Abstract
BACKGROUND
Elevated lipoprotein(a) [Lp(a)] is associated with atherosclerotic cardiovascular disease (ASCVD) risk, and vascular inflammation is one mechanism through which Lp(a) causes ASCVD.
OBJECTIVES
The authors aimed to evaluate whether interleukin-6 (IL-6), a biomarker associated with inflammation and cardiovascular disease, helps risk-stratify individuals with elevated Lp(a).
METHODS
Data from participants in the MESA (Multi-Ethnic Study of Atherosclerosis) (n = 6,514) and the UK Biobank (UKB) (n = 26,574) were used for this analysis. The associations between Lp(a) and IL-6 with coronary heart disease (CHD) (defined as myocardial infarction or resuscitated cardiac arrest), ASCVD (CHD and ischemic stroke), and peripheral vascular disease (PVD) were evaluated separately and with mutual adjustment in Cox proportional hazard models adjusted for traditional cardiovascular risk factors and high-sensitivity C-reactive protein (hsCRP). HRs were presented per standard deviation. Participants were also grouped by Lp(a) level (≤50 or >50 mg/dL [125 nmol/L]) and IL-6 level (≤ median or > median) in similar models.
RESULTS
Participants with higher IL-6 levels were more likely to have higher body mass index, systolic blood pressure, triglycerides, and hsCRP with lower high-density lipoprotein cholesterol. Lp(a) (HR: 1.13; 95% CI: 1.04–1.23 in MESA; HR: 1.11; 95% CI: 1.09–1.13 in UKB) and IL-6 (HR: 1.22; 95% CI: 1.10–1.35 in MESA; HR: 1.19; 95% CI: 1.15–1.24 in UKB) were both independently associated with CHD events when evaluated separately. When evaluated together, no significant change was noted, and interaction testing was not significant. Similar results were seen for ASCVD and PVD. When participants were categorized by both Lp(a) and IL-6 levels, the strongest association for each outcome was noted when both levels were high (for CHD: HR: 1.72; 95% CI: 1.25–2.36 in MESA; HR: 1.39; 95% CI: 1.12–1.72 in UKB).
CONCLUSIONS
In 2 independent primary prevention cohorts, Lp(a) and IL-6 were independent predictors of ASCVD risk, and their combination identified individuals at highest risk.
Keywords: inflammation, lipids, Lp(a), prevention, risk factors
CENTRAL ILLUSTRATION
Lp(a), IL-6, and Cardiovascular Risk
(Top) Interplay between Lp(a) and inflammatory pathways. Lp(a) is predominantly genetically determined by the LPA gene, leading to production of apolipoprotein (a) and ultimately assemble Lp(a). Inflammation leads to increased production of IL-6, which binds to response elements in the liver to increase CRP and apolipoprotein(a) synthesis. In individuals with genetically elevated Lp(a) levels, IL-6 stimulus may raise levels further. (Bottom) Findings from the present study. In 2 independent primary prevention cohorts, the association between Lp(a), IL-6, and cardiovascular outcomes was evaluated. Both Lp(a) and IL-6 were independently associated with CHD, ASCVD, and PVD in multivariable-adjusted models, including with adjusting for high-sensitivity CRP. When evaluated together, the greatest risk was noted when both were at elevated levels. Created with the use of BioRender.com. ASCVD = atherosclerotic cardiovascular disease; CHD = coronary heart disease; CRP = C-reactive protein; IL = interleukin; Lp(a) = lipoprotein(a); OxPL = oxidized phospholipids; PVD = peripheral vascular disease.
Elevated lipoprotein(a) [Lp(a)] is a common genetically determined risk factor for atherosclerotic cardiovascular disease (ASCVD), with levels associated with significantly increased risk present in ~20% of the global population. Lp(a) leads to ASCVD through multiple pathways, including the development of atherosclerosis, vascular inflammation, and potential antifibrinolytic mechanisms.1 Although the association between Lp(a) and cardiovascular risk is continuous at the population level,2 there is heterogeneity in terms of which individuals manifest clinical disease in clinical practice, and there is a need to better risk-stratify individuals with elevated Lp(a).
This heterogeneity may be explained in part by the interplay between Lp(a) and inflammation. Lp(a) is the predominant lipoprotein carrier of proinflammatory oxidized phospholipids on apolipoprotein B-100 (OxPL-ApoB), which are strongly associated with coronary artery disease.3,4 A recent analysis of the LoDoCo2 trial noted a greater absolute risk reduction with the antiinflammatory colchicine in individuals with higher Lp(a) and OxPL-ApoB levels.5 Interleukin (IL)-6 is a cytokine that has been causally associated with cardiovascular disease by means of mendelian randomization.6 In addition, acute increases in IL-6 up-regulate the LPA gene to produce apolipoprotein(a) leading to an increase in Lp(a) levels,7 and targeted inhibition of IL-6 signaling reduces Lp(a).8,9 In another analysis of the LoDoCo2 trial, in a secondary prevention setting, IL-6 modified the association between Lp(a) and ASCVD risk.10 Taken together, these findings support a role for inflammation in the pathophysiology of Lp(a)-mediated cardiovascular disease.
However, it remains unclear if markers of vascular inflammation modify the association between Lp(a) and cardiovascular risk in a clinically meaningful way in a primary prevention setting. Studies evaluating high-sensitivity C-reactive protein (hsCRP) have yielded conflicting results, and more recent meta-analyses suggest that hsCRP does not modify Lp(a)mediated risk.11–13 IL-6, however, is upstream of hsCRP,14 is more specific to pathways involved in atherosclerosis,10 and is associated with Lp(a) levels,7 and thus IL-6 may better reflect inflammatory risk associated with elevated Lp(a). We aimed to address whether IL-6 modifies the association between Lp(a) and cardiovascular risk, and whether IL-6 can be used to further risk-stratify individuals with elevated Lp(a) in a study of 2 independent primary prevention cohorts.
METHODS
STUDY POPULATION.
Data from participants from the MESA (Multi-Ethnic Study of Atherosclerosis) and the UK Biobank (UKB) were used. The design of MESA has been described previously.15 Briefly, MESA is a prospective cohort study of individuals without known baseline cardiovascular disease recruited from 6 field centers across the United States and followed for the development of cardiovascular events. Recruitment occurred from 2000 to 2002, and adjudicated cardiovascular events are available through 2018. MESA was approved by the institutional review boards of each field center, and all participants provided informed written consent. For the present study, all participants with available baseline Lp(a) and IL-6 measurements were included, and all participants with missing follow-up were excluded. The design of UKB has been previously described. Briefly, it is a large biomedical database that has been collecting data on >500,000 participants since 2006. Individuals living in the UK aged 40 to 69 years were enrolled and regularly provide biological samples. In addition, hospital admissions data are linked on a participant basis allowing for ascertainment of outcomes of interest. All individuals provided consent before enrolling in UKB. Individuals without prior ASCVD and with available covariate data were included for the present study.
COVARIATES AND OUTCOMES.
At recruitment, data were collected on demographics, comorbid medical conditions, and physical measures through questionnaires, physical examinations, and laboratory measurements on blood samples for both studies. Lp(a) was measured with the use of a latex-enhanced turbidimetric immunoassay (MESA: Denka Seiken; r = 0.91 compared with Lp(a) molar assay16; UKB: Randox Bioscience; r = 0.995 compared with other commercially available assays) and reported in mg/dL (MESA) or nmol/L (UKB). In MESA, IL-6 was measured with the use of an enzyme-linked immunosorbent assay (Quantikine HS Human IL6 Immunoassay17). In UKB, IL-6 was measured at Olink Proteomics (Uppsala, Sweden) by means of their proximity extension assay, an antibody-based immunoassay using polymerase chain reaction amplification for detection using previously described methods, and the vast majority of participants were randomly selected and similar to the overall UKB population18; results are reported in normalized protein expression, which cannot be converted to standard units. For this study, outcomes were coronary heart disease (CHD) (defined as myocardial infarction, resuscitated cardiac arrest or CHD death), ASCVD (defined as CHD and ischemic stroke), and peripheral vascular disease (PVD) (defined as symptomatic disease including lower extremity claudication, atherosclerosis of the lower extremity, arterial embolism or thrombosis of the lower extremity, and abdominal aortic aneurysm). UKB outcomes were constructed from International Classification of Diseases-10th edition (ICD-10) codes and death registry data using the earliest recorded occurrence of each respective ICD-10 code; myocardial infarction was defined as I21-I23, I24.1, I25.2; stroke was defined was I60-I61, I63-I64; cardiovascular death was defined as I20-I25 as cause of death. In UKB, only CHD and ASCVD were evaluated because data for PVD were not being readily available.
STATISTICAL METHODS.
Baseline characteristics were compared by quartile of IL-6. Continuous variables were compared by means of analysis of variance or Kruskal-Wallis testing as appropriate, and categoric variables were compared by means of chi-square tests. Baseline characteristics for those lost to follow-up (defined as those without an event who did not die with follow-up less than the median of 6,113 days) were compared with those not lost to follow-up. IL-6 and Lp(a) levels by race or ethnicity were visualized with the use of box plots. The associations between Lp(a) and IL-6 with each outcome were evaluated with the use of Cox proportional hazards regression models, first evaluating each biomarker separately with adjustment for age, sex, race/ethnicity, total cholesterol, high-density lipoprotein cholesterol, body mass index, systolic blood pressure, diabetes status, cigarette smoking status, medications for hypertension, and statin use. Lp(a), IL-6, and hsCRP were then evaluated in similar models with mutual adjustment, and the multiplicative interactions between IL-6 and Lp(a) and between IL-6 and hsCRP were tested with the use of continuous variables. Relative excess risk due to interaction (RERI) also was calculated for Lp(a) (cutoff of 50 mg/dL or 125 nmol/L) and IL-6 (cutoff of median) as well as for IL-6 and hsCRP (cutoff of 2 mg/dL) in the above models. Analyses were conducted also for the association between IL-6 and events, stratified by Lp(a) (cutoff of 50 mg/dL or 125 nmol/L) or hsCRP (cutoff of 2 mg/dL). Multiplicative interaction testing between IL-6 as a continuous variable and Lp(a) and hsCRP using the above categories was performed. For each biomarker, hazard ratios per standard deviation are presented. IL-6 and hsCRP were ln transformed because of nonlinearity in MESA. Lp(a) and hsCRP were ln transformed because of nonlinearity in UKB. The linear association between IL-6, Lp(a), and events was subsequently confirmed by visualizing the Mattingale residuals. In sensitivity analyses, the primary analysis was conducted with additional adjustment for education, health insurance status, depressive symptoms, and smoking pack-years and cigarettes per day instead of smoking status in MESA. An analysis with additional adjustment for family history of myocardial infarction in a first degree relative also was performed in MESA. A falsification endpoint analysis also was performed evaluating hemorrhagic stroke as an outcome using the same methodology as the primary analysis.
Participants were then categorized by Lp(a) (≤50 or >50 mg/dL in MESA, ≤125 or >125 nmol/L in UKB) and IL-6 (≤ or > the median), and the association with each outcome was evaluated in similar models as above. Participants were also categorized by IL-6 (≤ or > the median) and hsCRP (≤2 or >2 mg/dL) for a similar analysis. Additional thresholds (Lp(a) of 70 mg/dL or 175 nmol/L, IL-6 in quartile 4 vs quartiles 1–3, and hsCRP of 3 mg/dL) also were evaluated. Multiplicative interaction testing between IL-6, Lp(a), and hsCRP using the above categories was performed. Finally, improvement in predictive value with the addition of Lp(a), IL-6, and hsCRP to a model including the pooled cohort equations (PCEs), body mass index, and statin use was evaluated according to category-free net reclassification improvement (NRI) and Harrell’s C-index. Both cohorts used similar definitions for exposure, outcomes and covariates, but analyses were run separately by design. All analyses were performed with the use of R (version 4.3.1). Two-tailed P < 0.05 was considered to be statistically significant. The proportional hazards assumption was confirmed by means of Schoenfeld residuals.
RESULTS
The study included 6,514 individuals from MESA (mean age 62.1 years, 52.9% female, 38.9% White, 27.1% Black, 22.1% Hispanic, 12.0% Chinese, median follow-up 16.8 years [Q1–Q3: 12.5–17.5 years]) after excluding those with missing data for Lp(a) (n = 114) or IL-6 (n = 160) or without follow-up (n = 26). For the UKB analyses, 26,574 individuals (mean age 57.0 years, 55.0% female, 94.1% White, 1.8% Asian, 4.8% other, median follow-up 13.6 years [Q1–Q3: 12.9–14.2 years]) were included after excluding those with missing covariates (n = 474,450) or prior ASCVD (n = 1,168). For MESA, median IL-6 was 1.21 pg/mL [Q1–Q3: 0.77–1.88 pg/mL]. With increasing quartile of IL-6, there was generally a greater burden of cardiovascular risk factors in both cohorts including increasing age, hypertension, diabetes, and current smoking. Participants with higher IL-6 levels were more likely to have higher body mass index, systolic blood pressure, and triglyceride, high-density lipoprotein cholesterol, and hsCRP levels. Nonlinear trends for total cholesterol and low-density lipoprotein-cholesterol (LDL-C) were observed. No significant differences in Lp(a) levels by IL-6 quartile were noted in MESA, however, there was a small but significant increase in Lp(a) levels by IL-6 quartile in UKB participants (Table 1). In MESA, those lost to follow-up were more likely to be of non-White race, and less likely to have hypertension or active smoking status. There was a very weak correlation between IL-6 and Lp(a) in MESA (ρ = 0.02; P = 0.049) and UKB (ρ = 0.03; P < 0.001) and a modest correlation between IL-6 and hsCRP in MESA (ρ = 0.53; P < 0.001) and UKB (ρ = 0.55; P < 0.001) (Supplemental Figure 1). In those both with and without a subsequent event, the correlation between Lp(a) and IL-6 was weak in both MESA (ρ = 0.02 [P = 0.103] and ρ = 0.04 [P = 0.301], respectively) and UKB (ρ = 0.03 [P < 0.001] and ρ = 0.04 [P = 0.089], respectively). The correlation between IL-6 and hsCRP also was similar in those without and with events in both MESA (rho 0.54 [P < 0.001] and rho 0.50 [P < 0.001], respectively) and UKB (rho 0.54 [P < 0.001] and rho 0.55 [P < 0.001], respectively). There was significant variation in IL-6 and Lp(a) levels by race and ethnicity (Supplemental Figure 2).
TABLE 1.
Study Population Characteristics by Quartile of IL-6
| MESA, IL-6 in pg/mL | ||||||
|---|---|---|---|---|---|---|
| Overall (N = 6,514) | Quartile 1 (0.13–0.77) (n = 1,629) | Quartile 2 (0.77–1.21) (n = 1,628) | Quartile 3 (1.21–1.88) (n = 1,628) | Quartile 4 (1.88–13.00) (n = 1,629) | P Value | |
| Age, y | 62.06 ± 10.25 | 58.66 ± 9.48 | 62.07 ± 10.03 | 63.56 ± 10.39 | 63.94 ± 10.24 | <0.001a |
| Female | 3,449 (52.9) | 775 (47.6) | 830 (51.0) | 915 (56.2) | 929 (57.0) | <0.001a |
| Race/ethnicity | <0.001a | |||||
| Black | 1,764 (27.1) | 329 (20.2) | 423 (26.0) | 484 (29.7) | 528 (32.4) | |
| Chinese | 781 (12.0) | 340 (20.9) | 216 (13.3) | 120 (7.4) | 105 (6.4) | |
| Hispanic | 1,438 (22.1) | 254 (15.6) | 329 (20.2) | 409 (25.1) | 446 (27.4) | |
| White | 2,531 (38.9) | 706 (43.3) | 660 (40.5) | 615 (37.8) | 550 (33.8) | |
| Hypertension | 2,905 (44.6) | 489 (30.0) | 692 (42.5) | 845 (51.9) | 879 (54.0) | <0.001a |
| Diabetes | 810 (12.4) | 107 (6.6) | 181 (11.1) | 228 (14.0) | 294 (18.1) | <0.001a |
| Current smoking | 833 (12.8) | 169 (10.4) | 185 (11.4) | 212 (13.1) | 267 (16.5) | <0.001a |
| Hypertension medication use | 2,407 (37.0) | 405 (24.9) | 547 (33.6) | 713 (43.8) | 742 (45.5) | <0.001a |
| Statin use | 962 (14.8) | 220 (13.5) | 258 (15.9) | 244 (15.0) | 240 (14.8) | 0.298 |
| Body mass index, kg/m2 | 28.32 ± 5.45 | 25.44 ± 3.78 | 27.54 ± 4.38 | 29.23 ± 5.12 | 31.05 ± 6.48 | <0.001a |
| Systolic blood pressure, mm Hg | 126.41 ± 21.48 | 120.22 ± 19.82 | 126.15 ± 20.63 | 128.88 ± 21.71 | 130.41 ± 22.26 | <0.001a |
| hsCRP, mg/L | 1.90 (0.83–4.19) | 0.83 (0.44–1.58) | 1.49 (0.75–2.82) | 2.60 (1.25–4.74) | 4.44 (2.15–9.18) | <0.001a |
| Total cholesterol, mg/dL | 194.21 ± 35.68 | 195.88 ± 33.92 | 197.42 ± 36.07 | 194.52 ± 36.35 | 189.03 ± 35.79 | <0.001a |
| HDL cholesterol, mg/dL | 50.98 ± 14.77 | 54.28 ± 15.74 | 51.10 ± 14.39 | 50.13 ± 14.30 | 48.40 ± 13.99 | <0.001a |
| LDL cholesterol, mg/dL | 117.22 ± 31.41 | 118.13 ± 30.74 | 119.49 ± 31.09 | 117.39 ± 31.45 | 113.85 ± 32.10 | <0.001a |
| Triglycerides, mg/dL | 131.66 ± 89.08 | 117.67 ± 69.95 | 135.77 ± 96.71 | 136.17 ± 91.66 | 137.05 ± 94.08 | <0.001a |
| Lp(a), mg/dL | 17.10 (7.43–40.10) | 15.60 (7.60–37.10) | 17.90 (7.60–42.32) | 17.15 (7.40–41.87) | 18.10 (7.20–40.10) | 0.287 |
| IL-6, pg/mL | 1.21 (0.77–1.88) | 0.58 (0.47–0.68) | 0.97 (0.87–1.07) | 1.49 (1.33–1.67) | 2.69 (2.19–3.64) | <0.001a |
| UKB, IL-6 in Normalized Unitless Values | ||||||
|---|---|---|---|---|---|---|
| Overall (N = 33,203) | Quartile 1 (n = 8,301) | Quartile 2 (n = 8,02) | Quartile 3 (n = 8,301) | Quartile 4 (n = 8,299) | P Value | |
| Age, y | 57.0 ± 8.17 | 54.1 ± 8.21 | 56.9 ± 8.15 | 58.2 ± 7.84 | 58.9 ± 7.61 | <0.001a |
| Female | 1,8250 (55.0) | 4,726 (56.9) | 4,466 (53.8) | 4,483 (54.0) | 4,575 (55.1) | <0.001a |
| Race/ethnicity | 0.182 | |||||
| Asian | 583 (1.8) | 134 (1.6) | 129 (1.6) | 148 (1.8) | 172 (2.1) | |
| White | 3,1238 (94.1) | 7,823 (94.2) | 7,840 (94.4) | 7,821 (94.2) | 7,754 (93.4) | |
| Other | 1,382 (4.2) | 344 (4.1) | 333 (4.0) | 332 (4.0) | 373 (4.5) | |
| Hypertension | 22,702 (68.4%) | 4,497 (54.2%) | 5,683 (68.5%) | 6,176 (74.4%) | 6,346 (76.5%) | <0.001a |
| Diabetes | 1,379 (4.2%) | 145 (1.7%) | 234 (2.8%) | 374 (4.5%) | 626 (7.5%) | <0.001a |
| Current smoking | 1,379 (4.2%) | 707 (8.5%) | 825 (9.9%) | 865 (10.4%) | 1,138 (13.7%) | <0.001a |
| Hypertension medication use | 3,676 (11.1%) | 495 (6.0%) | 848 (10.2%) | 1,080 (13.0%) | 1,253 (15.1%) | <0.001a |
| Statin use | 3,342 (10.1%) | 814 (9.8%) | 974 (11.7%) | 1,024 (12.3%) | 1,024 (12.3%) | <0.001a |
| Body mass index, kg/m2 | 27.5 ± 4.77 | 25.0 ± 3.35 | 26.7 ± 3.81 | 28.2 ± 4.43 | 30.0 ± 5.71 | <0.001a |
| Systolic blood pressure, mm Hg | 138 ± 18.6 | 133 ± 17.8 | 138 ± 18.4 | 140 ± 18.3 | 141 ± 18.8 | <0.0012 |
| hsCRP, mg/L | 2.70 (0.68, 2.88) | 1.03 (0.38–1.22) | 1.60 (0.62–1.94) | 2.53 (0.97–3.06) | 5.63 (1.65–6.53) | <0.001a |
| Total cholesterol, mg/dL | 220 ± 44.3 | 219 ± 41.6 | 223 ± 44.0 | 222 ± 45.2 | 216 ± 45.9 | <0.001a |
| HDL cholesterol, mg/dL | 56.1 ± 14.8 | 60.5 ± 15.1 | 57.3 ± 14.8 | 54.4 ± 14.0 | 52.2 ± 13.9 | <0.001a |
| LDL cholesterol, mg/dL | 138 ± 33.7 | 135 ± 31.9 | 139 ± 33.5 | 140 ± 34.4 | 136 ± 34.7 | <0.001a |
| Triglycerides, mg/dL | 155 ± 89.8 | 129 ± 75.1 | 150 ± 86.5 | 167 ± 93.0 | 173 ± 96.6 | <0.001a |
| Lp(a), nmol/L | 44.9 (9.6–62.5) | 43.3 (9.3–59.9) | 44.1 (9.6–60.6) | 45.9 (9.7–64.7) | 46.1 (9.8–65.4) | <0.001a |
| IL-6, normalized unitless values | NA | NA | NA | NA | NA | |
Values are mean ± SD, n (%), median (Q1–Q3).
P < 0.05.
HDL = high-density lipoprotein; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6; LDL = low-density lipoprotein; Lp(a) = lipoprotein(a); MESA = Multi-Ethnic Study of Atherosclerosis; UKB = UK Biobank.
In MESA there were 493 (7.6%) CHD, 725 (11.1%) ASCVD, and 118 (1.8%) PVD events in 6,514 participants. In UKB there were 1,438 (5.4%) CHD and 2,115 (8.0%) ASCVD events in 26,574 participants. In multivariable-adjusted models, Lp(a) (HR: 1.13 [95% CI: 1.04–1.23] in MESA; HR per SD: 1.11 [95% CI: 1.09–1.13] in UKB) and IL-6 (HR per SD: 1.22 [95% CI: 1.10–1.35] in MESA; HR per SD: 1.1;[95% CI: 1.15–1.24] in UKB) were both independently associated with CHD events when evaluated separately. When evaluated together, both remained statistically significant without a significant multiplicative interaction, evaluating both continuously (MESA: P = 0.117; UKB: P = 0.577). Similar results were seen for ASCVD, and higher point estimates were noted for PVD. There was no meaningful difference in results with and without adjustment for hsCRP (Table 2). Additional multivariable adjustment for health insurance status, education level and depressive symptoms in MESA did not alter the results and so were not included in the final models. In addition, adjustment for smoking pack-years and cigarettes per day instead of smoking status did not meaningfully alter the results in MESA, so smoking status was used in the final models owing to missingness of data for pack-years in UKB. Finally, additional adjustment for family history of myocardial infarction in MESA did not alter the overall results. There was no significant multiplicative interaction between IL-6 and hsCRP for any outcome, when evaluated continuously (Table 2). RERI was not significant for CHD (−0.31 [95% CI −1.06 to 0.34] in MESA, 0.00 [95% CI −0.47 to 0.43] in UKB), ASCVD (−0.24 [95% CI −0.8 to 0.25] in MESA, 0.03 [95% CI −0.34 to 0.38] in UKB) or PVD (0.21 [95% CI −1.59 to 1.75] in MESA) when evaluating the interaction between Lp(a) and IL-6. For IL-6 and hsCRP there was a borderline significant RERI for CHD, suggesting antagonism in MESA (−0.46; 95% CI −1.05 to −0.01) but not in UKB (0.22; 95% CI −0.12 to 0.50), with nonsignificant RERI for ASCVD and PVD. In a falsification endpoint analysis, there was no significant association between IL-6 or Lp(a) with hemorrhagic stroke.
TABLE 2.
Association of IL-6 and Lp(a) With Cardiovascular Events
| CHD | ||||
|---|---|---|---|---|
| MESA (493 Events/n = 6,514) | UKB (1,438 Events/n = 26,583) | |||
| HR (95% CI) | P Value | HR (95% CI) | P Value | |
| Individual biomarkers | ||||
| IL-6 | 1.22 (1.10–1.35) | 0.001 | 1.19 (1.15–1.24) | <0.001 |
| Lp(a) | 1.13 (1.04–1.23) | 0.006 | 1.11 (1.09–1.13) | <0.001 |
| Mutually adjusted | ||||
| IL-6 | 1.22 (1.10–1.35) | <0.001 | 1.23 (1.17–1.29) | <0.001 |
| Lp(a) | 1.13 (1.03–1.23) | 0.007 | 1.05 (1.00–1.10) | 0.043 |
| Mutually adjusted + hsCRP | ||||
| IL-6 | 1.25 (1.12–1.40) | <0.001 | 1.15 (1.09–1.22) | <0.001 |
| Lp(a) | 1.12 (1.03–1.22) | 0.008 | 1.06 (1.01–1.12) | 0.032 |
| hsCRP | 0.93 (0.83–1.04) | 0.222 | 1.14 (1.07–1.23) | <0.001 |
| IL-6 × Lp(a) P for interaction | 0.117 | 0.577 | ||
| IL-6 × hsCRP P for interaction | 0.987 | 0.864 | ||
| ASCVD | PVD | |||||
|---|---|---|---|---|---|---|
| MESA (725 Events/n = 6,514) | UKB (2,115 Events/n = 26,583) | MESA (118 Events/n = 6,514) | ||||
| HR (95% CI) | P Value | HR (95% CI) | P Value | HR (95% CI) | P Value | |
| Individual biomarkers | ||||||
| IL-6 | 1.22 (1.12–1.32) | <0.001 | 1.17 (1.13–1.21) | <0.001 | 1.61 (1.32–1.95) | <0.001 |
| Lp(a) | 1.08 (1.00–1.16) | 0.049 | 1.08 (1.06–1.09) | <0.001 | 1.22 (1.05–1.41) | 0.010 |
| Mutually adjusted | ||||||
| IL-6 | 1.21 (1.12–1.32) | <0.001 | 1.20 (1.15–1.25) | <0.001 | 1.61 (1.32–1.95) | <0.001 |
| Lp(a) | 1.07 (1.00–1.16) | 0.060 | 1.04 (1.00–1.08) | 0.062 | 1.21 (1.04–1.40) | 0.012 |
| Mutually adjusted + hsCRP | ||||||
| IL-6 | 1.22 (1.11–1.33) | <0.001 | 1.14 (1.08–1.19) | <0.001 | 1.58 (1.27–1.96) | <0.001 |
| Lp(a) | 1.07 (1.00–1.15) | 0.064 | 1.04 (1.00–1.09) | 0.055 | 1.21 (1.04–1.40) | 0.012 |
| hsCRP | 0.99 (0.90–1.09) | 0.861 | 1.11 (1.05–1.18) | <0.001 | 1.04 (0.83–1.30) | 0.732 |
| IL-6 × Lp(a) P for interaction | 0.266 | 0.485 | 0.544 | |||
| IL-6 × hsCRP P for interaction | 0.740 | 0.763 | 0.988 | |||
Adjusted for age, sex, race/ethnicity, total cholesterol, HDL cholesterol, body mass index, systolic blood pressure, diabetes mellitus, cigarette smoking, hypertension medications, statin use. HRs are presented per SD. IL-6 and hsCRP were ln transformed in MESA owing to nonlinearity. SDs were 33.87 mg/dL for Lp(a), 1.22 pg/mL for IL-6, and 5.23 mg/dL for hsCRP. SD for ln-transformed IL-6 was 0.67 pg/mL and for hsCRP was 1.15 mg/dL. Lp(a) and hsCRP were ln transformed owing to nonlinearity in UKB. SDs were 1.12 nmol/L for ln(Lp(a)), 0.89 units for IL-6, and 1.07 mg/dL for ln(CRP). P values for interaction are presented for multiplicative interaction between IL-6 and Lp(a) or hsCRP as continuous variables.
ASCVD = atherosclerotic cardiovascular disease; CHD = coronary heart disease; PVD = peripheral vascular disease; other abbreviations as in Table 1.
When stratified by Lp(a) of 50 mg/dL (MESA) or 125 nmol/L (UKB), the association between IL-6 per SD and CHD, ASCVD, and PVD was similar regardless of Lp(a) level. No significant moderation was noted between IL-6 and Lp(a) >50 mg/dL or Lp(a) >125 nmol/L for any outcome, evaluating IL-6 continuously and Lp(a) categorically (Supplemental Figure 3). Similarly, the association between IL-6 and CHD, ASCVD, and PVD was similar regardless of hsCRP level (using a threshold of 2 mg/L). Interaction testing was negative except for IL-6 and hsCRP >2 mg/L in UKB (P = 0.002), evaluating IL-6 continuously and hsCRP categorically; however, the HRs for IL-6 stratified by hsCRP level were not meaningfully different (Supplemental Figure 4).
When participants were categorized by Lp(a) (MESA: ≤50 or >50 mg/dL, UKB: ≤125 or >125 nmol/L) and IL-6 (≤ or > median), the strongest association for each outcome was noted when both Lp(a) and IL-6 levels were high (CHD: HR: 1.72 [95% CI: 1.25–2.36] in MESA; HR: 1.39 [95% CI: 1.12–1.72] in UKB) in both cohorts. For participants with low Lp(a) but high IL-6, there was a significant association with each outcome (CHD: HR: 1.48 [95% CI: 1.16–1.88] in MESA; HR: 1.24 [95% CI: 1.11–1.38] in UKB). When Lp(a) was high, but IL-6 was low, there was a significant association with CHD (HR: 1.58; 95% CI: 1.13–2.21) and ASCVD (HR: 1.39; 95% CI: 1.05–1.84) in MESA, with nonsignificant association in UKB. Multiplicative interaction testing between IL-6 and Lp(a) assessed categorically was negative for all outcomes (Figure 1). When a threshold of IL-6 in the top quartile was used instead, greater risk was noted when both IL-6 and Lp(a) were elevated compared with the previous analysis for all outcomes (ie, for CHD: HR: 1.93 [95% CI: 1.36–2.76] in MESA, HR: 1.98 [95% CI: 1.58–2.49] in UKB). Again, no significant multiplicative interaction was noted, and RERI was nonsignificant. When thresholds of Lp(a) >70 mg/dL and the top quartile of IL-6 were evaluated, there was even greater risk when both were elevated (ie, for CHD: HR: 2.32 [95% CI: 1.53–3.52] in MESA and HR: 2.57 [95% CI: 1.634.07] in UKB). A significant multiplicative interaction between categoric Lp(a) and IL-6 was noted for CHD (P = 0.015) and ASCVD (P = 0.029) in MESA, but not in UKB (Supplemental Table 1). RERI was >0 for CHD (1.12; 95% CI: 0.27–2.29) and ASCVD (0.71; 95% CI: 0.09–1.51), suggesting a positive interaction in MESA, but the results were not significant in UKB (0.89; 95% CI −0.24 to 2.42 for CHD; 0.54; 95% CI −0.34 to 1.67 for ASCVD).
FIGURE 1. Combined Association of IL-6 and Lp(a) With CHD Events.
Adjusted for age, sex, race/ethnicity, total cholesterol, high-density lipoprotein cholesterol, body mass index, systolic blood pressure, diabetes mellitus, cigarette smoking, hypertension medications, statin use, and high-sensitivity C-reactive protein (log transformed). In MESA, median IL-6 was 1.21 pg/mL (Q1–Q3: 0.77–1.88 pg/mL). P values for interaction are presented for multiplicative interaction between Lp(a) >50 mg/dL or >125 nmol/L and IL-6 > median. ASCVD = atherosclerotic cardiovascular disease; CHD = coronary heart disease; IL-6 = interleukin-6; Lp(a) = lipoprotein(a); MESA = Multi-Ethnic Study of Atherosclerosis; PVD = peripheral vascular disease; UKB = UK Biobank.
When participants were categorized by hsCRP ≤2 or >2 mg/L and IL-6 ≤ or > median, increased CHD and ASCVD risk was noted when IL-6 was elevated but hsCRP was not in both cohorts. Multiplicative interaction testing between IL-6 and hsCRP assessed categorically was negative for all outcomes (Figure 2). When a threshold of IL-6 in the top quartile was used instead, similar results were noted, with negative interaction testing and nonsignificant RERI. When thresholds of hsCRP >3 mg/L and IL-6 in the top quartile were used, similar results were again noted, with negative interaction testing in MESA. In UKB, however, there was a significant interaction for CHD (P = 0.023), but not ASCVD; similarly, a significant RERI >0, suggesting a positive interaction, was noted for CHD (0.51; 95% CI: 0.17–0.83) but not ASCVD (Supplemental Table 2).
FIGURE 2. Combined Association of IL-6 and hsCRP With CHD Events.
Adjusted for age, sex, race/ethnicity, total cholesterol, high-density lipoprotein cholesterol, body mass index, systolic blood pressure, diabetes mellitus, cigarette smoking, hypertension medications, and statin use. In MESA, median IL-6 was 1.21 pg/mL (Q1–Q3: 0.77–1.88 pg/mL). P values for interaction are presented for multiplicative interaction between hsCRP >2 mg/L and IL-6 > median. hsCRP = high-sensitivity C-reactive protein; other abbreviations as in Figure 1.
Improvement in risk prediction for each outcome with the addition of Lp(a), IL-6, or hsCRP to the PCE was evaluated (Supplemental Table 3). In MESA, there was significant NRI with the addition of IL-6, but not Lp(a) or hsCRP, for CHD and ASCVD. All biomarkers resulted in significant NRI for PVD with greater improvement noted for Lp(a) + IL-6 and IL-6 alone than for hsCRP or Lp(a) alone. Borderline significant improvement was seen in the C-index for CHD and ASCVD with the addition of IL-6. Significant improvement in the C-index was noted for PVD with the addition of IL-6 alone and Lp(a) + IL-6. In UKB, there was significant NRI with the addition of all biomarkers to the PCE for CHD, and for all except Lp(a) for ASCVD. For both CHD and ASCVD, the greatest NRI was noted with the addition of both IL-6 and Lp(a). Borderline significant improvement was seen with the incorporation of IL-6 for CHD and ASCVD.
DISCUSSION
In this study of 2 independent large cohorts of individuals without prior ASCVD, Lp(a) and IL-6 were independently associated with risk for CHD, ASCVD, and PVD (Central Illustration). When evaluated together, these associations were unchanged and there was no significant interaction or effect modification between Lp(a) and IL-6 when evaluated continuously. The greatest risk for each outcome was noted when both Lp(a) and IL-6 were elevated. In addition, there was no multiplicative interaction between hsCRP and IL-6 for any outcome, but there was borderline antagonism for CHD in MESA but not in UKB. Together, these results suggest that Lp(a) and IL-6 are independent risk factors for multiple cardiovascular diseases, and risk is greater when both are elevated.
Findings that individuals with elevated Lp(a) have enhanced vascular inflammation19 and that potent and specific Lp(a) lowering attenuates the proinflammatory gene expression profile in circulating monocytes20 support the importance of inflammation in Lp(a)-mediated cardiovascular risk. Furthermore, both Lp(a) and OxPL-ApoB attributable cardiovascular risk has been conditioned to proinflammatory interleukin-1β genotypes,21 suggesting that biomarkers of inflammation may identify patients with elevated Lp(a) and elevated risk. Although interleukin-1β genotyping is not routinely available in clinical settings, previous studies have examined clinically available biomarkers of inflammation and ASCVD.
In particular, there have been several studies of hsCRP and Lp(a). In one study involving data from MESA, Lp(a) was associated with ASCVD risk when hsCRP was ≥2 mg/L, but not when hsCRP was <2 mg/L.11 In a recent study of multiple cohorts, including >300,000 primary-prevention patients and >34,000 secondary-prevention patients, Lp(a) was associated with major adverse cardiovascular events regardless of hsCRP level.12 Multiple other recent studies, including a meta-analysis, have suggested that the association between Lp(a) and ASCVD risk is independent from hsCRP.13,22 In total, these studies have demonstrated the importance of inflammatory risk in cardiovascular disease but have not clearly established the use of inflammatory biomarkers for risk-stratifying individuals with elevated Lp(a).
IL-6 is a proinflammatory cytokine produced primarily by T cells that is associated with atherosclerosis.6,8 IL-6 interacts with the LPA gene, which regulates Lp(a) levels, and IL-6 has been associated with increased Lp(a) levels.8 In addition, IL-6 inhibitors have been associated with decreases in Lp(a) levels,7 and there is an ongoing cardiovascular outcomes trial with IL-6 inhibition using ziltivekimab.23 Several studies have examined the association of IL-6 with cardiovascular risk, particularly in association with hsCRP. In an earlier analysis from MESA, IL-6 was more strongly associated with CVD than hsCRP was.24 Similarly, in an analysis of the ARIC (Atherosclerosis Risk In Communities) study, IL-6 was associated with CVD risk independently from hsCRP.25 In a long-term analysis of cohort studies with decades of follow up, IL-6 was independent from hsCRP as well as LDL-C for cardiovascular events.26
However, the interaction between Lp(a) and IL-6 for ASCVD risk and whether IL-6 levels can help to risk-stratify individuals with elevated Lp(a) in primary prevention are not well established. A recent secondary analysis of the LoDoCo2 trial evaluated the interaction between IL-6, Lp(a), and OxPL in a secondary prevention population. The original trial compared colchicine to placebo for the prevention of cardiovascular events in individuals with chronic coronary disease. In this analysis, there was a significant interaction between IL-6 levels above and below the median and Lp(a), as well as for OxPL on apolipoprotein(a) [Apo(a)] and OxPL-ApoB, but not with hsCRP. Lp(a), OxPL-Apo(a), and OxPL-ApoB were positively associated with cardiovascular events only when IL-6 was above the median level.10 These results notably contrast with our study, where there was no multiplicative interaction between Lp(a) and IL-6 levels for ASCVD risk in the primary analysis. There are several potential reasons for these differences. First, a limitation of the LoDoCo2 analysis was that biomarker measurements were conducted on end-of-study samples and thus may not reflect baseline risk. Second, the populations studied are different: Participants in LoDoCo2 were older, less often female, and with a higher prevalence of risk factors. In addition, most participants had prior acute coronary syndrome. This is reflected in the higher IL-6 levels in LoDoCo2 (median 3.2 pg/mL vs 1.2 pg/mL in MESA in our study) and may be supported by the positive interaction testing when using an Lp(a) threshold of 70 mg/dL and IL-6 threshold >1.8 pg/mL. Similarly to our study, there was a very weak and nonsignificant correlation between Lp(a) and IL-6. These results suggest that IL-6 does not modify Lp(a)-induced risk in relatively healthy and stable populations but may be more relevant in acute disease states or in chronic inflammatory conditions with higher IL-6 levels. Importantly, our results were adjusted for hsCRP which, as discussed above, has been inconsistently shown to modify Lp(a)-mediated risk. Indeed, IL-6 was independent from Lp(a), with the greatest risk noted when both biomarkers were at higher levels.
In a long-term analysis of the Women’s Health Study, the association of LDL-C, Lp(a), and hsCRP with cardiovascular events was evaluated in a primary prevention population. Notable findings were that each biomarker was independently associated with cardiovascular risk and that there was increasing risk with combined elevations of these biomarkers.27 IL-6, however, was not evaluated. Although IL-6 and Lp(a) were the focus of our study, the results demonstrate the association of Lp(a) with cardiovascular risk independently from IL-6. We also observed that hsCRP was independently associated with CHD and ASCVD risk in UKB but not in MESA, which may be due to the higher median hsCRP levels in UKB compared with MESA. The addition of IL-6 to a traditional risk factor model was associated with improvement in reclassification for CHD, ASCVD, and PVD, whereas hsCRP was associated with improvement to a lesser degree than IL-6, suggesting that IL-6 is associated with greater predictive value than hsCRP for cardiovascular events. However, further study is needed to compare these biomarkers and to evaluate the combined association of multiple biomarkers including IL-6.
We recently evaluated Lp(a) in the context of the novel AHA PREVENT equations in a study of MESA and UKB. In that study, we observed that elevated Lp(a) was associated with increased cardiovascular risk across PREVENT risk categories and that the addition of Lp(a) modestly improved risk prediction.28 The present study evaluated the PCE rather than the PREVENT equations, but similarly demonstrated modest improvement in risk prediction with the incorporation of Lp(a). In addition, although estimation of risk differs between the PCE and PREVENT equations, the independent associations of elevated Lp(a) with risk are similar for both.28,29 A novel aspect of the present study is evaluation of both Lp(a) and measures of inflammatory risk. In general, the incorporation of IL-6 was associated with greater improvement in risk prediction than the incorporation of Lp(a) or hsCRP. In addition, greater improvement was generally observed with the incorporation of both Lp(a) and IL-6, suggesting that assessment of inflammatory risk can further enhance risk prediction in addition to assessment of Lp(a).
The present study has clinical implications for risk assessment in primary prevention. Guidelines are in evolution: The 2018 multisociety cholesterol guidelines noted Lp(a) as a risk enhancer but did not endorse universal testing,30 but several more recent international guidelines have recommended universal Lp(a) testing at least once in adults.31–33 In the 2018 guidelines, hsCRP was considered as a risk enhancer and there was no specific recommendation regarding IL-6.30 The findings of the present study support the potential incorporation of IL-6 in a primary prevention setting in addition to Lp(a); at the minimum, IL-6 may be considered as a risk enhancer because it identifies individuals with increased cardiovascular risk independently from traditional risk factors hsCRP and Lp(a). Moreover, individuals with both elevated Lp(a) and IL-6 were at the highest risk, IL-6 was associated with improved predictive value for all outcomes when added to traditional risk factors, and the greatest improvement in risk prediction overall was noted with incorporation of IL-6. Our findings also suggest that IL-6 is a better marker of cardiovascular risk than hsCRP is, but further study is needed. We also observed borderline significant antagonism between hsCRP and IL-6 in MESA for CHD, which may be due to both capturing overlapping components of inflammatory risk, or it may be artifactual as it was not replicated in UKB and was not observed for other outcomes or when using higher thresholds of IL-6 and hsCRP. There was also evidence for a positive interaction between IL-6 and Lp(a) when using higher thresholds. Although these findings should be interpreted with caution because the interaction testing was negative when evaluating both continuously or using lower thresholds and was not present in UKB, these findings support a need for further study in individuals with high levels of both biomarkers. Another important finding is that the greatest risk among outcomes evaluated was for PVD. Given that PVD is often underrecognized, this should prompt clinicians to have heightened awareness of PVD in individuals with elevated Lp(a) or IL-6.
Our findings also have implications for trials of future IL-6- and Lp(a)-lowering therapies. The greatest risk for all outcomes was noted when both Lp(a) and IL-6 were elevated. This suggests that each may be used as enrichment criteria for the other, particularly for novel therapies in a primary-prevention setting where there is a desire to identify the highest risk individuals to have sufficient power to demonstrate efficacy. In the ACCLAIM-Lp(a) trial, for example, hsCRP ≥2 mg/L is a criterion for potential inclusion in the primary prevention arm in addition to elevated Lp(a) (NCT06292013). A potential area for future study in these trials would also be prespecified subgroups by Lp(a) or IL-6 levels to determine if individuals with elevation of both derive greater benefit from targeted therapies, particularly given the positive interaction between Lp(a) and IL-6 when using higher thresholds for each. Another notable finding was that the greatest risk associated with Lp(a) and IL-6 was for PVD. This suggests that trials of targeted therapy for Lp(a) and IL-6 should consider PVD as an endpoint. This is not the case for the current phase 3 trials of Lp(a)-lowering therapy (NCT04023552, NCT05581303, NCT06292013) and IL-6–lowering therapy (NCT05021835, NCT0611828).
STUDY LIMITATIONS.
Our study has notable limitations. Although Lp(a) levels are considered to be relatively stable throughout life, IL-6 levels, as markers of inflammation, may vary over time, and our results use a single measure of IL-6. In addition, measured IL-6 levels were not available in UKB; they were determined by means of proteomics analysis and cannot be presented in standard units; however, results were consistent between MESA and UKB. In addition, although IL-6 levels were available for the majority of MESA participants, proteomics data were available on only a subset of the UKB population, which may lead to selection bias. However, UKB proteomics measurements were performed on a primarily random sample that had similar characteristics to the overall UKB population and were not done specifically for IL-6 measurements. Moreover, the results between MESA and UKB were generally consistent. Lead time bias may also be present because individuals with better access to care or more comorbidities may seek medical care sooner, leading to earlier recorded outcomes; we attempted to address this by adjusting for insurance status and education level in MESA; these adjustments did not change the primary results. IL-6 is known to regulate Lp(a) levels, but there was a very weak correlation between IL-6 and Lp(a) levels in our study. This may be due to having a relatively healthy population, and different results may be obtained in other populations such as in acute coronary syndrome or other secondary prevention settings. We did observe a similar correlation between IL-6 and Lp(a) in those with and without subsequent CVD events, but we were limited to evaluating baseline levels of the biomarkers. Therefore, further study is needed. Finally, as an observational study, our results are subject to residual confounding, particularly regarding IL-6 levels.
CONCLUSIONS
In 2 independent primary prevention cohorts, Lp(a) and IL-6 were independent predictors of ASCVD risk, and their combination identified individuals at highest risk.
Supplementary Material
ACKNOWLEDGMENTS
The authors thank the other investigators, the staff, and the participants of MESA for their valuable contributions. A full list of participating MESA investigators and institutions is available at www.mesa-nhlbi.org. This research was conducted using the UKB Resource under application no. 97439.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
Dr Bhatia is supported by National Institutes of Health grant 1K08HL166962. Dr Yeang has received research support from the National Institutes of Health1K08HL150271-01 and Kaneka Corp. Dr Tsimikas is supported by National Heart, Lung, and Blood Institute grants R01 HL159156 and HL170224. MESA was supported by contracts 75N92020D00001, HHSN268201500003I, N01-HC-95159, 75N92020D00005, N01-HC-95160, 75N92020D00002, N01-HC-95161, 75N92020D00003, N01-HC-95162, 75N92020D00006, N01-HC-95163, 75N92020D00004, N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, and N01-HC-95169 from the National Heart, Lung, and Blood Institute and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from the National Center for Advancing Translational Sciences. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Dr Bhatia is a consultant or advisor for Kaneka, Novartis, Arrowhead, Abbott, and New-Amsterdam. Dr Tsai is a co-inventor on and receives royalties for patents owned by the University of California San Diego (UCSD), is a co-founder and has equity interest in Oxitope and Kleanthi Diagnostics, is a consultant for Novartis, and has a dual appointment at UCSD and Ionis Pharmaceuticals as reviewed and approved by UCSD in accordance with its conflict of interest policies. Dr Shapiro is supported by institutional grants from Amgen, Arrowhead, Boehringer Ingelheim, 89Bio, Esperion, Novartis, Ionis, Merck, New Amsterdam, and Cleerly, has participated in scientific advisory boards for Amgen, Arrowhead Ionis, Novartis, New Amsterdam, Tourmaline, and Merck, and has served as a consultant for Ionis, Novartis, Regeneron, Aidoc, Novo Nordisk, Arrowhead, and Tourmaline. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
ABBREVIATIONS AND ACRONYMS
- ASCVD
atherosclerotic cardiovascular disease
- CHD
coronary heart disease
- hsCRP
high-sensitivity C-reactive protein
- IL
interleukin
- LDL-C
low-density lipoprotein-cholesterol
- Lp(a)
lipoprotein(a)
- NRI
net reclassification improvement
- OxPL-ApoB
oxidized phospholipids on apolipoprotein B-100
- PCE
pooled cohort equation
- PVD
peripheral vascular disease
- RERI
relative excess risk due to interaction
APPENDIX
For supplemental figures and tables, please see the online version of this paper.
Footnotes
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
REFERENCES
- 1.Bhatia HS, Wilkinson MJ. Lipoprotein(a): evidence for role as a causal risk factor in cardiovascular disease and emerging therapies. J Clin Med. 2022;11:6040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bhatia HS, Wandel S, Willeit P, et al. Independence of lipoprotein(a) and low-density lipoprotein cholesterol–mediated cardiovascular risk: a participant-level meta-analysis. Circulation. 2025;151:312–321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Taleb A, Witztum JL, Tsimikas S. Oxidized phospholipids on ApoB-100–containing lipoproteins: a biomarker predicting cardiovascular disease and cardiovascular events. Biomark Med. 2011;5:673–694. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tsimikas S, Brilakis ES, Miller ER, et al. Oxidized phospholipids, Lp(a) lipoprotein, and coronary artery disease. N Engl J Med. 2005;353:46–57. [DOI] [PubMed] [Google Scholar]
- 5.Mohammadnia N, van Broekhoven A, Bax WA, et al. The effects of colchicine on lipoprotein(a) and oxidized phospholipid associated cardiovascular disease risk. Eur J Prev Cardiol. 2025;32: 758–765. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Swerdlow DI, Holmes MV, Kuchenbaecker KB, et al. The interleukin-6 receptor as a target for prevention of coronary heart disease: a mendelian randomisation analysis. Lancet. 2012;379:1214–1224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Müller N, Schulte DM, Türk K, et al. IL-6 blockade by monoclonal antibodies inhibits apolipoprotein(a) expression and lipoprotein(a) synthesis in humans. J Lipid Res. 2015;56:1034–1042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Makris A, Barkas F, Sfikakis PP, et al. Lipoprotein(a), interleukin-6 inhibitors, and atherosclerotic cardiovascular disease: is there an association? Atheroscler Plus. 2023;54:1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ridker PM, Devalaraja M, Baeres FMM, et al. IL-6 inhibition with ziltivekimab in patients at high atherosclerotic risk (RESCUE): a double-blind, randomised, placebo-controlled, phase 2 trial. Lancet. 2021;397:2060–2069. [DOI] [PubMed] [Google Scholar]
- 10.Mohammadnia N, van Broekhoven A, Bax WA, et al. Interleukin-6 modifies lipoprotein(a) and oxidized phospholipids associated cardiovascular disease risk in a secondary prevention cohort. Atherosclerosis. 2025;405:119211. [DOI] [PubMed] [Google Scholar]
- 11.Zhang W, Speiser JL, Ye F, et al. High-sensitivity C-reactive protein modifies the cardiovascular risk of lipoprotein(a). J Am Coll Cardiol. 2021;78:1083–1094. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Small AM, Pournamdari A, Melloni GEM, et al. Lipoprotein(a), C-reactive protein, and cardiovascular risk in primary and secondary prevention populations. JAMA Cardiol. 2024;9:385–391. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Alebna PL, Han CY, Ambrosio M, et al. Association of lipoprotein(a) with major adverse cardiovascular events across hs-CRP: a systematic review and meta-analysis. JACC Adv. 2024;3: 101409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-Reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA. 2001;286:327–334. [DOI] [PubMed] [Google Scholar]
- 15.Bild DE, Bluemke DA, Burke GL, et al. Multi-Ethnic Study of Atherosclerosis: objectives and design. Am J Epidemiol. 2002;156:871–881. [DOI] [PubMed] [Google Scholar]
- 16.Cao J, Steffen BT, Guan W, et al. Evaluation of lipoprotein(a) electrophoretic and immunoassay methods in discriminating risk of calcific aortic valve disease and incident coronary heart disease: the Multi-Ethnic Study of Atherosclerosis. Clin Chem. 2017;63:1705–1713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Steffen BT, Steffen LM, Tracy R, et al. Obesity modifies the association between plasma phospholipid polyunsaturated fatty acids and markers of inflammation: the Multi-Ethnic Study of Atherosclerosis. Int J Obes (Lond). 2012;36:797–804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sun BB, Chiou J, Traylor M, et al. Plasma proteomic associations with genetics and health in the UK Biobank. Nature. 2023;622:329–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.van der Valk FM, Bekkering S, Kroon J, et al. Oxidized phospholipids on lipoprotein(a) elicit arterial wall inflammation and an inflammatory monocyte response in humans. Circulation. 2016;134:611–624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Stiekema LCA, Prange KHM, Hoogeveen RM, et al. Potent lipoprotein(a) lowering following apolipoprotein(a) antisense treatment reduces the pro-inflammatory activation of circulating monocytes in patients with elevated lipoprotein (a). Eur Heart J. 2020;41:2262–2271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tsimikas S, Duff GW, Berger PB, et al. Proinflammatory interleukin-1 genotypes potentiate the risk of coronary artery disease and cardiovascular events mediated by oxidized phospholipids and lipoprotein(a). J Am Coll Cardiol. 2014;63:1724–1734. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Thomas PE, Vedel-Krogh S, Afzal S, Nordestgaard BG, Kamstrup PR. Lipoprotein(a)-associated risk of atherosclerotic cardiovascular disease is independent of C-reactive protein. Eur J Prev Cardiol. 2025;32:783–785. [DOI] [PubMed] [Google Scholar]
- 23.Perkovic V, Tuttle K, Sattar N, et al. WCN25888 Design of the ZEUS trial: interleukin-6 inhibition with ziltivekimab for cardiovascular protection in chronic kidney disease. Kidney Int Rep. 2025;10:S767–S768. [Google Scholar]
- 24.Ferreira JP, Vasques-Nóvoa F, Neves JS, Zannad F, Leite-Moreira A. Comparison of interleukin-6 and high-sensitivity C-reactive protein for cardiovascular risk assessment: findings from the MESA study. Atherosclerosis. 2024;390: 117461. [DOI] [PubMed] [Google Scholar]
- 25.Jia X, Buckley L, Sun C, et al. Association of interleukin-6 and interleukin-18 with cardiovascular disease in older adults: Atherosclerosis Risk in Communities study. Eur J Prev Cardiol. 2023;30:1731–1740. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ridker PM, MacFadyen JG, Glynn RJ, Bradwin G, Hasan AA, Rifai N. Comparison of interleukin-6, C-reactive protein, and low-density lipoprotein cholesterol as biomarkers of residual risk in contemporary practice: secondary analyses from the Cardiovascular Inflammation Reduction Trial. Eur Heart J. 2020;41:2952–2961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ridker PM, Moorthy MV, Cook NR, Rifai N, Lee I-M, Buring JE. Inflammation, cholesterol, lipoprotein(a), and 30-year cardiovascular outcomes in women. N Engl J Med. 2024;391:2087–2097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bhatia HS, Ambrosio M, Razavi AC, et al. AHA PREVENT equations and lipoprotein(a) for cardiovascular disease risk: insights from MESA and the UK Biobank. JAMA Cardiol. 2025;10:810–818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Bhatia HS, Rikhi R, Allen TS, et al. Lipoprotein(a) and the pooled cohort equations for ASCVD risk prediction: the Multi-Ethnic Study of Atherosclerosis. Atherosclerosis. 2023;381: 117217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Grundy SM, Stone NJ, Bailey AL, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2019;73:e285–e350. [DOI] [PubMed] [Google Scholar]
- 31.Mach F, Baigent C, Catapano AL, et al. 2019 ESC/EAS guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk. Eur Heart J. 2020;41:111–188. [DOI] [PubMed] [Google Scholar]
- 32.Kronenberg F, Mora S, Stroes ESG, et al. Lipoprotein(a) in atherosclerotic cardiovascular disease and aortic stenosis: a European Atherosclerosis Society consensus statement. Eur Heart J. 2022;43:3925–3946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Koschinsky ML, Bajaj A, Boffa MB, et al. A focused update to the 2019 NLA scientific statement on use of lipoprotein(a) in clinical practice. J Clin Lipidol. 2024;18:e308–e319. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.