Increased Remnant Cholesterol Explains Part of Residual Risk of All-Cause Mortality in 5414 Patients with Ischemic Heart Disease

Abstract

BACKGROUND

Increased concentrations of remnant cholesterol are causally associated with increased risk of ischemic heart disease. We tested the hypothesis that increased remnant cholesterol is a risk factor for all-cause mortality in patients with ischemic heart disease.

METHODS

We included 5414 Danish patients diagnosed with ischemic heart disease. Patients on statins were not excluded. Calculated remnant cholesterol was nonfasting total cholesterol minus LDL and HDL cholesterol. During 35836 person-years of follow-up, 1319 patients died.

RESULTS

We examined both calculated and directly measured remnant cholesterol; importantly, however, measured remnant cholesterol made up only 9% of calculated remnant cholesterol at nonfasting triglyceride concentrations <1 mmol/L (89 mg/dL) and only 43% at triglycerides >5 mmol/L (443 mg/dL). Multivariable-adjusted hazard ratios for all-cause mortality compared with patients with calculated remnant cholesterol concentrations in the 0 to 60th percentiles were 1.2 (95% CI, 1.1–1.4) for patients in the 61st to 80th percentiles, 1.3 (1.1–1.5) for the 81st to 90th percentiles, 1.5 (1.1–1.8) for the 91st to 95th percentiles, and 1.6 (1.2–2.0) for patients in the 96th to 100th percentiles (trend, P < 0.001). Corresponding values for measured remnant cholesterol were 1.0 (0.8–1.1), 1.2 (1.0–1.4), 1.1 (0.9–1.5), and 1.3 (1.1–1.7) (trend, P = 0.006), and for measured LDL cholesterol 1.0 (0.9–1.1), 1.0 (0.8–1.2), 1.0 (0.8–1.3), and 1.1 (0.8–1.4) (trend, P = 0.88). Cumulative survival was reduced in patients with calculated remnant cholesterol ≥1 mmol/L (39 mg/dL) vs <1 mmol/L [log-rank, P = 9 × 10⁻⁶; hazard ratio 1.3 (1.2–1.5)], but not in patients with measured LDL cholesterol ≥3 mmol/L (116 mg/dL) vs <3 mmol/L [P = 0.76; hazard ratio 1.0 (0.9–1.1)].

CONCLUSIONS

Increased concentrations of both calculated and measured remnant cholesterol were associated with increased all-cause mortality in patients with ischemic heart disease, which was not the case for increased concentrations of measured LDL cholesterol. This suggests that increased concentrations of remnant cholesterol explain part of the residual risk of all-cause mortality in patients with ischemic heart disease.

Mean triglyceride concentrations have risen with the growing epidemic of obesity and type 2 diabetes, and roughly 30% of adults in Europe and the US have increased concentrations of triglycerides (1, 2). Importantly, increased triglycerides are associated with increased risk of cardiovascular disease (3–6) and total mortality (4, 6, 7). Also, in patients with acute coronary syndrome treated with statins, increased triglycerides predict risk of recurring cardiovascular events (8, 9). Because triglycerides can be readily metabolized by most cells of the body, increased triglycerides per se are unlikely to directly cause cardiovascular disease; however, increased concentrations of triglycerides mark increased concentrations of potentially atherogenic remnant cholesterol (4, 6, 7, 10).

Remnant cholesterol is the cholesterol content of triglyceride-rich lipoproteins and is composed of very-low-density lipoproteins (VLDL)⁴ and intermediate-density lipoproteins (IDL) in the fasting state, and of VLDL, IDL, and chylomicron remnants in the nonfasting state. Increased remnant cholesterol, like increased LDL cholesterol, is causally associated with increased risk of ischemic heart disease (10). Furthermore, increased remnant cholesterol is also causally associated with low-grade inflammation, whereas increased LDL cholesterol is not (11). Finally, extreme high concentrations of remnant cholesterol, but not extreme high concentrations of LDL cholesterol, have been associated with increased risk of all-cause mortality in the general population (12).

After reduction of LDL cholesterol to recommended concentrations, there is still considerable residual risk of recurring cardiovascular events (13–15). Although statins lower LDL cholesterol and cardiovascular risk in survivors of acute coronary syndrome, statins only have a small effect in lowering triglycerides (9). Therefore, it is possible that increased triglycerides and thus increased remnant cholesterol may explain part of the residual risk in patients with ischemic heart disease.

In this study including 5414 patients with ischemic heart disease from the Copenhagen Ischemic Heart Disease Study (CIHDS), we tested the hypothesis that increased remnant cholesterol is a risk factor for all-cause mortality in patients with ischemic heart disease. Directly measured remnant cholesterol captures only about 13% of the calculated value across all individuals in the general population (12), and therefore we included both calculated and measured remnant cholesterol in our analyses.

Methods

THE COPENHAGEN ISCHEMIC HEART DISEASE STUDY

The CIHDS includes patients from the greater Copenhagen area referred for coronary angiography to Rigshospitalet, Copenhagen University Hospital, from 2001 to 2012. All patients had been examined by a physician before they were referred to the cardiology department. Patients were referred from another hospital, from a primary care physician, or from an emergency doctor. Blood samples were drawn at admission. In addition to a diagnosis of ischemic heart disease, these patients had stenosis/atherosclerosis on coronary angiography and/or a positive exercise electrocardiography test. In this study, we included 5414 patients with data on measured and calculated remnant cholesterol as well as on other lipids, lipoproteins, and apolipoproteins. The study was approved by Herlev and Gentofte Hospital and by a Danish ethics committee and was conducted according to the Declaration of Helsinki with written informed consent from all participants. All patients were whites of Danish descent.

ISCHEMIC HEART DISEASE AND ALL-CAUSE MORTALITY

Information on the diagnosis of ischemic heart disease (International Classification of Diseases, Eighth Revision codes 410–414 and Tenth Revision codes I20–I25) was collected from January 1977 to April 2013 by reviewing all hospital admissions entered in the national Danish Patient Registry and all causes of death entered in the national Danish Causes of Death Registry. Information on death was obtained from the national Danish Civil Registration System. Every individual in Denmark is assigned a personal identification number at birth or immigration and can thereby be traced in the national registers without any loss to follow-up.

LABORATORY ANALYSES

All lipid, lipoprotein, and apolipoprotein measurements were done on thawed blood samples, frozen at −80 °C from the day of examination until measurement in 2014. Standard colorimetric and turbidimetric assays were used to measure nonfasting plasma concentrations of triglycerides, total cholesterol, LDL cholesterol (direct), HDL cholesterol (direct), apolipoprotein A1, and apolipoprotein B (Konelab, Helsinki, Finland). Remnant cholesterol was calculated as nonfasting total cholesterol minus measured HDL cholesterol minus measured LDL cholesterol. In addition, remnant cholesterol was measured directly with an automated assay by Denka Seiken, measuring the cholesterol content in chylomicron and VLDL remnants specifically, with the aid of enzymes and surfactants (12) in 2 steps. The first step removes the cholesterol in other lipoproteins, and the second step determines the cholesterol in the remaining remnant lipoprotein particles (see Supplemental Fig. 1, which accompanies the online version of this article at http://www.clinchem.org/content/vol62/issue4). Small dense LDL and HDL3 were also measured with assays from Denka Seiken. HDL2 cholesterol was calculated as total HDL cholesterol minus HDL3 cholesterol.

OTHER COVARIATES

Smokers were current smokers. Hypertension was systolic blood pressure >160 mmHg, diastolic blood pressure >90 mmHg, and/or use of antihypertensive medication. Statin use was self-reported at study entry. Diabetes mellitus was self-reported, blood glucose >11 mmol/L (198 mg/dL), use of insulin or other antidiabetic drugs, and/or a diagnosis of diabetes in national registers (International Classification of Diseases, Tenth Revision codes E11, E13–E14). Body mass index was divided into 5 categories: <25, 25–29, 30–34, 35–40, and ≥40 kg/m². High alcohol consumption was >7 U per week for women and >14 U per week for men in accordance with the current recommendations from the Danish Health Authority (1 unit, approximately 12 g).

STATISTICAL ANALYSIS

We used Stata/SE 13.1. Data on hypertension were available for 3112 patients, on diabetes for 3196 patients, on alcohol consumption for 1561 patients, on body mass index for 1768 patients, and on smoking and statin use for 2744 patients. For multivariable-adjusted analyses, missing values for these variables were imputed on the basis of age, sex, and date of birth. Analyses were purposely not adjusted for body mass index, alcohol consumption, and diabetes, because these variables in themselves may affect lipid, lipoprotein, and apolipoprotein concentrations through biological pathways. Patients were not excluded if on statin treatment.

Patients were divided into tertiles on the basis of calculated remnant cholesterol, measured remnant cholesterol, measured LDL cholesterol, measured small dense LDL, measured HDL cholesterol, measured HDL3 cholesterol, calculated HDL2 cholesterol, measured apolipoprotein A1, and measured apolipoprotein B. In different analyses, patients were stratified into 4 groups with focus on clinically relevant extreme high concentrations according to calculated remnant cholesterol (<0.5, 0.5–0.99, 1–1.49, and ≥1.5 mmol/L; <19, 19–38, 39–58, and ≥59 mg/dL), measured remnant cholesterol (<0.05, 0.05–0.19, 0.2–0.29, and ≥0.3 mmol/L; <2, 2–7, 8–11, ≥12 mg/dL), and measured LDL cholesterol (<2, 2–2.99, 3–3.99, ≥4 mmol/L; <77, 77–116, 116–154, ≥155 mg/dL). We aimed at having similar fractions of patients in the top groups of calculated remnant cholesterol, measured remnant cholesterol, and measured LDL cholesterol to allow direct comparison of extreme concentrations of these lipoproteins, as done previously (12). For comparison, patients were furthermore divided into extreme percentiles according to concentrations of calculated remnant cholesterol, measured remnant cholesterol, and measured LDL cholesterol.

We used Cox proportional hazards models with age as time scale and examination date as date of entry to estimate hazard ratios (HRs). Multivariable adjustment was for age, sex, smoking, hypertension, and statin use, or additionally for body mass index and diabetes. We tested for interaction on all-cause mortality between calculated remnant cholesterol, measured remnant cholesterol, and measured LDL cholesterol and sex, hypertension, statin use, alcohol consumption, diabetes, body mass index, and smoking status by introducing 2-factor interaction terms in the multivariable-adjusted Cox regression models; no significant interactions were detected (see online Supplemental Table 1). Cumulative survival was plotted with Kaplan–Meier curves, and log-rank tests were used for comparison of cumulative survival between groups.

The association between lipoprotein cholesterol concentrations and all-cause mortality was also examined with a restricted cubic spline model having 3 knots at default values, by use of sex- and age-adjusted HRs from Cox regression on rounded values of calculated remnant cholesterol to 1 decimal, measured remnant cholesterol to 2 decimals, and measured LDL cholesterol to 1 decimal.

Population-attributable risk was calculated as [f (HR − 1)]/[1 + f (HR − 1)], where f is the frequency of the risk factor in the population at baseline and HR is the hazard ratio for all-cause mortality. These calculations were also conducted by use of HRs corrected for regression dilution bias: HRs and CIs were corrected for regression dilution bias with a nonparametric method (16), by use of lipoprotein values from 4253 individuals without lipid-lowering therapy participating in both the 1991–1994 and 2011–2003 examinations of the Copenhagen City Heart Study (10); this correction helps avoid underestimation of risk estimates, but it does not affect whether results are statistically significant. Regression dilution ratios of 2.2 and 1.6 were computed for calculated remnant cholesterol and LDL cholesterol.

Results

Table 1 shows baseline characteristics of patients as a function of tertiles of calculated remnant cholesterol, measured remnant cholesterol, and measured LDL cholesterol. Although not all patients were on statins at baseline, we suspect that the majority were prescribed a statin during hospital visits, as this has been the standard practice at the cardiology department recruiting these patients during the entire study period (per authors A. Langsted, L.E. Bang, and B.G. Nordestgaard). During 35836 person-years of follow-up, 1319 individuals died, of whom 555 (42%) died of cardiovascular causes, 324 (25%) died of cancer, 386 (29%) died of other causes, and 54 (4%) had unknown cause of death. Median follow-up was 7 years (range 0–12 years).

LIPIDS, LIPOPROTEINS, AND APOLIPOPROTEINS AND RISK OF ALL-CAUSE MORTALITY

Associations of nonfasting lipids, lipoproteins, and apolipoproteins in tertiles with risk of all-cause mortality are shown in Fig. 1. Multivariable-adjusted HRs for the highest vs lowest tertile were 1.3 (95% CI, 1.2–1.5) for calculated remnant cholesterol, 1.1 (1.0–1.3) for measured remnant cholesterol, 0.9 (0.8–1.0) for measured LDL cholesterol, 1.0 (0.8–1.1) for measured small dense LDL, 0.8 (0.7–0.9) for measured HDL cholesterol, 0.7 (0.6–0.8) for measured HDL3 cholesterol, 0.8 (0.7–1.0) for calculated HDL2 cholesterol, 0.8 (0.7–0.9) for measured apolipoprotein A1, 1.1 (0.9–1.2) for measured apolipoprotein B, 1.0 (0.9–1.1) for measured total cholesterol, and 1.2 (1.0–1.3) for measured triglycerides.

Risk of all-cause mortality as a function of nonfasting lipids, lipoproteins, and apolipoproteins in 5414 patients with ischemic heart disease.

ASSOCIATIONS AND CORRELATIONS OF LIPOPROTEIN CHOLESTEROL AND TRIGLYCERIDES

Increased concentrations of plasma triglycerides were associated with increased concentrations of both calculated and measured remnant cholesterol and with reduced concentrations of HDL cholesterol (Fig. 2); intercorrelation of all lipids, lipoproteins, and apolipoproteins are given in Fig. 2 and online Supplemental Table 2. With increasing plasma concentrations of triglycerides, measured remnant cholesterol captured an increasingly larger portion of calculated remnant cholesterol. At nonfasting triglyceride concentrations <1 mmol/L (89 mg/dL), measured remnant cholesterol made up 9% of calculated remnant cholesterol. This was 12% for triglycerides of 1–1.99 mmol/L (89–176 mg/dL), 18% for 2–2.99 mmol/L (177–265 mg/dL), 25% for 3–3.99 mmol/L (266–353 mg/dL), 32% for 4–4.99 mmol/L (354–442 mg/dL), and 43% for ≥5 mmol/L (≥443 mg/dL). As remnant cholesterol is causally related with ischemic heart disease (6, 10, 11), like LDL cholesterol and unlike the other lipids, lipoproteins, and apolipoproteins, the remainder of the data analyses focused on remnant and LDL cholesterol only.

Associations and correlations of lipoprotein cholesterol and triglycerides in 5414 patients with ischemic heart disease.

REMNANT AND LDL CHOLESTEROL AND RISK OF ALL-CAUSE MORTALITY

To study more clinically relevant cutpoints as compared to tertiles, patients were divided into 4 groups according to extreme concentrations of calculated remnant cholesterol, measured remnant cholesterol, and measured LDL cholesterol (Fig. 3, top). For calculated remnant cholesterol, multivariable-adjusted HRs for all-cause mortality were 1.0 (95% CI, 0.9–1.2) for patients with 0.5–0.99 mmol/L (19–38 mg/dL), 1.3 (1.1–1.6) for 1.0–1.49 mmol/L (39–58 mg/dL), and 1.5 (1.2–2.0) for ≥1.5 mmol/L (≥59 mg/dL), compared with patients with remnant cholesterol <0.5 mmol/L (<19 mg/dL) (trend, P < 0.001). For measured remnant cholesterol, multivariable-adjusted HRs for all-cause mortality were 1.0 (0.8–1.1) for patients with 0.05–0.19 mmol/L (2–7 mg/dL), 1.1 (0.9–1.4) for 0.2–0.29 mmol/L (8–11 mg/dL), and 1.2 (1.0–1.5) for ≥0.3 mmol/L (≥12 mg/dL), compared with patients with measured remnant cholesterol <0.05 mmol/L (<2 mg/dL) (trend, P = 0.04). For measured LDL cholesterol, multivariable-adjusted HRs for all-cause mortality were 0.9 (0.8–1.0) for patients with 2–2.99 mmol/L (77–116 mg/dL), 0.9 (0.8–1.1) for 3–3.99 mmol/L (116–154 mg/dL), and 0.9 (0.8–1.2) for ≥4 mmol/L (>155 mg/dL), compared with patients with measured LDL cholesterol <2 mmol/L (<77 mg/dL) (trend, P = 0.35).

Risk of all-cause mortality as a function of remnant and LDL cholesterol concentrations in 5414 patients with ischemic heart disease.

When patients were divided into percentiles also with focus on extreme concentrations, patients with increased concentrations of calculated remnant cholesterol vs patients in the 0 to 60th percentile had multivariable-adjusted HRs for all-cause mortality of 1.2 (95% CI, 1.1–1.4) for patients in the 61st to 80th percentiles, 1.3 (1.1–1.5) for the 81st to 90th percentiles, 1.5 (1.1–1.8) for the 91st to 95th percentiles, and 1.6 (1.2–2.0) for the 96th to 100th percentiles (trend, P < 0.001) (Fig. 3, bottom). Corresponding values for measured remnant cholesterol were 1.0 (0.8–1.1), 1.2 (1.0–1.4), 1.1 (0.9–1.5), and 1.3 (1.1–1.7) (trend, P = 0.006), and for measured LDL cholesterol, 1.0 (0.9–1.1), 1.0 (0.8–1.2), 1.0 (0.8–1.3), and 1.1 (0.8–1.4) (trend, P = 0.88). In analyses including only individuals with complete data, and thus reduced statistical power, results were similar (see online Supplemental Fig. 2).

Fig. 4 (left) shows the sex- and age-adjusted HRs for all-cause mortality comparing patients with cholesterol concentrations greater than vs equal to or less than cutpoints of calculated remnant cholesterol, measured remnant cholesterol, and measured LDL cholesterol. With higher cutpoints for calculated and measured remnant cholesterol, the HR for all-cause mortality increased, although this was not statistically significant for measured remnant cholesterol or calculated remnant cholesterol >2.1 mmol/L (81 mg/dL). Associations between cholesterol concentrations of these 3 lipoprotein fractions and risk of all-cause mortality, when expressed on a continuous scale and analyzed by use of restricted cubic splines, showed patterns of association similar to those observed by use of different cutpoints of the lipoproteins (Fig. 4, right).

Risk of all-cause mortality as a function of different cutpoint values, comparing patients with cholesterol concentrations above a cutpoint to patients with concentrations below or equal to the cutpoint (left panels) and restricted cubic spline plot of the association between cholesterol concentrations and risk of all-cause mortality (right panels).

CUMULATIVE SURVIVAL BY REMNANT AND LDL CHOLESTEROL

Cumulative survival was reduced in patients with calculated remnant cholesterol ≥1 mmol/L (39 mg/dL) vs <1 mmol/L (log-rank, P = 9 × 10⁻⁶), but not in patients with measured LDL cholesterol ≥3 mmol/L (116 mg/dL) vs <3 mmol/L (P = 0.76) (Fig. 5). Sex- and age-adjusted HR for all-cause mortality in patients with calculated remnant cholesterol ≥1 mmol/L (39 mg/dL) vs <1 mmol/L was 1.3 (95% CI, 1.2–1.5) and in patients with measured LDL cholesterol ≥3 mmol/L (116 mg/dL) vs <3 mmol/L, 1.0 (0.9–1.1); corresponding values after correction for regression dilution bias were 1.8 (1.4–2.4) and 1.0 (0.9–1.2). Because 27% of patients had calculated remnant cholesterol ≥1 mmol/L (39 mg/dL), the population-attributable risk of all-cause mortality for calculated remnant cholesterol was 8% and 18% for uncorrected and for regression dilution bias–corrected HRs.

Cumulative survival in patients with ischemic heart disease according to high and low concentrations of calculated remnant cholesterol and measured LDL cholesterol.

REMNANT AND LDL CHOLESTEROL AND RISK OF RECURRENT CARDIOVASCULAR EVENTS

Patients with calculated remnant cholesterol in the 96th to 100th percentile vs the 0 to 60th percentile had a multivariable-adjusted HR for recurrent fatal or nonfatal myocardial infarction or stroke of 1.4 (95% CI, 1.1–1.8) (see online Supplemental Fig. 3). Corresponding values were 1.2 (0.9–1.5) for measured remnant cholesterol and 1.4 (1.1–1.9) for measured LDL cholesterol.

Discussion

By studying 5414 patients with ischemic heart disease, we observed that the risk of all-cause mortality increased with higher remnant cholesterol concentrations, but not with higher concentrations of LDL cholesterol. To our knowledge, this is the first study showing such an association in patients with ischemic heart disease.

Triglycerides and remnant cholesterol are highly correlated, and high triglycerides are a marker of high remnant concentrations. Therefore, associations found for remnant cholesterol will be similar for triglycerides, and high triglycerides have previously been shown to be associated strongly with ischemic heart disease, myocardial infarction, and all-cause mortality in the Danish general population (4, 7). However, because triglycerides, but not cholesterol, can be readily metabolized by most cells of the body, triglycerides per se are unlikely to be causal. Importantly, no previous studies have examined remnant cholesterol and only a single study has examined triglycerides (9) for prediction of all-cause mortality in patients who have already been diagnosed with ischemic heart disease, making the present findings novel.

Increased concentrations of remnant cholesterol are thought to cause atherosclerosis just as increased LDL cholesterol does (6, 17). Mechanistically, remnant particles including chylomicron remnants, VLDL, and IDL penetrate the atrial wall (18–20) and appear to be selectively retained within the arterial intima (21, 22), leading to accumulation of cholesterol, foam cell formation, and development of atherosclerotic plaque (23). Also, in vitro models have shown that triglyceride-rich lipoproteins can increase the expression of inflammatory proteins, adhesion molecules, and coagulation factors in endothelial cells and enhance the recruitment and attachment of monocytes, promoting the formation of foam cells [reviewed in (1)]. Further, increased remnant cholesterol is associated causally with ischemic heart disease just as increased LDL cholesterol is (10), and increased remnant cholesterol is associated with whole-body low-grade inflammation, unlike increased LDL cholesterol (11). Taken together, these data therefore make it plausible that increased concentrations of remnant cholesterol or triglyceride-rich lipoproteins in patients with ischemic heart disease could explain part of residual risk with respect to increased mortality, as demonstrated in the present study.

The U-shaped association of increased LDL cholesterol with decreased risk of all-cause mortality is consistent with what has been found in the general population (12), and might be attributed to the fact that we looked at all-cause mortality and not cardiovascular mortality. Indeed, noncardiovascular diseases associated with increased mortality, such as cancer, severe respiratory diseases, and inflammatory diseases, are associated with reduced concentrations of LDL and total cholesterol (24–26). Furthermore, patients admitted to the hospital with cardiovascular disease would most likely have been given statin treatment if they were not already on lipid-lowering therapy, and this could attenuate the effect of high LDL cholesterol at baseline on all-cause mortality risk.

A recent mendelian randomization study has shown that genetically increased remnant cholesterol was associated with low-grade inflammation, whereas genetically increased LDL cholesterol was not (11). This suggests that the inflammatory component of atherosclerosis might be driven by increased remnant cholesterol rather than by increased LDL cholesterol. In support of the present findings, we have previously found a similar observational association between increased remnant cholesterol and increased all-cause mortality in the general population (12), and both observational and causal associations between increased concentrations of nonfasting triglycerides (a marker of increased remnant cholesterol) and all-cause mortality in the general population (4, 6, 7, 27).

In further support of our findings, a recent post hoc analysis of 2 large randomized controlled trials showed that among patients with acute coronary syndrome treated effectively with statins, high fasting triglycerides (a marker of high remnant cholesterol) predict long-term and short-term cardiovascular risk (9). In dal-OUTCOMES, a randomized, double-blind comparison of dalcetrapib with placebo in 15871 patients with a recent acute coronary syndrome, long-term risk increased across quintiles of baseline triglycerides, and the HR for recurring cardiovascular events or death in the highest vs the lowest quintiles was 1.6 (95% CI, 1.3–1.9). Also, in the Myocardial Ischemia Reduction with Acute Cholesterol Lowering (MIRACL) trial, a randomized, double-blind comparison of atorvastatin 80 mg daily with placebo in 2086 patients with acute coronary syndrome, the short-term risk increased across tertiles of baseline triglycerides, with a HR for recurring cardiovascular events or death of 1.5 (1.1–2.2) in the highest vs the lowest tertiles in the atorvastatin arm. Importantly, the relationship between triglycerides and thus remnant particles and cardiovascular risk was independent of LDL cholesterol in both studies.

If following the current guidelines, most physicians would treat patients with dyslipidemia and high risk of cardiovascular disease with statins, and thus focus mainly on reducing total and LDL cholesterol concentrations rather than on reducing triglycerides and remnant cholesterol (6). Increased triglycerides have primarily been treated to prevent acute pancreatitis, and there has not been consensus on whether to treat moderately increased triglycerides to prevent cardiovascular disease, although such treatment is now recommended by several consensus statements (2, 14) and guidelines (6). Our results provide further support for the use of nonfasting triglycerides or remnant cholesterol as potential targets for treatment, particularly in those with ischemic heart disease. In support of this, in subgroup analyses from fibrate trials in participants with baseline triglyceride concentrations >2 mmol/L (>177 mg/dL), a 0.1-mmol/L (9-mg/dL) decrease in triglycerides was associated with a 5% (95% CI, 1%–10%) reduction in major coronary events (28). Furthermore, results from the Evaluation of Cinacalcet Hydrochloride Therapy to Lower Cardiovascular Events (EVOLVE) trial (29) showed a reduction in plasma triglycerides, non-HDL cholesterol, VLDL cholesterol, remnant-like particle cholesterol, and apolipoprotein CIII for participants treated with ω-3 free fatty acids compared with placebo (olive oil); in addition, a recent metaanalysis (30) of 11 randomized controlled trials investigating the effect of high-dose (≥1 g/day) ω-3 fatty acid supplementation for patients with existing cardiovascular disease supplied evidence that long-term high dose ω-3 fatty acid supplementation may be beneficial for secondary prevention to reduce cardiovascular disease risk. Other possible drugs for lowering plasma triglycerides, remnant cholesterol, and triglyceride-rich lipoprotein particles include (in addition to potent statins and niacin), peroxisome proliferator-activated receptor agonists, inhibitors of diacylglycerol O-acyltransferase-1, apolipoprotein C3, proprotein convertase subtilisin/kexin type-9 inhibitors, and microsomal triglyceride protein (2, 6, 31). However, no trial has tested the impact of remnant cholesterol lowering on cardiovascular morbidity or mortality.

A limitation is that we only studied white patients of Danish descent, and therefore our results may not necessarily apply to other racial groups; however, we are not aware of data to suggest that the present results should not be applicable to most races and countries. Another limitation is that we used a single measurement of remnant and LDL cholesterol to categorize the patients, presuming that concentrations were constant over time. Given the variability of remnant and LDL cholesterol concentrations, the use of a single measurement of a patient's lipid profile at study enrollment without repeated sampling will lead to regression dilution bias; thus, the present results should be viewed as minimal estimates, whereas the real risk increase in all-cause mortality for increased remnant cholesterol likely is higher than that observed in the present study. Also, we did not have complete information on statin use at baseline and during follow-up; however, all patients with ischemic heart disease were prescribed high-dose statins (per authors A. Langsted, L.E. Bang, B.G. Nordestgaard). Nevertheless, we do not know which patients stopped taking statins during follow-up, although we suspect that doing so would not be influenced by the individuals' concentrations of lipoprotein cholesterol at baseline. Finally, we lacked information on all relevant cardiovascular risk factors, and our data set was not complete for the risk factors we did have information on; however, in parallel studies of individuals in the general population with adjustment for a large number of relevant cardiovascular risk factors, we observed results similar to those in the present study (12).

Strengths of the present study include a large sample size, no losses to follow-up, use of the hardest end point of all, that is, all-cause mortality, which is registered 100% complete in Denmark, and the use of both calculated and measured remnant cholesterol. It is also a strength that we show similar results for HDL cholesterol, HDL3 cholesterol, HDL2 cholesterol, and apolipoprotein A1, all of which are strongly inversely correlated with increased concentrations of remnant cholesterol, but by themselves are unlikely to be causally related to cardiovascular disease (10, 32). Finally, it is reassuring that apolipoprotein B, which marks LDL and remnant cholesterol combined, has risk estimated to be in between those for LDL and remnant cholesterol; however, it is slightly surprising that small dense LDL cholesterol was not associated with all-cause mortality, as small dense LDL cholesterol (like HDL cholesterol) is inversely correlated with triglycerides and remnant cholesterol and has previously been associated with cardiovascular disease events with the same assay as used here (33).

In patients with ischemic heart disease, we observed that increased concentrations of both calculated and measured remnant cholesterol were associated with increased all-cause mortality, which was not the case for increased concentrations of LDL cholesterol. Our data suggest that increased concentrations of remnant cholesterol explain 8%–18% of residual risk of all-cause mortality in patients suffering from ischemic heart disease. Future intervention studies should therefore be focused not only on lowering LDL cholesterol concentrations, but also on lowering triglycerides and remnant cholesterol.

4 Nonstandard abbreviations

 
• VLDL

  very-low-density lipoprotein

 
• IDL

  intermediate-density lipoprotein

 
• CIHDS

  Copenhagen Ischemic Heart Disease Study

 
• HR

  hazard ratio

 
• MIRACL

  Myocardial Ischemia Reduction with Acute Cholesterol Lowering

 
• EVOLVE

  Evaluation of Cinacalcet Hydrochloride Therapy to Lower Cardiovascular Events.

Acknowledgments

We are indebted to staff and participants of the CIHDS for their important contributions. We thank Mia Hansen, Anders Marcuslund-Reuss, and Anitta Pedersen for excellent technical assistance.

References