Despite major advances in prevention and treatment strategies in recent decades, atherosclerotic cardiovascular disease (ASCVD) remains the leading cause of death worldwide. The known risk factors are largely modifiable, including hyperlipidaemia, smoking, hypertension, diabetes, obesity, and physical inactivity. In the context of hyperlipidaemia, there is overwhelming evidence that low-density lipoprotein cholesterol (LDL-C) is a causal risk factor for ASCVD.1 However, numerous epidemiological and genetic studies have also indicated that remnant cholesterol, the cholesterol contained in triglyceride-rich lipoproteins (TRL), i.e., remnants of chylomicrons, very low-density lipoproteins (VLDL) and intermediate-density lipoproteins, may be causally associated with ASCVD.2 Nevertheless, there is currently no conclusive evidence to support the use of TRL and remnant cholesterol reduction therapy alongside statins for the prevention of ASCVD. Furthermore, TRL penetrate the arterial intima and accumulate in atherosclerotic plaques together with low-density lipoproteins (LDL).3

TRL contain triglycerides (TG), cholesterol, phospholipids, and various apolipoproteins. For clinical purposes, two simple measurements can be used to assess the concentrations of these lipoproteins in plasma: plasma TG and remnant cholesterol. These measurements can be obtained using a standard lipid profile, which consists of plasma concentrations of TG, total cholesterol, LDL-C, and HDL-C. Remnant cholesterol, such as non-HDL cholesterol, can be calculated or estimated at no additional cost.4

The population distribution of plasma TG is skewed towards higher levels. In the Copenhagen general population study, 27% of adults have mild to moderately elevated TG levels of 2 to 10 mmol/l (176–880 mg/dl), while only .1% of adults have very high TG levels >10 mmol/l (880 mg/dl).5 Mild to moderately elevated concentrations are largely due to overweight, obesity, and diabetes mellitus; genetic variations and alcohol consumption also contribute. Very high TG is usually due to poorly controlled diabetes mellitus, alcoholism and, in rare cases, homozygosity for mutations in genes involved in TRL metabolism.

The advantage of TG testing is that it provides a direct and precise measurement of all TG in plasma. However, TG per se are unlikely to be the cause of arteriosclerosis, since, unlike cholesterol, they can be broken down by most cells in the body. Therefore, intuitively, and rightly, many scientists and doctors are sceptical about elevated TG being a cause of atherosclerosis and ASCVD. Furthermore, unlike cholesterol, TG do not accumulate in atherosclerotic plaque. Therefore, high TG levels should simply be considered a marker of high cholesterol levels in TRL. However, a limitation of evaluating raised TG alone is that the cholesterol content relative to TG content in TRL can vary. Therefore, it seems more clinically appropriate to estimate, calculate, or measure the cholesterol content of TRL, i.e., remnant cholesterol.2,6

In this review, we analysed the results of findings from: a) observational studies, which show that elevated remnant cholesterol confers a higher risk of ASCVD beyond conventional risk factors, and b) Mendelian randomisation studies, which show that elevated remnant cholesterol probably causally increases the risk of ASCVD. In addition, we summarised the evidence suggesting effects of TRL beyond remnant cholesterol.

Since the late 1980s, evidence has accumulated to suggest that TG plays an independent role in the development of coronary heart disease.7 An early meta-analysis of six studies demonstrated univariate relative risks of 1.32 and 1.76 for each 1 mmol/L increase in plasma TG levels for men and women, respectively.8 After multivariable adjustment, these relative risks were attenuated to 1.14 and 1.37, respectively, but remained significant.8 A subsequent, larger, meta-analysis including 10,158 patients with incident coronary heart disease and 262,525 participants from 29 studies consistently demonstrated a moderately strong association between TG concentration and risk for coronary heart disease.9 However, other studies have shown a loss of statistical significance after adjustment for other risk factors, such as diabetes, body mass index (BMI), glucose levels, hypertension, and smoking. Nevertheless, the results of these studies also suggested that even slightly elevated TG levels are associated with an increased risk of recurrent ASCVD events in patients treated with statins and should be considered a useful risk marker.10 A 1993 study did not demonstrate an independent association between blood TG levels and coronary heart disease mortality after 12 years of follow-up.11 In the subgroup with low LDL-C and HDL-C levels, TG appeared to have a greater effect on coronary heart disease mortality; however, this interaction lost significance when fasting glucose levels were included in the multivariate model.11 The Chinese Multi-provincial Cohort Study,12 followed 30,378 participants for 15 years and found that raised TG levels predicted coronary heart disease and low HDL-C levels predicted the incidence of ischaemic stroke, but only in patients with low LDL-C levels. It has been traditionally theorised that, although hypertriglyceridaemia may be associated with an increased risk of ASCVD, the association weakens when adjusted for other risk factors, particularly low HDL-C levels, which often accompany elevated plasma TG levels.13 However, even after adjusting for HDL-C levels, elevated TG levels remain a risk factor for cardiovascular disease (CVD).14

In subjects with hypertriglyceridaemia, postprandial triglyceride (TG) levels are elevated throughout the day, and it has also been hypothesised that postprandial TRL and their remnants are important in the pathogenesis of atherosclerosis.2,15 Therefore, it is interesting that postprandial TG have been associated with the risk of ASCVD,16,17 despite the fact that they are quite variable. However, unlike with raised LDL-C levels, the magnitude of the TG elevation does not appear to correlate with the degree of ASCVD risk. In particular, very severe hypertriglyceridaemia per se does not invariably seem to increase the risk of ASCVD, potentially due to the accumulated chylomicrons being too large to enter the arterial intima2 and produce atherosclerosis.

The association between elevated remnant cholesterol, which is calculated from plasma TG levels - a common marker of TRL - and an increased risk of ASCVD is well established.18–20 Furthermore, recent studies that have used direct methods and nuclear magnetic resonance spectroscopy (NMR) to measure remnant cholesterol have confirmed that it is the cholesterol content, rather than the TG content that explains the association between elevated TRL and an increased risk of ischaemic heart disease and myocardial infarction.21–23 This is also supported by biology, as TG can be broken down in the human body through hydrolysis, and do not accumulate in atherosclerotic plaques as cholesterol does, and therefore TG level per se is unlikely to be responsible for the formation of atherosclerosis.2 Elevated plasma TG levels are a reliable indicator of high remnant cholesterol levels in most individuals. In fact, the Friedewald equation, which is used to calculate LDL-C, assumes 1 mg of remnant cholesterol for every 5 mg of plasma TG. However, in individuals with very high plasma TG levels and low non-HDL cholesterol levels, there may be as little as .5 mg of remnant cholesterol for every 5 mg of plasma TG, as these individuals carry much of their TG in large, cholesterol-poor TRL.24 Consequently, higher plasma TG levels do not necessarily reflect higher levels of remnant cholesterol in individuals with low non-HDL cholesterol, a situation that is likely to become more common due to the use of statins, ezetimibe, and PCSK9 inhibitors. To accurately estimate remnant cholesterol in individuals with elevated plasma TG and low non-HDL cholesterol, remnant cholesterol should be measured using direct assays or updated formulas.24 Furthermore, a recent study of the general population in Copenhagen found that direct measurement of remnant cholesterol versus calculated remnant cholesterol identified 5% of individuals (with normal calculated remnant cholesterol) in the general population with cholesterol-rich and TG-poor remnants and at a 1.8-fold higher risk of myocardial infarction.21 It should be noted that statin use is associated with lower levels of remnant cholesterol measured by NMR, but not with calculated remnant cholesterol levels. This illustrates that the calculation of LDL-C and remnant cholesterol from TG levels can be modified by the presence of lipid-lowering therapy. In addition, in a recent primary prevention study of participants at high cardiovascular risk in the PREDIMED trial, baseline TG levels, estimated remnant cholesterol and non-HDL cholesterol, but not LDL-C or HDL-C, were associated with the risk of developing major cardiovascular events, regardless of the intervention group, other clinical phenotypes (obesity and diabetes), lifestyle confounders related to lipid concentrations and cardiovascular risk, and lipid-lowering treatment.18

Recent human genetic studies have provided important information on the contribution of TG to ASCVD. Several genetic approaches, including sequencing of candidate genes, GWAS of common DNA sequence variants, and genetic analysis of TG phenotypes, have revealed new proteins and genetic variants involved in the regulation of plasma TG. Some genetic variants that influence TG levels appear to be associated with an increased risk of ASCVD even after adjusting for their effects on other lipid traits.

Mendelian randomisation studies are a highly valid tool for studying the effect of plasma lipid exposures and for making causal inferences about the associations between lipoproteins and ASCVD risk.25 These designs greatly mitigate residual confounding factors, measurement errors, and reverse causality that often limit traditional observational studies. Several Mendelian randomisation studies have shown that remnant cholesterol calculated from plasma TG is causally associated with an increased risk of ASCVD.26–28 In these studies, genetic variants that increase remnant cholesterol had little effect on LDL-C, while genetic variants that increase LDL-C had little effect on remnant cholesterol. More recent studies using directly measured remnant cholesterol and LDL-C have shown that genetic variants that mimic the effect of statins, ezetimibe, PCSK9 inhibitors, bempedoic acid, and angiopoietin-like protein (ANGPTL)3 significantly reduce both remnant cholesterol and LDL-C.29,30 In contrast, apolipoprotein C3 inhibition, ANGPTL4 inhibition, and lipoprotein lipase (LPL) enhancement can only reduce remnant cholesterol and not LDL-C.29,31

Other new genetic studies involving several proteins that influence lipoprotein lipase function, in the APOA5, APOC3, and ANGPTL3 genes, have also strongly supported the idea that elevated TRL are causally related to ASCVD.6,32 Lipoprotein lipase is the key enzyme in the degradation of TG in plasma, while apoC3, apoA5, and ANGPTL3 modulate lipoprotein lipase function and influence hepatic uptake of remnants.

Previous studies have concluded that genetic variants that reduce TG levels in the LPL gene decrease the risk of coronary heart disease to the same extent as genetic variants that reduce LDL-C in the LDL receptor gene, given the same reduction per mg/dl of apolipoprotein B, apolipoprotein B, which is a marker of the total number of TRL and LDL combined.33 However, new evidence contradicts this finding by identifying that elevated TRL may increase the risk of coronary heart disease in addition to apolipoprotein B;34,35 furthermore, remnant cholesterol was found to increase the risk of ischaemic heart disease more than LDL-C for every 1 mmol/l (39 mg/dl),34 suggesting that the results were not solely due to TRL transporting more cholesterol per particle compared to LDL. Genetic variants in ANGPTL4 and LPL associated with lower TRL levels appear to decrease the risk of coronary heart disease more than variants in ANGPTL3 and APOC3, despite having similar associations with postprandial plasma lipid levels.29,31 As ANGPTL4 and LPL can only influence plasma lipid levels by modulating LPL function, while ANGPTL3 and APOC3 probably also have other effects on lipid metabolism,36,37 it is possible that the effect of LPL on ASCVD risk is not fully explained by its effect on plasma lipid levels. Indeed, genetic variation in LPL associated with lower TRL levels may decrease the risk of non-alcoholic fatty liver disease,38 which in turn may decrease the risk of ASCVD by decreasing endothelial dysfunction and inflammation, or by the effects of reduced ectopic fat deposition in other organs such as the epicardium.39

In the case of genetic variations in APOA5, a twofold increase in genetically determined elevated remnant cholesterol levels was causally associated with a 2.2-fold risk of myocardial infarction and an estimate in observational studies of 1.7-fold26; in the case of a genetically determined twofold increase in non-fasting TG, the corresponding risk increases were 1.9-fold in causal genetic studies and 1.6-fold in observational studies. Similar findings were observed in another large Mendelian randomisation study using a single APOA5 genetic variant.40

In 2008, Pollin et al.41 observed a reduction in TG and remnant cholesterol in APOC3 loss-of-function heterozygotes, and a parallel association with a reduction in coronary artery calcification, a biomarker of subclinical atherosclerosis and predictor of ASCVD. In 2014, a 44% reduction in non-fasting TG and a 41% reduction in ASCVD was observed in heterozygotes for APOC3 loss-of-function mutations in individuals in the general population of Copenhagen.42 In a parallel study published consecutively and including 18 different combined cohorts, APOC3 loss-of-function heterozygosity was found to cause a 39% reduction in TG and a 40% reduction in the risk of ASCVD.43 While it has been suggested that these reductions in CVPA risk could be explained by reductions in LDL-C rather than reductions in TRL,44 in the studies by Pollin et al.,41 Jørgensen et al.,42 and Crosby et al.43 reductions in LDL-C due to APOC3 loss-of-function heterozygosity were only 3%, 9%, and 3%, reductions that are unlikely to explain a 40% reduction in the risk of ASCVD.

Furthermore, other studies have suggested that a reduction in genetically mediated TRLs will not only likely lead to a lower incidence of ASCVD, but will also likely lead to a substantial reduction in all-cause mortality.45

Chylomicrons and large TG-rich VLDL particles cannot penetrate the arterial wall and cause atherosclerotic lesions. Lipoprotein particles in the circulation normally flow in and out of the arterial wall by transcytosis in specialised clathrin-coated vesicles. These transport vesicles have an average diameter of about 100 nm, which allows for the transport of particles ≤70 nm only, such as chylomicron remnants and VLDL. Therefore, these smaller TG-rich remnants can penetrate the arterial intima and bind to the connective tissue matrix and be retained by it.46,47 These remnant particles, which contain both TG and are enriched with cholesterol esters, can be directly taken up by arterial macrophages, inducing a massive cholesterol load and the formation of foam cells in the coronary arteries.48 Therefore, elevated levels of remnant TG in the blood have been linked to the progression of coronary heart disease by directly contributing to the formation and progression of atherosclerotic plaque.49,50

Some researchers suggest that it is the cholesterol content of remnant particles that contributes primarily to the progression of atherosclerosis, rather than TG.6 Although the assessment of the specific contribution of raised TG levels is unclear, several direct mechanisms have been suggested, including smooth muscle damage and impaired vascular repair processes.48,49 TG-rich remnants have also been shown to promote endothelial dysfunction, which is one of the first steps in atherogenesis.51 Furthermore, excess TG in lipoprotein particles could lead to abnormal and dysfunctional reverse cholesterol transport.52,53 Hypertriglyceridaemia is also associated with a preponderance of small, dense LDL particles, reduced levels of HDL-C and, in metabolic syndrome, with abnormalities in HDL composition. HDL particles in some hypertriglyceridaemic states, for example in association with metabolic syndrome, may be dysfunctional with regard to their cholesterol efflux, anti-inflammatory and antioxidant properties.

Furthermore, there is growing evidence that TG, or more specifically TRL and their remnants, undergo LPL-mediated hydrolysis of their TG and produce high concentrations of lipolytic products, such as oxidised free fatty acids, which are associated with an increased risk of atherosclerosis and CVD through various mechanisms.4 These mechanisms include the production of interleukins and proinflammatory cytokines, fibrinogen, coagulation factors (in particular, through the upregulation of tissue factor, plasminogen activator inhibitor-1 and plasminogen activator inhibitor-1 antigen expression) and impaired fibrinolysis. In addition, TRL remnants have been shown to upregulate endothelial expression of adhesion molecules, such as intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1), leading to endothelial monocyte adhesion and increased vascular inflammatory response.32,54 In 2016, a study confirmed the link between elevated blood TG levels, endothelial dysfunction, and subclinical atherosclerosis.55 In a cross-sectional analysis of 4,887 individuals from the FMD-Japan Registry, flow-mediated dilation (a marker of endothelial function) was inversely correlated with serum TG levels (r = −.12, p < .001). Flow-mediated dilation decreased significantly with increasing TG levels, and after adjustment for age, sex, and cardiovascular risk factors, including HDL level, a serum TG level greater than 1.20 mmol/L was independently associated with the lowest quartile of flow-mediated dilation, suggesting that TG is an independent predictor of endothelial function.55 In a study of 1,447 people in China, lower blood TG levels were significantly associated with a reduction in carotid-femoral pulse wave velocity, indicating improved arterial function, as arterial stiffness is a strong predictor of ASCVD.56

In order to cause atherosclerosis and ASCVD, TRL must be able to enter the arterial intima, which ultimately leads to the development of atheromatous plaques and ASCVD. Studies in both humans and animals have shown that medium-sized TRL can enter the intima, albeit at a slightly slower rate than small, dense LDL particles.57 However, once these TRL have entered the intima, they may become trapped in the arterial intima to a greater extent than LDL, possibly due to their larger molecular size, which makes re-entry into the arterial lumen against a blood pressure gradient even more difficult than for LDL particles, or due to becoming trapped by components of the arterial intima.58 When entering and becoming trapped in the arterial intima, it seems likely that lipoprotein lipase, either on the endothelial surface or in the arterial intima, degrades TG, leading to the release of free fatty acids and monoacylglycerols, both of which are toxic to tissues and therefore likely to generate local inflammation.58,59 Furthermore, lipoprotein lipase is expressed in macrophages and foam cells in human atherosclerotic plaques,60 and lipoprotein lipase may play a stimulatory role in the formation of human foam cells from TRL.

High TRL levels can also lead to low-grade systemic inflammation, as indicated by slightly elevated plasma CRP levels. To investigate this, 60,608 individuals in Copenhagen were studied using a Mendelian randomisation design,28 revealing that for every 39 mg/dl increase in remnant cholesterol, there was a 37% increase in CRP in observational studies and a 28% increase in causal studies using Mendelian randomisation. In contrast, for LDL-C, a 39 mg/dl increase was associated with only a 7% increase in CRP level. Therefore, these data demonstrate that elevated remnant cholesterol and TRL are causally associated with the development of low-grade inflammation, whereas elevated LDL-C is not. Supporting this finding, patients with heterozygous familial hypercholesterolemia and genetically elevated LDL levels as a result of an LDL receptor defect showed no difference in CRP levels between patients and controls.61

Therefore, hypertriglyceridaemia could accelerate atherosclerosis through several mechanisms, all of which could increase the risk of ASCVD.

This article is part of the supplement entitled “Reduction of cardiovascular events in patients with hypertriglyceridaemia,” which was sponsored by the Spanish Society of Atherosclerosis, with funding from Amarin.

This study was sponsored by the Spanish Society of Arteriosclerosis with funding from Amarin, which did not participate in the design or preparation of this manuscript.