke, heart disease and all-cause mortality across cardiovascular-kidney-metabolic syndrome stages 0-3: a national cohort study. Nutrients. 2026;18:205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bian F, Yang X, Zhou F, Wu PH, Xing S, Xu G, et al. C-reactive protein promotes atherosclerosis by increasing ldl transcytosis across endothelial cells. Br J Pharmacol. 2014;171:2671–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Yang L, Yue Q, Fang F, Zhang Y, Liu P, Zhang Z, et al. Effect of dual residual risk of cholesterol and inflammation on all-cause mortality in patients with cardiovascular disease. Cardiovasc Diabetol. 2023;22:96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Doi T, Langsted A, Nordestgaard BG. Dual elevated remnant cholesterol and c-reactive protein in myocardial infarction, atherosclerotic cardiovascular disease, and mortality. Atherosclerosis. 2023;379:117141. [DOI] [PubMed] [Google Scholar]
- 12.Claudel SE, Schmidt IM, Waikar SS, Verma A. Cumulative incidence of mortality associated with cardiovascular-kidney-metabolic (CKM) syndrome. J Am Soc Nephrol. 2025;36:1343–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.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. 2025;76:543–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wang F, Li Q. Association between residual cholesterol and vulnerable non-culprit lesions progressing to major adverse cardiovascular events. Front Endocrinol (Lausanne). 2025;16:1603907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Navarese EP, Vine D, Proctor S, Grzelakowska K, Berti S, Kubica J, et al. Independent causal effect of remnant cholesterol on atherosclerotic cardiovascular outcomes: a mendelian randomization study. Arterioscler Thromb Vasc Biol. 2023;43:e373–80. [DOI] [PubMed] [Google Scholar]
- 16.Miller YI, Choi SH, Fang L, Tsimikas S. Lipoprotein modification and macrophage uptake: role of pathologic cholesterol transport in atherogenesis. Subcell Biochem. 2010;51:229–51. [DOI] [PubMed] [Google Scholar]
- 17.Chen X, Li LH. Remnant cholesterol, a valuable biomarker for assessing arteriosclerosis and cardiovascular risk: a systematic review. Cureus. 2023;15:e44202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lin A, Nerlekar N, Rajagopalan A, Yuvaraj J, Modi R, Mirzaee S, et al. Remnant cholesterol and coronary atherosclerotic plaque burden assessed by computed tomography coronary angiography. Atherosclerosis. 2019;284:24–30. [DOI] [PubMed] [Google Scholar]
- 19.Kunimura A, Miura K. Is remnant cholesterol a new therapeutic target for preventing hypertension? Hypertens Res. 2024;47:1227–8. [DOI] [PubMed] [Google Scholar]
- 20.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;79:837–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sheedy FJ, Grebe A, Rayner KJ, Kalantari P, Ramkhelawon B, Carpenter SB, et al. Cd36 coordinates nlrp3 inflammasome activation by facilitating intracellular nucleation of soluble ligands into particulate ligands in sterile inflammation. Nat Immunol. 2013;14:812–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Balashkevich N, Kazymov M, Syzdykbayev M, Adylova A. Molecular basis of acute coronary syndrome. J Res Med Sci. 2022;27:40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Guzik TJ, Nosalski R, Maffia P, Drummond GR. Immune and inflammatory mechanisms in hypertension. Nat Rev Cardiol. 2024;21:396–416. [DOI] [PubMed] [Google Scholar]
- 24.Prabhu SD, Frangogiannis NG. The biological basis for cardiac repair after myocardial infarction: From inflammation to fibrosis. Circ Res. 2016;119:91–112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Shen Y, Dong Z, Fan F, Li K, Zhu S, Dai R, et al. Targeting cytokine-like protein FAM3D lowers blood pressure in hypertension. Cell Rep Med. 2023;4:101072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Patel S, Rauf A, Khan H, Abu-Izneid T. Renin-angiotensin-aldosterone (RAAS): the ubiquitous system for homeostasis and pathologies. Biomed Pharmacother. 2017;94:317–25. [DOI] [PubMed] [Google Scholar]
- 27.Li G, Luo C, Ge T, He K, Zhang M, Hu J, et al. Evaluating the impact of metabolic indicators and scores on cardiovascular events using machine learning. Diabetol Metab Syndr. 2025;17:180. 10.1186/s13098-025-01753-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Li G, Zhang T, He K, Zhang M, Hu J, Ge T, et al. Deciphering the influence of heavy metals on adverse cardiovascular events using machine learning. Toxicol Lett. 2025;411:101. 10.1016/j.toxlet.2025.07.1414. [DOI] [PubMed] [Google Scholar]
- 29.Li G, Lin D, Fan X, Peng B. Exploring hypertrophic cardiomyopathy biomarkers through integrated bioinformatics analysis: uncovering novel diagnostic candidates. Cardiol Res Pract. 2024;2024:4639334. 10.1155/2024/4639334. [DOI] [PMC free article] [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 analyzed during the current study are available in the CHARLS repository for researchers who meet the criteria for access to confidential data. The datasets analyzed during the current study are available in the CHARLS repository (http://charls.pku.edu.cn/index/en.html) and the UK Biobank resource (https://www.ukbiobank.ac.uk/) for researchers who meet the criteria for access to confidential data.