Cholesterol, Lipoproteins & Lipids: Understanding CVD Risk

In Sigma Statements by Alan Flanagan17 Comments

This is the first published 'Sigma Statement', which represents what we feel to be the best current interpretation of the evidence related to an important, confusing and/or controversial topic in nutritional science. Click here to learn more about Sigma Statements.
For reference, at the bottom of this page there is a list of all the acronyms used in this statement.

Introduction & Context

This statement is the first of three statements which collectively make up our "Diet & CVD" series. The goal of this series is to examine the relationship between diet and cardiovascular disease (CVD), and specifically coronary heart disease (CHD). But before discussing the connection between diet and CVD outcomes, it is crucial to first examine:

  1. If (and how) blood lipids and lipoproteins are implicated in atherosclerosis development and increased CHD/CVD risk
  2. How diet impacts levels of circulating blood lipids and lipoproteins

In this statement we will address point #1 above; i.e. answering the question:

What are the roles of lipids, cholesterol and lipoproteins in atherosclerosis & CVD development?

The next statement in this series deals with point #2 above, whilst the third statement will circle back to our original concern: how does diet impact CHD/CVD risk?

For clarification, atherosclerosis is the building up of plaque in the arteries. Atherosclerosis development requires lipid-carrying particles (lipoproteins) to penetrate the arterial wall. This is what allows deposits of lipids, cholesterol and other substances to form a plaque. Atherosclerosis can be a precursor to cardiovascular events, including CHD.

Lipids 101

‘Blood lipids’ is the broad term for various lipids that are circulating in the bloodstream, either as free (unbound) molecules or bound to other structures. Because lipids are hydrophobic, they are not soluble in water and thus in order to be transported through the blood they require transport within structures that act as 'carriers'. These structures that transport fats and cholesterol around the body are known as lipoproteins.

These lipoprotein structures are comprised phospholipids, free cholesterol and, importantly, proteins called apolipoproteins. The significance of this will emerge later in this statement. There are seven classes of lipoproteins that are classified according to their size and their density:

  1. Chylomicrons (CM)
  2. Chylomicron remnants (CMr)
  3. Very-low-density lipoprotein (VLDL)
  4. Intermediate density lipoprotein (IDL)
  5. Low-density lipoprotein (LDL)
  6. High-density lipoprotein (HDL)
  7. Lipoprotein(a) or Lp(a) - technically a sub-type of the LDL class

‘Density’ refers to the amount of lipid relative to protein (apolipoprotein) of the particle. For example, the composition of VLDL is roughly 92% lipids and 8% protein; because lipids are large compounds, this means that VLDL are large lipoproteins, but with low-density. Conversely, the composition of HDL is roughly 58% lipids and 42% protein; this high protein composition makes HDL quite ‘dense’, and the smallest of all lipoproteins subclasses. So more lipid and less protein means a larger, less dense particle. Whilse less lipid and more protein means a smaller, more dense particle.

Image credit: A new perspective on lipid research in age-related macular degeneration - Scientific Figure on ResearchGate. Available from:

The main form of fat ingested through the diet is triglyceride (TG), formed of three fatty acids and a backbone of a sugar alcohol, known as glycerol. Triglycerides and dietary cholesterol (albeit smaller amounts) are absorbed into enterocytes (intestinal cells) and packaged into chylomicrons. This is the pathway of lipids coming into circulation from external, dietary sources (referred to as the exogenous pathway). Chylomicrons enter circulation and the triglycerides they carry are broken down by the lipoprotein lipase enzyme (LPL) and either utilised or stored by adipose and muscle tissues. This process reduces the amount of lipids carried in the chylomicron. When the chylomicrons lose their lipid contents like this, what remains are referred to as “chylomicron remnants”,  which are taken up by the liver. 

The liver is the site of the ‘endogenous pathway’, denoting that the liver is where VLDL is formed in order to transport new triglycerides which may be created (from circulating free fatty acids or from an overconsumption of simple sugars). Like chylomicrons, the triglycerides in VLDL are broken down by LPL and utilised or stored by adipose and muscle tissues. As their lipid content is now reduced, VLDL now forms IDL and the process of triglyceride breakdown continues to a point where IDL in turn forms LDL. Because this process begins in the liver and results in the transport of lipids to tissues and progressive evolution of VLDL to LDL, it is often dubbed ‘forward cholesterol transport’. 

Forward Cholesterol Transport

‘Reverse cholesterol transport’ occurs with HDL, which is formed in the liver. HDL gathers cholesterol that effluxes [i.e., leaves the cell having been utilised within it] from cells throughout the body, and can transport cholesterol by either:

  1. Transporting cholesterol directly to the liver
  2. Transferring cholesterol to VLDL and chylomicrons

 In cases where HDL returns cholesterol directly to the liver,  the cholesterol is transported into the liver, where it is oxidised and removed via bile. When cholesterol is taken into the liver, it depletes the HDL particle of its cholesterol content and the HDL particle can then return into circulation to continue gathering cholesterol efflux. 

In the second case, HDL can transfer cholesterol to VLDL and chylomicrons in return for an equal weight of triglycerides. This process is mediated by cholesterol-ester transfer protein (CETP).  These VLDL and chylomicron particles and their cholesterol contents can then be rapidly removed by the liver, clearing excess cholesterol from circulation while preserving HDL in circulation. However, this transfer process can be overburdened, which will be discussed below in more detail in the ‘Lipoprotein Remodelling’ section. 

A key feature of ‘forward cholesterol transport’ is the progressive breakdown of triglycerides carried in chylomicrons and VLDL, leading in turn to the formation of chylomicron remnants, IDL, and LDL. With less lipid in the form of triglycerides, these lipoproteins become characterised by their enrichment with cholesterol. The capacity of cholesterol-enriched lipoproteins to penetrate the arteries is a function of their size. The following list indicates the diameter size of the different lipoprotein classes (nm = nanometers):

  • Chylomicrons: 75–1,200 nm
  • VLDL: 30–80 nm
  • IDL: 25–35 nm 
  • Lipoprotein(a): 25–30 nm 
  • LDL: 18–25 nm
  • HDL: 5–12 nm

The size and density of the lipoproteins is critical to understanding the capacity of these compounds to enter into the artery. Lipoproteins with a diameter of >75nm are too large to penetrate the artery; thus, chylomicrons and large VLDL particles are not atherogenic (‘atherogenic’ meaning capable of forming fatty deposits in the arteries). The smaller particles, namely: VLDL, IDL, LDL, and Lp(a), are all pro-atherogenic lipoproteins. Due to its very small size and density, HDL is also capable of penetrating the arteries, however, as the smallest lipoprotein it also has the capacity to exit via the adventitia of the artery (see image below) and thus HDL does not build up in the artery. Conversely, small  VLDL, IDL, LDL, and Lp(a) are all small enough to penetrate into the arteries, but large enough that they cannot exit via the adventitia. The only way for these lipoproteins to be removed form the artery is by the same route of entry, however this reverse transport goes against the blood pressure gradient, and consequently these lipoproteins and their cholesterol contents become trapped within the arterial wall, initiating the processes of atherosclerosis (the formation and buildup of arterial plaque).


With this basic overview of the role of lipids in atherosclerosis, and thus CHD/CVD, we’ll look briefly at the emergence of the role of cholesterol as the causative agent in atherosclerosis, before looking in slightly more detail at the role of the major lipoprotein classes implicated in CHD/CVD risk. 

Emergence of the ‘Lipid-Heart Hypothesis’

The first suggestion that cholesterol was the causative agent in atherosclerosis was a seminal paper published in 1913 by Nikolai N. Anitschkow, an experimental pathologist, and his student S. Chalatow. Anitschkow and his research group had conducted a series of experiments in rabbits using milks, meats, egg whites, whole eggs and egg yolks; the rabbits only developed arterial lesions in response to the whole eggs and egg yolks, which ultimately led to Anitschkow isolating cholesterol. By isolating cholesterol from egg yolks, and emulsifying it with vegetable oil (which their experiments had shown had no effect), Anitschkow was able to demonstrate that the high-cholesterol feed led to profound increases in blood cholesterol levels, which resulted in the development of arterial lesions. Further experiments revealed a species-specific responsiveness to the effects of their experimental high-cholesterol diets: guinea-pigs rapidly developed atherosclerosis, while rats and dogs were resistant to any adverse effects, due to their highly efficient conversion of cholesterol to bile acids, and are thus resistance to increases in blood cholesterol levels. However, experiments in dogs with experimentally induced low LDL-receptor activity (which clears cholesterol from circulation by uptake into cells) showed that high blood cholesterol levels induced by diet then lead to the development of atherosclerosis. 

The most significant contribution of Anitschkow’s work was clearly demonstrating that the development of atherosclerosis was a two-step process conditional on increases in blood cholesterol levels; only when blood levels of cholesterol were elevated did atherosclerosis occur. However, his work and the implications of his findings were not definitively accepted at the time, largely because his work in rabbits drew criticism that rabbits are herbivores and their habitual diet contains no cholesterol. Thus, questions remained over whether these findings and their clinical implications was relevant for human CHD/CVD. These questions were provided more comprehensive answers in the 1950’s, with metabolic ward studies demonstrating the effects on blood cholesterol of different dietary fats (see part 2 of this series for more detail), and the seminal Framingham Study published in 1957 showing that high total cholesterol was strongly associated with the development of new heart disease in men aged 45-62. These associations were subsequently observed in the Seven Countries Study, correlating elevated blood cholesterol levels with heart disease mortality. 

To determine what is causitive (and what is not) in atherosclerosis development, let us examine the role of each of the following separately:

  1. Total Cholesterol
  2. Chylomicrons & VLDL
  3. Triglycerides
  4. LDL-C, ApoB and LDL-P
  5. HDL-C

Total Cholesterol

Total cholesterol (TC), as the name suggests, is the total cholesterol content within the three major lipoproteins; HDL, LDL and VLDL. The informative value of measuring TC has been questioned; with it being claimed that it is too insensitive to accurately reflect risk factors such as LDL-C. In 2009, the Emerging Risk Factors Collaboration found that replacing TC and HDL with ratios or more specific lipid measures, like the total cholesterol: HDL-C ratio; non–HDL-C, apolipoprotein-B and lipoprotein(a), did not result in any improvement in CVD risk prediction. However, it should be noted that this conclusion was in relation to broad population-wide risk assessment where each measure is examined as an independent biomarker (i.e., does one have more predictive power over another when analysed in a constant context). But actually the links between lipoprotein classes and their cholesterol content may differ (discordance). The nuances of this point will be discussed in further detail below in the section on LDL-C, ApoB, and Lp(a).

So at a population-level, TC still retains value as a clinical measure to be factored into long-term CVD risk assessment, for example the Systematic Coronary Risk Estimation [SCORE] recommended by the European Atherosclerosis Society. The addition of further lipoprotein particles can further refine the accuracy of predictive risk (as we'll discuss later).

One point of controversy that has been discussed since the seminal Framingham paper is that in the Framingham cohort up to 35% of CHD incidence occurred in subjects with a total cholesterol level of less than 5.2mmol/L (<200mg/dL). The fact that CHD incidence can occur in individuals with ’normal’ TC should not, however, be taken to mean that the relationship between TC and CHD is entirely invalid or useless. First, as seen in the image below, there is an exponential increase in risk as TC rises, quadrupling from a TC score of 5mmol/L to 7.8mmol/L (200 to 300mg/dL).


Secondly, and more importantly, is that fact that even with a TC score of <5.2mmol/L (200mg/L), there is high probability that LDL-C may be at a level of >3mmol/L (116mg/dL), which is a sufficient exposure over the course of a 20-40 year period to contribute to atherosclerosis (discussed in LDL section below). This is a crucial factor, because what this indicates, and a fact that has been confirmed in statin interventions, is that the definition of ’normal’ for TC (<5.2mmol/L or 200mg/dL) implies that increased risk only occurs over that point. However, interventions targeting cholesterol lowering indicate that atherosclerosis may still progress in the range that are currently defined as normal for TC.

Thus, while there may be a degree of insensitivity to TC as a means of fully quantifying CHD risk alone, TC remains a potentially valid marker for use in assessments of CHD/CVD risk, especially at a population level. The linear, graded associations between blood TC levels and CHD/CVD mortality is evident from converging lines of evidence. Of particular note in this regard is that populations still living traditional ‘hunter-gatherer’ lifestyles and absent evidence of atherosclerotic CVD often exhibit TC levels often <3.1mmol/L (120mg/dL). 

Chylomicrons and VLDL

As stated previously, the diameter size of chylomicrons and large VLDL (>75nm) renders these lipoproteins too large to penetrate the arteries. Lipoproteins of this size are consequently not atherogenic per se. However, the progressive breakdown of triglycerides carried by these lipoproteins generates ‘remnant lipoproteins’, which are smaller particles that are capable of arterial penetration. As chylomicrons and large VLDL contain predominantly triglyceride (86% and 55%, respectively), the breakdown of these triglycerides reduces their particle size and increases the relative composition of cholesterol. These cholesterol-enriched chylomicrons are pro-atherogenic, as are smaller VLDL and IDL particles. These remnant particles are prevalent in combined hypercholesterolaemia, in which TC levels are elevated and triglycerides are also elevated in the range of 2-10mmol/L (176-880mg/dL), leading to an increase in chylomicron remnants and triglyceride-rich VLDL and IDL.


High circulating triglycerides (TGs) have historically been considered an independent risk factor for CVD. However, after adjusting for non-HDL-cholesterol (which is a simple formula of TC minus HDL-C, with the remaining value representing cholesterol content of all pro-atherogenic lipoproteins) this association has been shown to be null.

Graph based on: Emerging Risk Factors Collaboration - JAMA. 2009;302(18):1993-2000

Evidence for a lack of direct effect of TGs per se comes from the condition, familial chylomicronaemia (FC), where individuals lack the LPL enzyme that breaks down triglycerides. While individuals with FC exhibit significantly elevated triglycerides carried in chylomicrons and large VLDL, atherosclerosis does not develop (instead the primary adverse health outcome association with FC is pancreatitis) due to the inability of these large lipoproteins to penetrate the arteries. Comparisons of the effect of pharmacological lowering of TGs with lowering LDL-C has shown that when assessed as reductions in non-HDL-C, the effect on reducing cardiovascular events is the same. Factoring these lines of evidence together, it appears that circulating TGs as a risk for CVD may in fact be somewhat of a proxy for all atherogenic triglyceride-rich lipoproteins, which may be estimated by non-HDL-C, rather TGs being the issue per se.

This is important when considering the argument some make that low TGs (<2mmol/L;176mg/dL) mean a lipid profile is "good", even in the context of elevated LDL-C (or LDL-P), thus downplaying the effects of elevated LDL-C/LDL-P. However, it is important to stress that elevated atherogenic lipoprotein levels, even in the context of low TGs, is still sufficient to cause atherosclerosis. In this context there is less atherogenic contribution of remnant lipoproteins but LDL, VLDL, and IDL all remain pro-atherogenic.


First it's important to make a distinction between two terms (and markers) that are often erronously used interchangeably:

  1. Low-density lipoprotein: denoted as LDL (or a count of these particles is denoted as LDL-P)
  2. Low-density lipoprotein cholesterol: denoted as LDL-C

The distinguishing factor between these, as the names would suggest, is that LDL refers to the lipoprotein particle itself, whereas LDL-C refers to the cholesterol content within those LDL particles. It is LDL-C that is routinely measured on a standard blood lipid panel. It is possible to measure the number of LDL particles (LDL-P), although this testing is currently uncommon. Rather than a direct measure of LDL-P, it is more common to see a measurement of plasma apolipoprotein B (apoB) levels used to estimate LDL particle concentration. However, as apolipoprotein B is an apolipoprotein attached to not only LDL but also chylomicrons, VLDL, and IDL particles, the apoB measurement assesses total numbers of all of these particles. As will be discussed in a later section, in the majority of cases there is concordance between LDL-C and ApoB or LDL-C and LDL-P (meaning that they scale with one another; e.g. high LDL-C means high LDL-P). However, for some individuals, there will be discordance between these measures. This discordance may explain why data has consistently shown apoB (and LDL-P) to be more strongly associated with CVD than LDL-C.

The causal role of LDL-C in atherosclerosis development in humans has been comprehensively demonstrated by a body of converging lines of evidence. These converging lines of evidence encompass:

  • over 200 prospective cohort studies
  • randomised controlled clinical intervention trials with a collective two-million participants
  • 20-million person-years of follow-up during which over 150,000 CVD events occurred
  • Mendelian randomisation studies* on genetic predispositions to elevated or reduced lifelong exposure to LDL-C

*Click button below for side bar on Medelian randomisation studies

One particularly important aspect to this risk is that it relates not only to the magnitude of elevated LDL-C, but the duration of exposure. Thus, the role of LDL-C in driving atherosclerosis is a cumulative, integrated exposure over the course of the lifespan. This fact underpins why the relative risk reduction of interventions to lower cholesterol in older populations (70-89yrs) is significantly smaller in magnitude than the risk reduction from interventions earlier in life (40-49yrs). 


Evidence from familial hypercholesterolaemia (FH) is itself sufficient evidence for a causal role of LDL-C in atherosclerosis. FH is a genetic condition in which there is a loss of function of the LDL-receptor, which is responsible for cholesterol uptake from LDL into cells and out of the circulation. With this loss of function, LDL-C levels become exponentially elevated, resulting in premature atherosclerosis and, if left untreated, early CHD mortality. FH provides evidence that exposure to LDL-C from early in life leads to atherosclerosis, and the extent is related to the magnitude of the exposure. 

In contrast, there are a number of genetic polymorphisms that result in reduced LDL-C levels over the course of the lifespan. Mendelian randomisation studies have indicated that the reduction in lifelong CHD/CVD risk from different genetic polymorphisms relates to the magnitude by which they reduce LDL-C levels. Analysing each polymorphism based on a 1mmol/L (38.7mg/dL) lower LDL-C level indicates a 54.5% reduction in CHD risk per unit reduction. Different polymorphisms correlate to different magnitude of reductions in LDL-C levels. For example, individuals with a polymorphism in the NPC1L1 gene display an average 0.06mmol/L (2.4mg/dL) lower LDL-C level, while those with a combination of both the NPC1L1 and HMGCR polymorphisms exhibit an average 0.15mmol/L (5.8mg/dL) lower LDL-C. To quantify the effect of exposure to lower LDL-C on CHD risk, it is possible to analyse these different polymorphisms against a standardised unit measure of LDL-C; such analysis highlights that per 0.25mmol/L (9.7mg/dL) reduction in LDL-C, CHD risk is reduced by a similar average of 18%. Consistent with the concept of risk being a cumulative exposure integrating magnitude of LDL-C levels and duration of exposure, the risk reduction for CHD was far greater with earlier exposure to lower LDL than comparable level of LDL-C lowering from statin interventions later in life.

A point sometimes made in relation to the assessment of risk associated with LDL-C relates to the potential discrepancy between estimating LDL-C (from TC and HDL measures) and direct measurement of LDL-C. Under most circumstances for the general population both directly measured LDL-C and estimated LDL-C are strongly correlated. There are exceptions, in particular with high TGs the calculation of LDL-C may underestimate the actual concentration of LDL-C. And also in metabolic syndrome, diabetes, and central abdominal obesity there may be weaker correlation due to the remodelling of LDL into smaller, denser, subparticles that may occur under these conditions (discussed further below).

Finally,  it is worth addressing the popular misconception that only smaller, dense LDL-C particles are atherogenic. This is a misnomer which fails to account for the degree of cholesterol enrichment of larger or smaller LDL-C particles; larger particles will carry more cholesterol, while smaller particles will be depleted of their cholesterol content relative to larger particles. Thus, if a greater number of smaller particles are trapped within the arterial wall, given their lower cholesterol content per particle, the amount of cholesterol deposited is similar to a lesser number of larger particles with a higher cholesterol content per particle. Ultimately, all LDL-C particles are equally atherogenic irrespective of size.

What to Measure?: Comparing LDL-C, ApoB, LDL-P and non-HDL-C

Let's refresh on two points made earlier in this statement:

  1. LDL-P is the number of LDL particles in circulation, whilst LDL-C is a measure of the cholesterol content within those particles.
  2. All lipoproteins have a protein structure wrapped around them, called an apolipoprotein.

There are various types (and sub-types) of apolipoproteins. However, each of the lipoproteins that are atherogenic (LDL, Lp(a), IDL,VLDL & CM) have an apolipoproteinB (apoB) attached to them. Therefore ApoB provides a measure of the actual number of particles for all such atherogenic lipoproteins.

In the 2009 assessment of lipoproteins and CVD risk prediction, the Emerging Risk Factors Collaboration found ApoB levels correlated strongly with non-HDL-C. An important differentiation between non-HDL-C and ApoB is that while non-HDL-C provides a measure of the cholesterol concentration in all atherogenic lipoproteins (i.e. the cholesterol contained within all lipoproteins except for HDL), ApoB is a measure of the number of lipoprotein particles. This mirrors the differentiation between LDL-C and LDL-P.

ApoB may correlate strongly with non-HDL-C (and LDL-C) when ApoB particles contain average cholesterol content. However, when ApoB becomes either depleted or enriched with cholesterol, there is discordance between the values of ApoB, LDL-C, and non-HDL-C. In these circumstances, a direct measure of ApoB provides a better predictor of both coronary arterial calcification (CAC) and CVD risk than LDL-C or non-HDL-C. The subset of the population for whom this discordance appears to be present across cohort studies (and thus for which ApoB provided a stronger predictive risk) is in the range of 10-20% (see images below).

However, independent of discordance, ApoB provides a direct measure of the number of atherogenic lipoproteins in circulation. Consequently, the 2019 European Atherosclerosis Society guidelines recommend direct measure of ApoB in ideal circumstances. This should not be construed as invalidating the clinical utility of ‘traditional’ measures, like LDL-C and calculation of non-HDL-C, but rather the emergence of ApoB provides an additional tool for more refined risk assessment. Indeed, the EAS consensus recommendations still place LDL-C as the primary target of lipid-lowering therapy.


Lp(a) & PCSK9 Inhibitors

The role of Lp(a) has been more difficult to fully elucidate, as interventions targeting lowering Lp(a) levels failed to demonstrate a reduction in risk beyond that associated with lowering atherogenic lipoproteins, despite higher Lp(a) correlating to increased risk. In this respect, Mendelian randomisation studies have shown that lifelong exposure to high levels of Lp(a) is causally associated with CVD risk. The emergence of PCSK9 inhibitors, which inhibit a liver protein that moderates blood cholesterol levels, has provided further insight into Lp(a) as these drugs lead to significant 25-30% reductions in Lp(a). In the FOURIER trial of the PCSK9 inhibitor, evolocumab, the highest risk for CVD was evident in participants with the highest baseline Lp(a) levels, which was independent of LDL-C levels. The greatest reductions in risk were found in those with highest baseline Lp(a) levels. The greatest magnitude of risk reduction was found in participants with the lowest achieved levels of both LDL-C and Lp(a). PCSK9 inhibitors may thus provide an emerging therapeutic intervention to reduce CVD risk in patients with high baseline Lp(a) levels. 

The Enigma of HDL-C

In epidemiology, high HDL levels have been associated with lower risk or CHD/CVD, observations which have generated interest into the potential for deliberate raising of HDL to reduce risk. However, interventions targeting increased HDL-C have found no benefit to reducing CHD/CVD risk. Another aspect of framing that occurs in defence of dietary practices that raise LDL-C levels is that often there is a concurrent increase in HDL-C levels; this is framed as a positive, suggested to mean the increase in LDL-C is not a concern. It would appear, however, that caution may be warranted in this regard. There is emerging evidence suggesting that in effect, HDL-C tracks LDL-C. The increase in HDL-C in this context is thus a compensatory response to the increase in LDL-C, but the increase in LDL-C remains the concern given the causal relationship described above. At this juncture the observations of a benefit to higher HDL-C levels for reducing risk are not supported by direct interventions or Mendelian randomisation studies. This stands in contrast to the causal role of LDL-C in atherosclerosis and CHD/CVD. Currently, primary benefit of considering HDL-C appears to be less as a therapeutic target, and more as a valuable element of the overall risk assessment process.

Lipoprotein Remodelling

Recall that density of lipoproteins is a function of their lipid and protein composition:

  • Increased density = increased protein content and decreased lipid content
  • Decreased density = decreased protein content and increased lipid content

So lipoproteins get smaller as their capacity to hold lipid is reduced, and their density increases as their protein content increases (with less lipid storage).  As mentioned earlier, HDL-C is capable of transferring cholesterol to VLDL and chylomicrons in return for an equal weight of triglycerides (a process mediated by cholesterol-ester transfer protein).  This allows VLDL to remove cholesterol to the liver. However, when circulating TGs increase beyond a certain level (a threshold of 1.5-1.6nM), this usual exchange is impaired, resulting in VLDL overloading HDL and LDL with too much TG. The process of removing these TGs from the HDL particles causes HDL to remodel into small dense subparticles. These small dense subparticles are rapidly catabolized in the liver and result in low circulating HDL levels. LDL also remodels into smaller, denser subparticles which are not well recognised by the LDL-receptor, and result in prolonged periods in circulation and increased atherogenic potential (as the longer an atherogenic lipoprotein is in circulation, the greater its probability of entering the arterial wall).

The result is a clear set of characteristics:

  1. high blood LDL-C
  2. high small, dense LDL subparticles
  3. low HDL-C levels
  4. high circulating TGs

Collectively these characteristics are known as the “atherogenic lipoprotein phenotype” (ALP). This phenotype is particularly prevalent in Type-2 Diabetes, and may be one reason (along with post-prandial hyperglycaemia) for the pronounced increase in CVD risk in that population. The ALP may also be common in metabolic syndrome, in particular with increased visceral fat depots. A final point in relation to lipoprotein remodelling is that it is also often used to frame a narrative that LDL-C is not a major risk per se, but the primary culprit is small, dense LDL sub particles. While there is categorically no doubt that remodelled LDL subparticles are highly atherogenic, it is not exhibited in all persons at risk for CHD/CVD. LDL-C itself remains an established risk factor for CHD/CVD, given the causative role of LDL particles in atherosclerosis development. 


  1. VLDL, IDL, LDL, and Lp(a), can all be considered pro-atherogenic particles.
  2. An overwhelming body of multiple, converging lines of evidence has established a causal role for LDL in atherosclerosis and CHD/CVD progression as a fact, beyond a hypothesis.
  3. Much framing has occurred in the popular and scientific literature to mitigate the role of LDL-C, in particular the ‘low TG, high LDL-C but also high HDL-C’ phenotype. However, an elevated atherogenic lipoprotein load, even in the context of low TGs (and/or high HDL-C), is still sufficient to cause atherosclerosis.
  4. While high HDL-C levels have been shown to be protective in epidemiology, evidence for therapeutic benefits to increasing HDL-C in direct interventions remains lacking. Currently, the value of HDL-C may be its functionality in overall risk prediction. 
  5. While high TGs have strong associations with CHD/CVD, this association is not evident once non-HDL is adjusted for, indicating that it is in fact triglyceride-enriched lipoproteins that are atherogenic. This is supported by evidence from FC, where isolated very high TGs does not result in atherosclerosis, as a function of the size of chylomicrons and large VLDL being unable to penetrate the arteries. 
  6. Remnant lipoproteins warrant consideration, and chylomicron and VLDL remnants may increase significantly in conditions of combined high TC and high TGs. The role of remnant lipoproteins in atherosclerosis remains to be fully elucidated. 
Acronyms Used

CVD = cardiovascular disease
CHD = coronary heart disease
SFA = saturated fat
PUFA = polyunsaturated fat
MUFA = monounsaturated fat
TFA = trans fat
CHO = carbohydrate
TC = total cholesterol
CM = chylomicrons
CMr = chylomicron remnants
VLDL = very-low-density lipoprotein
IDL = intermediate-density lipoprotein
LDL = low-density lipoprotein
HDL = high density lipoprotein
Lp(a) = lipoprotein(a)

Questions about this statement?


  1. I work for the NHS as a Physical Activity Tutor.
    I’m currently working on various Infographics (Arthritis/ Cholesterol) for our patients and I’d be grateful for any input you could possibly give?

    Thank you

  2. Would love a list of studies referred to at the end! Great content though and nice easy t read breakdown of this all too often complicated subject.

    1. Author

      Thanks Kieron! You should find any relevant studies hyperlinked from the body of the main text. We do this rather than reference lists so you can just get a link straight to the PubMed or journal page of the study.

        1. Hi Selina,

          Which links are not working for you? I checked all the links within the article (i.e. any text in red) but they all seem to be working fine.

          Can you please advise on where you’re having difficulty?

  3. Hi, thanks for the amazing content. I had a few questions on the article I was hoping you could help clarify:

    1) It says that chylomicrons carry the TG from exogenous sources. Part of the TG are then broken down by LPL and used for energy or stored, and the remaining is carried by chylomicron remnants to the liver? Is the chylomicron still carrying TGs? And, in the liver, are chylomicrons turned into VLDL or does the liver produce the VLDL? Or is this a different mechanism?

    Then, the article says that “A key feature of ‘forward cholesterol transport’ is the progressive breakdown of triglycerides carried in chylomicrons and VLDL, leading in turn to the formation of chylomicron remnants, IDL, and LDL. With less lipid in the form of triglycerides, these lipoproteins become characterised by their enrichment with cholesterol. ” What sparks the foward cholesterol transport effect, is it driven by energy needs, presence of TG, presence of FFA?


    1. Author

      Hi Gabriel,

      Appreciate you getting in touch, glad you find the content helpful 🙂

      To your questions:

      1) Yes, chylomicron remnants continue to carry TGs, even with the breakdown of TGs – because chylomicrons are primarily composed of TGs – chylomicron remnants are still rich in TGs. Chylomicron TGs taken up in the liver may be stored, which is one of the reasons why the chylomicron pathway is important for the development of fatty liver. The liver itself does produce VLDL, so chylomicrons are not turned into VLDL. However, excess TGs in the liver may be repackaged into VLDL and exported from the liver into the circulation as VLDL-TG.

      2) ‘Forward cholesterol transport’ is a perfectly normal process, as cholesterol is required by different tissues. So the process is one of delivering cholesterol to cells; ‘reverse cholesterol transport’ being driven primarily by HDL after cholesterol efflux back from the cell. So yes, you could say ‘energy needs’ in a broad sense.

      Hope that helps?

      1. “As their lipid content is now reduced, VLDL now forms IDL and the process of triglyceride breakdown continues to a point where IDL in turn forms LDL.”

        So if we decrease the formation of VLDL shouldn’t that decrease LDL concentration? Which in turn means that if triglycerides are low in amount LDL’s concentration will be low too?

        Also, what’s you take on body excreting excess cholesterol via bile?

  4. Do you have any info on all cause mortality and cholesterol? I have been seeing mixed studies and am trying to find more info on it.

    1. Hi Scott,

      This is an issue with important nuance to unpack. Is there a specific aspect you’d like to discuss.

      I’m aware of studies that people sometimes point to, that show either an inverse or U-shaped relationship between LDL-C and ACM, with people interpreting that to mean low LDL-C is “bad” or that lowering LDL-C is a risk. However, this is a misinterpretation. There are a number of reasons for such an observed relationship, for example, such disease states will cause reductions in LDL-C, not the other way around. Similarly, people placed on statins later in life, will see LDL-C drop, but folks placed on medication (vs. not needing medication) will have of course been in a higher risk group. And there are also issues with individual studies too.

      Often I see the ACM point being raised as a distractor from the main discussion of how elevated LDL-C impacts blood lipids, and hence ASCVD.

      This article goes into some of the explanations for this observation, particularly as it relates to cancer:

  5. This statement is incredibly helpful!

    I have a quick question that maybe you can help clarify for me:

    Is there any way to infer whether or not someone is in the minority of those who have a discordance between their ApoB and LDL-C by looking at a standard Lipid lanel (which usually doesn’t include ApoB).

    I ask this because my partner recently had her lipids taken – her TG’s (2.55mmol/L) and Non HDL-C (2.6mmol/L) were above the reference range, yet her LDL-C (1.42 mmol/L) was low.

    Based on these results is it reasonable to infer that she may likely has discordance between her LDL-C and total ApoB?

    She is currently taking a bile acid sequestrate (Cholestyramine) for the management of BAM (bile acid malabsorption) – So i’m not sure if this medication may be somehow responsible for the discordance in her results.

    Thank you so much for all the work you do, your podcast is amazing!

    1. Hey Jordan,

      First, this is certainly something for your partner to discuss with her doctor.

      However, to clarify on your question, did you mean to write “non HDL-C” or “HDL-C” for the 2.6 mmol/L measurement? Because on most standard lipid panels a non-HDL-C that would be high would would be above 4 mmol/L. So perhaps you meant HDL-C?

      If you did, then the only way to know if LDL-C was concordant or discordant with ApoB would be to get a measure of ApoB. Simiarly, non-HDL-C has been shown to be a better predictor than LDL-C, so you could look to see what non-HDL-C is, and if that is in or out of range.

      1. In full it says “Se non HDL cholesterol level” (I assume this is referring to non LDL-C?) – The result was [2.6 mmol/L] and the reference range displayed is [0.00 – 2.50mmol/L]

        It was flagged up as red, however her GP only seemed concerned about her elevated Triglycerides [2.55mmol/L] – with the reference range shown as [0.80 – 1.80mmol/L]

        Her GP advised weight loss, exercise and reducing carbohydrates to lower her Triglycerides.

        Thanks so much for your help and information – I’m currently re-listening to episode 315

        1. Sorry my previous comment had a typo:

          In full it says “Se non HDL cholesterol level” (I assume this is referring to non HDL-C?)***

  6. Hi,

    Truly some of the best and most comprehensive content on the topic.

    Since chylomicrons are too large to be atherogenic, does their inclusion in non-HDL-C negatively impact the value of this measure?


Leave a Comment