# Cholesterol efflux capacity: HDL, ABCA1 and biomarker limits

- Canonical: https://nutrifit.health/knowledge/cholesterol-efflux-capacity-hdl-subfractions-abca1
- Language: en
- Published: 2026-10-08T21:19:26.155Z
- Updated: 2026-10-08T21:19:26.155Z
- Author: NutriFit Editorial

> HDL biogenesis and subfractions, ABCA1/ABCG1/SR-BI, cellular CEC assays, inflammatory modifications and hepatic–intestinal transport. Cohorts and trials distinguish functional measurements from proven clinical benefit.

Cholesterol efflux capacity (CEC) measures how well a defined extracellular acceptor preparation removes labelled cholesterol from a defined cell system. It describes an important component of HDL biology that a cholesterol concentration cannot capture. HDL cholesterol (HDL-C) measures cargo; HDL particle concentration measures particle abundance; apoA-I concentration measures a major protein component. None of these quantities is interchangeable with the others, and CEC itself does not measure every protective function attributed to HDL.

The distinction matters because macrophages in an arterial lesion continuously receive cholesterol from retained apoB-containing lipoproteins. They must store, mobilize or export this lipid. HDL and lipid-poor apolipoproteins can accept free cholesterol, potentially limiting foam-cell formation. Yet laboratory efflux is one step of a larger network: subsequent transfer between lipoproteins, hepatic uptake, biliary secretion, intestinal absorption and faecal elimination determine whether cholesterol ultimately leaves the body. A high ex vivo CEC therefore cannot establish that an individual's arteries are being cleared at a particular rate.

Observational studies have associated higher CEC with lower coronary or cardiovascular risk after adjustment for HDL-C. Those findings support investigation of HDL function, but they do not validate a universal diagnostic threshold or prove that increasing CEC with a drug or supplement prevents events. Some subsequent cohorts have not found independent associations. The negative primary outcome of the large AEGIS-II apoA-I infusion trial is particularly important when translating attractive mechanisms into treatment claims.

## HDL is a family of changing particles

### ApoA-I structure and lipid binding

Mature human apoA-I is a 243-residue protein of approximately 28 kDa, encoded by APOA1 within the chromosome 11 apolipoprotein gene cluster. It supplies much of HDL's protein scaffold. Repeated amphipathic alpha-helical segments place hydrophobic side chains against lipid and polar or charged side chains toward water. The N-terminal helix bundle and flexible C-terminal lipid-binding region rearrange as the protein associates with phospholipid surfaces. The fraction of the protein that is helical depends on lipidation and particle environment; a single fixed percentage does not describe every form.

Lipid-free and lipid-poor apoA-I are related but distinct states. The C-terminal region helps initiate membrane interaction, while opening of the helix bundle permits extensive lipid binding. In reconstituted discoidal HDL, two or more apoA-I molecules can arrange around a phospholipid bilayer patch in a belt-like configuration. Spherical plasma HDL requires a different three-dimensional organization around a lipid monolayer and neutral-lipid core. Cross-linking and mass-spectrometry studies support adaptable cage-like arrangements rather than one rigid structure shared by every HDL particle (Huang et al., 2011).

### Nascent HDL, density fractions and electrophoretic fractions

The liver and intestine secrete apoA-I. ABCA1-dependent lipidation generates nascent particles containing phospholipid and free cholesterol. Initially discoidal particles may subsequently acquire a neutral-lipid core and become spherical. Lipid-poor apoA-I can also be regenerated through remodelling of existing HDL, so biogenesis is a cycle rather than a one-way assembly line.

The labels pre-beta and alpha refer to electrophoretic mobility, whereas HDL2 and HDL3 traditionally refer to density fractions. Size-resolved, density-resolved, electrophoretic and NMR-defined categories overlap but are not identical. It is therefore misleading to assign every pre-beta particle one diameter, or equate every HDL3 particle with a particular alpha band. A nascent disc is not simply a small mature sphere.

| HDL preparation or category | Typical organization | Functional interpretation |
|---|---|---|
| Lipid-free or lipid-poor apoA-I, including much pre-beta-1 material | Protein with absent or limited lipid | Efficient acceptor for ABCA1-dependent lipidation; abundance alone does not establish intact protein function |
| Nascent discoidal HDL | Phospholipid bilayer patch stabilized by apoA-I | Early HDL biogenesis; composition and size vary with the donor cell and lipid supply |
| HDL3 density fraction | Relatively small, dense, protein-rich particles | Often effective acceptors for ABCG1-related and passive efflux; assay normalization matters |
| HDL2 density fraction | Larger, more lipid-rich particles | Participates in remodelling and selective hepatic lipid uptake; large size is not synonymous with protection |

Traditional ultracentrifugation separates HDL2 at approximately 1.063–1.125 g/mL and HDL3 at 1.125–1.210 g/mL. Typical diameters are roughly 8.8–12 nm and 7.2–8.8 nm, respectively, but the boundaries depend on the method. The number of apoA-I molecules is variable. Activities measured per milligram of HDL protein, per particle, per apoA-I molecule or per volume of plasma can yield different rankings of the same fractions. Small, dense HDL3 can carry a favourable phospholipid composition and show strong activity in experimental assays; this does not make an HDL3 concentration a universal clinical treatment target (Camont et al., 2013).

### LCAT and formation of the neutral-lipid core

Lecithin:cholesterol acyltransferase (LCAT) is a circulating enzyme produced predominantly by the liver. ApoA-I and the lipid surface support its activity. LCAT transfers an acyl group from phosphatidylcholine to the hydroxyl group of free cholesterol, generating lysophosphatidylcholine and cholesteryl ester:

$$
\text{Phosphatidylcholine}+\text{cholesterol}
\xrightarrow{\mathrm{LCAT}}
\text{lysophosphatidylcholine}+\text{cholesteryl ester}.
$$

The enzyme forms an acyl-enzyme intermediate; the reaction is not simply release of a freely diffusing fatty acid followed by spontaneous cholesterol esterification. The resulting ester is more hydrophobic than free cholesterol and partitions into the particle's core, supporting the transition from disc to sphere. This can help maintain the acceptor surface's cholesterol activity, but it does not reduce surface free cholesterol to zero or guarantee unlimited efflux. LCAT activity, acceptor availability and downstream clearance interact. Severe LCAT deficiency produces major lipoprotein abnormalities, including lipoprotein X, corneal changes and renal disease; these manifestations do not establish a simple linear relationship between LCAT activity and coronary protection.

### CETP, PLTP and lipases continually remodel HDL

Cholesteryl ester transfer protein (CETP) transfers neutral lipids between HDL and apoB-containing particles, including VLDL, remnants and LDL. Exchange of HDL cholesteryl esters for triglycerides is a useful simplified description, but transfer need not be strictly equimolar. Proposed shuttle and bridge mechanisms should not be reduced to an obligatory tunnel-forming dimer. Transfer to apoB particles can contribute to delivery of HDL-derived cholesterol to the liver through receptor-mediated uptake of those particles.

Triglyceride-enriched HDL becomes a substrate for hepatic lipase, which hydrolyses triglycerides and phospholipids, often producing smaller particles and releasing lipid-poor apoA-I. Endothelial lipase preferentially hydrolyses phospholipids. Phospholipid transfer protein (PLTP) transfers surface lipids released during triglyceride-rich lipoprotein lipolysis and participates in HDL fusion, enlargement and release of lipid-poor apoA-I. These processes alter the acceptor pool for subsequent ABCA1-dependent lipidation. Remodelling can facilitate transport or produce changes that are unfavourable in a particular metabolic setting; neither CETP nor PLTP has an invariably harmful or beneficial role independent of context.

### Proteins, lipids and signalling cargo

HDL carries a heterogeneous proteome and lipidome. Proteomic inventories depend on isolation method, analytical sensitivity and contamination control; they do not imply that every particle carries dozens of different proteins simultaneously. ApoA-II, apoE, apoM, PON1 and acute-phase proteins characterize subsets or states of HDL.

Paraoxonase-1 (PON1) is a calcium-dependent HDL-associated enzyme with lactonase, arylesterase and paraoxonase activities. Its name reflects a laboratory substrate, not a complete account of physiological function. ApoA-I and the lipid environment help maintain activity. Experiments support antioxidant effects, but it is too strong to describe PON1 as an established universal hydrolase for every lipid hydroperoxide or to assume that any one artificial-substrate assay measures the entire antioxidant capacity of HDL (Mackness and Mackness, 2015).

ApoM is a lipocalin that binds sphingosine-1-phosphate (S1P), a signalling sphingolipid. HDL-associated S1P can activate endothelial S1P receptors, support adherens junctions and influence Akt/eNOS signalling. A substantial part of plasma S1P is carried by HDL, with albumin carrying another pool. Carrier-dependent receptor trafficking and signalling help explain why HDL-S1P and albumin-S1P need not have identical effects. ApoE can participate in receptor interactions and particle clearance; apoA-II modulates HDL structure and remodelling. These functions are biologically relevant but are not all captured by a cholesterol-efflux assay (Christoffersen et al., 2011).

## How macrophages export cholesterol

Foam cells contain both free cholesterol and cholesteryl esters in lipid droplets. Ester hydrolysis must make free cholesterol available before it can leave the cell. Lysosomal processing, intracellular transport, membrane accessibility and acceptor availability all affect export. The first step of macrophage reverse cholesterol transport is important, but calling efflux the universally rate-limiting step of whole-body cholesterol elimination oversimplifies a network with several possible bottlenecks.

### ABCA1: ATP-dependent lipidation of apolipoproteins

ABCA1 is a large full ABC transporter, with two transmembrane regions, two nucleotide-binding domains and extensive extracellular domains. It couples ATP-dependent conformational changes to phospholipid export and HDL biogenesis. Lipid-free or lipid-poor apoA-I is a particularly efficient partner. Other apolipoproteins, including apoE, can participate in appropriate systems. Mature HDL is generally a less efficient acceptor for this pathway, but the word “exclusively” is too rigid because particles can remodel or contain exchangeable apoA-I.

ABCA1-dependent lipid rearrangement produces an accessible membrane environment and/or exported lipid that apoA-I can solubilize into nascent HDL. ApoA-I binding can also stabilize ABCA1 against degradation. The classic membrane-curvature and microsolubilization model describes how amphipathic protein helices interact with lipid packing defects. It should be presented as a mechanistic model rather than literal proof that apoA-I scrapes a fully specified membrane protrusion in every cell (Phillips, 2014).

Biallelic loss-of-function ABCA1 variants cause Tangier disease, with extremely low HDL, tissue cholesterol accumulation, characteristic orange tonsils and sometimes peripheral neuropathy. Cardiovascular manifestations vary; severe premature atherosclerosis is not inevitable in every affected individual. The disease demonstrates the importance of ABCA1 for HDL formation without reducing the full phenotype to HDL-C alone.

### Structural findings and the limits of an alternating-access cartoon

The cryo-EM structure reported by Qian et al. in 2017, including PDB 5XJY, resolved ABCA1's unusually extensive extracellular architecture. The study was published in Cell (DOI 10.1016/j.cell.2017.05.020). Structural snapshots constrain mechanistic hypotheses but do not directly film every step of lipid export or apoA-I release.

A conventional ABC-transporter cartoon proposes inward access to substrate, ATP-driven nucleotide-domain closure and outward delivery. Applying that entire sequence uncritically to ABCA1 is problematic. Segrest et al. (2022) combined modelling and mutagenesis to propose extraction of phospholipid from the outer membrane leaflet through a gateway and annulus into an extracellular hydrophobic tunnel. Their model differs from a simple inner-to-outer leaflet floppase. It also leaves questions about coupling phospholipid export to cholesterol transfer and nascent HDL assembly unresolved. ATP dependence and the importance of extracellular domains are well supported; the exact lipid-entry route and all conformational transitions should retain their experimental scope. A mutation can alter transport, trafficking or stability without every point mutation causing complete loss of function.

### ABCG1 and mature acceptors

ABCG1 is an ABC half-transporter that functions in a dimeric organization. It redistributes sterol pools and promotes availability of cholesterol for removal by lipid-containing acceptors, including mature HDL. It does not generate HDL from lipid-free apoA-I in the same way as ABCA1. Transport can involve plasma-membrane and intracellular compartments. Protection against selected oxysterols has been demonstrated in experimental systems, but a list of exported sterols must not imply identical handling of every oxysterol in every tissue.

ABCA1 and ABCG1 can cooperate: ABCA1 generates nascent acceptors and ABCG1 promotes further export to lipidated particles. That synergy is not a compulsory one-particle conveyor in which every HDL molecule passes through ABCA1, LCAT and ABCG1 in a fixed order. The contribution of each transporter varies with species, cell differentiation, cholesterol loading and experimental acceptor concentration.

### SR-BI and aqueous diffusion

SR-BI, encoded by SCARB1, binds HDL and supports selective lipid uptake as well as bidirectional free-cholesterol exchange. It does not hydrolyse ATP. Free-cholesterol net movement depends on sterol chemical activity, membrane organization and particle composition, not simply total cholesterol concentration. Proposed nonpolar transfer routes are mechanistic models; a precisely measured universal 10-nm open tunnel should not be presented as established human SR-BI anatomy.

The hepatic role of SR-BI is especially important: selective uptake moves HDL cholesteryl esters into hepatocytes without requiring degradation of the whole HDL particle. Macrophage SR-BI can facilitate efflux under appropriate conditions, but animal experiments do not make it interchangeable with ABCA1 or ABCG1 in macrophage-to-faeces transport.

Aqueous diffusion allows individual cholesterol molecules to desorb from a donor surface, cross an aqueous boundary and partition into an acceptor. Cholesterol's low water solubility makes desorption important, while phospholipid packing and sterol activity influence its rate. A useful conceptual balance is:

$$
J_{\mathrm{net}}=J_{\mathrm{out}}-J_{\mathrm{in}}.
$$

Here the two terms denote outward and inward cholesterol mass flux under the specified conditions. Label can move outward even when a simultaneous inward flux means that net cholesterol mass removal is small. Aqueous diffusion can make a substantial contribution in some assays; a universal 10–15% ceiling or one membrane-desorption half-life is not justified across cell systems.

[![Macrophage cholesterol efflux through ABCA1 to lipid-poor ApoA-I, ABCG1 to mature HDL, bidirectional SR-BI exchange and aqueous diffusion.](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261002-cholesterol-efflux-capacity-hdl-subfractions-abca1-routes-en-v1.png)](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261002-cholesterol-efflux-capacity-hdl-subfractions-abca1-routes-en-v1.png)

### LXR/RXR regulation connects sterol loading to export

LXR-alpha and LXR-beta, encoded by NR1H3 and NR1H2, respond to selected oxysterols and related sterol metabolites. With RXR they regulate response elements in genes including ABCA1, ABCG1 and APOE. CYP27A1-dependent sterol oxidation can contribute to this sensing system. Nuclear-receptor activation changes transcription, while protein turnover and trafficking add further layers of control.

This response can reduce cholesterol accumulation and alter inflammatory signalling in macrophages. Nevertheless, neither immediate conversion of all excess cholesterol to one oxysterol nor a fixed five- to tenfold transporter induction applies universally. LXR activation also affects hepatic lipogenesis. A response that is favourable in an arterial macrophage can therefore be undesirable when the same receptor programme raises triglycerides in the liver.

## What a CEC assay actually measures

### Donor cells, acceptor preparations and tracer choice

Common research assays load J774 macrophages, differentiated THP-1 cells or other donor systems with radiolabelled cholesterol or a fluorescent sterol probe. Some protocols inhibit cellular ACAT/SOAT esterification to emphasize the free-cholesterol pool, then use cAMP stimulation or another induction strategy to increase ABCA1 activity. These are choices in an experimental method, not one universally standardized procedure. Unstimulated J774 cells still have biological activity; they are not necessarily ABCA1-null controls.

ApoB-depleted serum is frequently used as the acceptor. Removing LDL, VLDL and related particles helps focus the assay on HDL-containing preparations, but the remaining serum also contains albumin, exchangeable apolipoproteins and other factors. The result is not exclusively intrinsic HDL quality. CEC to a fixed serum volume combines acceptor quantity, composition and function. CEC to isolated HDL at equal protein concentration asks a different question, and ultracentrifugation or precipitation can itself change the preparation.

Cells are washed and incubated with a specified acceptor concentration for a defined time. Published assays often use a few percent apoB-depleted serum and several hours of efflux, but concentration, duration, tracer equilibration and temperature must be reported rather than borrowed as universal operating conditions. Cell viability and leakage require attention: high tracer release from dying cells is not desirable cholesterol export.

For a simple fractional tracer-efflux calculation:

$$
E=100\frac{M}{M+C},
$$

where M is recovered tracer signal in medium and C is recovered cell-associated signal, with appropriate instrument background and recovery handling. The reported E is a percentage over the specified incubation period, not automatically a percentage per hour. Many protocols subtract no-acceptor release:

$$
E_{\mathrm{corrected}}=E_{\mathrm{acceptor}}-E_{\mathrm{no\ acceptor}}.
$$

A plate-normalized index can then be expressed as:

$$
I_{\mathrm{CEC}}=\frac{E_{\mathrm{sample, corrected}}}{E_{\mathrm{reference, corrected}}}.
$$

The reference is a defined pooled control preparation. An index of 1 indicates equality to that reference under that assay, not 100% export, a population median, or a universal normal value. For illustration, a corrected sample efflux of 8% and reference efflux of 10% yield an index of 0.8. That arithmetic does not diagnose dysfunctional HDL.

The BODIPY-cholesterol method enables sensitive, scalable fluorescence measurements and can emphasize ABCA1-related efflux. Sankaranarayanan et al. (2011) compared it with tritiated cholesterol and demonstrated useful assay performance. Yet the fluorophore changes sterol properties and pathway weighting. Significant correlation does not mean the methods are numerically interchangeable or that correlations above 0.85–0.90 hold in all preparations. Radioactivity and fluorescence assays need their own validation, dynamic range, controls and reference materials.

### Pathway-specific measurements require explicit contrasts

A difference between stimulated and unstimulated donor cells is often used to estimate an ABCA1-dependent component:

$$
E_{\mathrm{ABCA1\ estimate}}=E_{\mathrm{stimulated}}-E_{\mathrm{unstimulated}}.
$$

This is an operational estimate. Induction can change more than one cellular process, and the unstimulated cells may retain ABCA1. Transfected systems, genetic loss-of-function controls or inhibitors can provide complementary information, each with its own limitations. Total CEC is also not necessarily an equal sum of four independent components: donor cells may express little of one pathway, and pathways share accessible lipid pools.

ABCA1-related efflux is strongly influenced by lipid-poor acceptor abundance and apoA-I integrity. ABCG1-related assays depend on lipidated acceptor composition. SR-BI assays can reflect bidirectional exchange. Comparing diseases or interventions therefore requires matching the biological system and normalization. A finding in one pathway cannot be generalized to all HDL functions.

[![Cellular CEC assay with medium M and cell C tracer signals and100M/(M+C), separate from liver, bile, intestine and faecal export.](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261002-cholesterol-efflux-capacity-hdl-subfractions-abca1-assay-en-v1.png)](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261002-cholesterol-efflux-capacity-hdl-subfractions-abca1-assay-en-v1.png)

### Harmonization is essential, but universal thresholds remain unestablished

Important sources of variation include cell passage and differentiation, induction strength, tracer specific activity, acceptor preparation, residual apoB particles, serum versus plasma, incubation time, storage and freeze–thaw history. PEG depletion can retain non-HDL components or alter the acceptor preparation; incomplete depletion is not the only possible analytical problem. Standard J774 assays do not inherently require transfection with a human receptor.

Reference pools, technical replicates, blinded sample placement, recovery controls and within- and between-run coefficients of variation improve reliability. A low CV achieved in one laboratory does not become a universal international performance specification. Agreement across laboratories requires common materials and carefully matched protocols, not simply labelling an index “normalized”. There is no internationally validated clinical decision threshold applicable across all CEC methods.

Tracer redistribution is not identical to net cholesterol mass removal, and an ex vivo acceptor assay does not measure the patient's endogenous macrophage transporter abundance, arterial cholesterol-ester mobilization or hepatic clearance. These limitations are part of the result, not reasons to discard the method. CEC remains valuable for mechanistic and translational research when the measured quantity is defined precisely.

## Epidemiology: promising associations and incomplete consistency

Khera et al. (2011) examined CEC in 203 healthy volunteers assessed for carotid intima–media thickness, 442 patients with angiographically confirmed coronary disease and 351 without such confirmed disease. The inverse association with disease status supported HDL-function research, but this was not the Dallas prospective cohort and did not follow those groups for 9.4 years to establish incident events.

Rohatgi et al. (2014) measured baseline CEC in 2,924 Dallas Heart Study participants without cardiovascular disease. Over a median 9.4 years, higher CEC was associated with lower incident cardiovascular risk. The fully adjusted highest-versus-lowest quartile hazard ratio was 0.33, with 95% confidence interval 0.19–0.55. Describing this as 67% lower relative hazard is appropriate; describing it as a guaranteed 67% reduction in an individual's risk after raising CEC is not. HDL-C did not show the same adjusted association in that analysis.

Saleheen et al. (2015), published in The Lancet Diabetes & Endocrinology, studied 1,745 incident coronary cases and 1,749 controls nested within EPIC-Norfolk. The upper-versus-lower tertile adjusted odds ratio was 0.64 (95% CI 0.51–0.80). CEC correlated with HDL-C at r=0.40 and apoA-I at r=0.22. The numerical association is compatible with a distinct functional dimension, but adjustment does not establish complete biological independence or remove every confounder.

Kuusisto et al. (2022) provided an important counterweight: six cohorts with 14,438 participants and 1,570 incident coronary events did not show an independent CEC association after adjustment including HDL-C. Differences in populations, assay systems and shared determinants complicate comparison. The evidence supports a research biomarker with context-dependent associations, rather than unanimous validation of a clinical surrogate.

| Study | Design and measured question | Principal interpretation |
|---|---|---|
| Khera 2011 | Cross-sectional coronary disease and carotid measurements | Association with prevalent disease; not prospective outcome causation |
| Rohatgi 2014, Dallas | Baseline acceptor CEC and incident cardiovascular events | Highest versus lowest quartile HR 0.33; observational association |
| Saleheen 2015, EPIC-Norfolk | Nested case–control analysis of incident coronary disease | Highest versus lowest tertile OR 0.64; assay and population specific |
| Kuusisto 2022 | Multiple cohorts and genetic analyses | Independent association not consistently reproduced |
| Gibson 2024, AEGIS-II | Randomized apoA-I infusion after myocardial infarction | No significant reduction in the primary 90-day composite outcome |

A useful biomarker must do more than associate with outcomes. Routine application requires reproducibility, incremental clinical information, calibrated decision thresholds and evidence that acting on the result improves care. CEC has not established a general replacement for conventional risk assessment. The separate problem of [apoB particle burden versus LDL cholesterol cargo](/knowledge/apob-vs-ldl-c-atherogenic-particle-number-risk) illustrates why different lipid measurements should retain their own meanings: attractive HDL function does not cancel exposure to atherogenic particles.

## Why raising HDL-C did not reliably prevent events

The historical inverse association between HDL-C and cardiovascular disease motivated HDL-raising therapies. It did not by itself prove that HDL-C was a causal treatment target. Voight et al. (2012) used genetic approaches, including an HDL-raising LIPG variant, to question that simple causal model. Mendelian randomization has instrument and pleiotropy limitations; it does not show that every HDL function is irrelevant.

In ILLUMINATE, torcetrapib raised HDL-C substantially but increased cardiovascular events and mortality. The trial documented blood-pressure, aldosterone and electrolyte effects consistent with important off-target toxicity (Barter et al., 2007). Dalcetrapib in dal-OUTCOMES did not significantly reduce events (Schwartz et al., 2012). Evacetrapib in ACCELERATE likewise produced large lipid changes without improving the primary outcome (Lincoff et al., 2017). Niacin added to contemporary statin therapy in AIM-HIGH and HPS2-THRIVE did not establish the hoped-for outcome benefit and introduced adverse effects.

These trials do not prove that CETP inhibition necessarily traps all HDL as inert, overfilled particles. CETP inhibitors can increase measured efflux, and inhibition can reduce apoB-related lipid exposure. REVEAL found fewer major coronary events with anacetrapib: 10.8% versus 11.8% over a median 4.1 years, rate ratio 0.91 (95% CI 0.85–0.97). Non-HDL-C was also lower, so this result cannot be attributed solely to raising HDL-C. Taken together, the trials separate a concentration change from a proven mechanism-specific clinical benefit; they do not support a blanket story of universal HDL paralysis.

## Inflammation and chemical modification can change HDL function

### MPO oxidation: residues and functions must be distinguished

Myeloperoxidase (MPO) uses hydrogen peroxide and chloride to generate hypochlorous acid and can produce other oxidants depending on substrate availability. ApoA-I is a target in atherosclerotic tissue and in experimental oxidation systems (Zheng et al., 2004). Chlorotyrosine is a useful marker of chlorinating chemistry, but it is not a complete molecular explanation of functional loss.

Tyr192 is a prominent site of chlorination in apoA-I. Methionine and tryptophan oxidation can also affect activity. The consequence depends on residue, oxidation extent, lipidation and experimental endpoint. A universal 75–85% fall in efflux caused only by Tyr192 chlorination is not justified. Met148 oxidation is particularly relevant to impaired LCAT activation: Shao et al. (2008) showed that reversing methionine oxidation could restore activity in their experimental system. This should not be conflated with ABCA1 binding, nor interpreted as proof that an oral antioxidant repairs apoA-I in human plaques.

### Carbamylation in uraemia and smoking

Cyanate/isocyanic acid arising from urea chemistry or MPO-related reactions can modify lysine amino groups to carbamyllysine, also called homocitrulline. High urea concentrations and smoking-related thiocyanate exposure provide plausible contexts for increased modification. The reaction changes charge and protein interactions, but it is not necessary to invent one universal set of four lysines whose modification abolishes all ABCA1 binding.

Holzer et al. (2011) identified carbamylation in lesion-derived HDL and showed that modified HDL could promote macrophage cholesterol accumulation through an SR-BI-dependent pathway. This is distinct from a generic claim that carbamylated HDL necessarily switches to SR-A1 or LOX-1. The findings support a mechanism of dysfunction, while clinical interpretation still depends on tissue versus plasma measurements and assay conditions.

### SAA enrichment is not an invariant toxic switch

Inflammation increases serum amyloid A (SAA), which associates with HDL and changes its protein composition. ApoA-I abundance, lipid composition, particle remodelling and clearance can also change. Displacement of apoA-I is one proposed process, but enrichment may occur without simple wholesale displacement.

Some experimental studies show impaired cholesterol acceptance or anti-inflammatory activity of SAA-rich HDL. Others find preserved or increased efflux for particular pathways, donor cells or normalizations. Lipid-free SAA can itself accept lipid in model systems. Thus SAA concentration alone cannot define HDL as uniformly toxic or justify a universal 60–80% CEC reduction. Acute endotoxin models, chronic inflammatory disease and isolated HDL enriched in vitro are different biological settings (Vaisar et al., 2015; de Beer et al., 2013).

### Glycation and endothelial signalling

Hyperglycaemia and reactive carbonyls such as methylglyoxal can modify apoA-I and other HDL components. Glycation can impair protein interactions, LCAT-related activity or PON1 function in experimental systems. Diabetes also changes triglyceride metabolism, particle composition, inflammation and renal function, making attribution to one chemical modification difficult. Fixed percentages of PON1 inactivation should not be generalized to every patient with diabetes.

Normal HDL can support endothelial NO signalling through SR-BI-associated pathways and S1P receptors. HDL from coronary disease populations can show impaired eNOS activation and altered inflammatory signalling; Besler et al. (2011) linked these findings to mechanisms including reduced PON1 activity and LOX-1-related signalling. Changes in eNOS phosphorylation, oxidative stress and uncoupling are mechanistically meaningful, but not every modified HDL preparation triggers an identical cascade or turns endothelial NO production into superoxide production. These assays complement CEC rather than duplicate it. The relationship between glucose exposure and vascular damage is discussed separately in [glycaemic variability and postprandial endothelial effects](/knowledge/glycemic-variability-cgm-postprandial-endothelial-damage).

### S1P and microRNAs add functions beyond efflux

ApoM-bound S1P can help maintain endothelial junctions, influence Akt/eNOS signalling and limit inflammatory responses. S1P content and carrier context can change with disease. A severalfold reduction reported in one population cannot become a diagnostic threshold for all dysfunctional HDL.

HDL can also carry regulatory microRNAs. Tabet et al. (2014) demonstrated transfer of miR-223 and regulation of endothelial ICAM-1 expression. IKK-beta is an intracellular kinase rather than a receptor. A specific endothelial ICAM-1 result does not establish a universal macrophage pathway; cellular delivery mechanisms and targets require direct evidence. The abundance and physiological delivery efficiency of HDL-associated RNA remain important questions. Finding a microRNA in isolated HDL does not establish a validated clinical RNA-based repair pathway or prove that every inflammatory HDL sample has the same protective-to-harmful microRNA switch.

## Efflux is followed by hepatic and intestinal handling

Hepatic SR-BI supports selective uptake of cholesteryl esters without obligatory uptake of the whole HDL protein particle. Lipid removal can remodel the remaining particle, which can re-enter circulation, but it need not always become one fixed disc or HDL3 species. CETP-mediated transfer to apoB particles supplies another indirect hepatic route. These routes coexist in humans; common mouse models differ because mice lack normal human-like CETP activity.

Cholesteryl esters must be hydrolysed to free cholesterol for further use. Neutral esterases contribute in nonlysosomal compartments; lysosomal acid lipase (LIPA) handles substrates reaching lysosomes. Assigning all selective HDL-derived ester hydrolysis to hormone-sensitive lipase and lysosomal acid lipase as interchangeable enzymes ignores compartmental differences. Hepatic free cholesterol can be resecreted, stored, used for membranes, converted to bile acids or secreted into bile.

CYP7A1 initiates the classical bile-acid synthesis pathway. Conversion changes the cholesterol molecule but does not destroy its steroid nucleus. Other enzymes determine subsequent hydroxylation and side-chain shortening. Regulation includes bile-acid feedback through FXR-associated hepatic and intestinal signalling, including intestinal FGF19 in humans. Direct LXR induction of CYP7A1 observed in rodents should not be assumed in humans, whose promoter regulation differs (Goodwin et al., 2003).

ABCG5/ABCG8 forms an obligatory heterodimer that exports neutral sterols across the canalicular membrane into bile and also limits sterol absorption at the enterocyte apical membrane. BSEP/ABCB11 exports bile salts, while ABCB4 supplies phosphatidylcholine to support mixed micelles. These are different substrates and transporters; ABCG5/8 does not export bile acids as its principal described function.

In the intestine, NPC1L1 contributes to uptake of dietary and biliary cholesterol. Absorbed sterol can return through chylomicron pathways. Bile acids have their own efficient enterohepatic circulation; cholesterol absorption should not be equated with bile-acid recycling. Intestinal microbes can convert cholesterol to coprostanol, but not every unabsorbed cholesterol molecule must undergo that conversion before elimination. Transintestinal cholesterol excretion provides an additional route. Faecal sterol output integrates several inputs and does not automatically identify cholesterol originating specifically in arterial macrophages.

SCARB1 genetics illustrates how a high HDL-C concentration can coexist with impaired transport. Zanoni et al. (2016) characterized a rare P376L loss-of-function variant, including a homozygous woman with very high HDL-C, and reported increased coronary odds among carriers. The reported odds ratio of 1.79 was not a finding confined to homozygotes. Other SCARB1 variant studies have shown different risk associations, so neither this example nor LCAT deficiency permits diagnosing every high-HDL individual with proatherogenic HDL. Rare genetic phenotypes help distinguish cargo accumulation from transport flux while preserving genotype-specific uncertainty.

## Interventions: distinguish pharmacodynamics from clinical benefit

### ApoA-I infusions and mimetic peptides

ApoA-I Milano contains an Arg173Cys substitution that permits disulfide-linked forms, unlike ordinary mature apoA-I. In the small Nissen et al. (2003) pilot, five weekly ETC-216 infusions were followed by IVUS changes. Fifty-seven participants were randomized and 47 completed the protocol. The combined treatment group showed a 1.06-percentage-point reduction in percent atheroma volume from baseline; the separate 4.2% figure described reduction of total atheroma volume from baseline. These are different measurements, and a within-group change is not proof of superiority for clinical events. Later Milano formulations and trials must be considered separately; MILANO-PILOT did not establish an incremental atheroma benefit over contemporary therapy (Nicholls et al., 2018).

CSL112 uses plasma-derived human apoA-I reconstituted with phospholipid; it is not simply recombinant Milano. AEGIS-I demonstrated large, transient increases in total and ABCA1-related CEC and evaluated safety after infarction (Gibson et al., 2016). Those assays supported a pharmacodynamic effect, not direct measurement of rapid plaque emptying.

AEGIS-II then randomized 18,219 high-risk patients after acute myocardial infarction to four weekly CSL112 or placebo infusions. The primary 90-day composite of cardiovascular death, myocardial infarction or stroke occurred in 4.8% versus 5.2%; HR 0.93 (95% CI 0.81–1.05), P=0.24 (Gibson et al., 2024). Exploratory or later analyses do not convert that primary result into a proven general indication. The result does not disprove every role of efflux, but it decisively limits the claim that increasing this biomarker automatically prevents events.

Short amphipathic apoA-I mimetics such as 4F/D-4F have been investigated for lipid interactions and anti-inflammatory effects. Strong binding to selected oxidized lipids in vitro does not establish a universal millionfold superiority, human plaque regression or outcome efficacy. An apoJ mimetic paper is not interchangeable evidence for an apoA-I peptide. Formulation, exposure and clinical endpoints matter.

### Nuclear-receptor agonists

LXR agonists can increase ABCA1/ABCG1 expression in experimental macrophages, but systemic activation can stimulate SREBP-1c-dependent hepatic lipogenesis and hypertriglyceridaemia. Tissue-targeted delivery and receptor-selective strategies are research approaches whose mechanistic promise does not establish routine CEC treatment.

PPAR-alpha agonism affects apoA-I, apoC-III and triglyceride metabolism and can connect to macrophage LXR programmes. A fixed 15–25% CEC increase cannot be assigned to every fibrate-treated patient. Outcome effects must be assessed in the relevant trial population. PROMINENT found no reduction of the primary cardiovascular outcome with pemafibrate in 10,497 participants with diabetes and dyslipidaemia, despite favourable triglyceride changes: HR 1.03 (95% CI 0.91–1.15) (Das Pradhan et al., 2022). Drugs used for other lipid indications should not be repurposed through an unvalidated CEC threshold.

### Diet, omega-3 preparations and activity

The PREDIMED functional substudy assessed 296 high-risk participants after one year of dietary intervention. Both Mediterranean diet arms increased CEC relative to baseline, while the olive-oil-enriched pattern improved selected other HDL functions compared with the low-fat comparator (Hernáez et al., 2017). A whole dietary pattern, a laboratory effect and prevention of clinical events are related research questions, not identical endpoints. The substudy does not establish that one isolated phenol produces all benefits or that CEC mediates the clinical effect.

REDUCE-IT tested prescription icosapent ethyl in a selected population and showed cardiovascular benefit (Bhatt et al., 2019). It did not prove that the benefit was caused by greater ABCG1 efflux, reduced CETP substrate supply or HDL membrane fluidity. The finding cannot be extended automatically to arbitrary EPA/DHA supplements. Lipid incorporation, triglyceride effects and particle composition are plausible biological mechanisms that need their own measurements.

Exercise and weight management can alter lipoprotein lipolysis, surface-lipid transfer, insulin sensitivity and inflammation. These changes may affect HDL composition or CEC, but responses vary with training, population and assay. A mandatory high-intensity regimen or guaranteed regeneration of pre-beta HDL is not established. Their broader health rationale does not depend on achieving one research efflux index.

### Nutraceutical findings do not supply a validated combination protocol

Zhu et al. (2014) studied purified anthocyanins in 122 participants with hypercholesterolaemia, using 320 mg/day for 24 weeks, and reported changes in PON1-related activity and efflux. The trial was published in The Journal of Clinical Endocrinology & Metabolism. A research dose, a laboratory endpoint and a selected population do not constitute a general cardiovascular-prevention prescription. In this trial, the reported change in CEC was 20.0% with anthocyanins versus 0.2% with placebo; these are relative changes in an assay endpoint, not cardiovascular-event reductions.

Quercetin, curcumin, lycopene, coenzyme Q10, pomegranate compounds, resveratrol and N-acetylcysteine have been examined in diverse cell, animal or human settings. Reported AhR/PON1 or PPAR/LXR/ABCA1 effects require the compound, exposure, formulation and model to be specified. They do not establish that combining these agents restores arterial HDL, prevents Tyr192 or Met148 modification in vivo, or lowers cardiovascular events through CEC. Bioavailability and drug interactions further limit direct extrapolation from cell experiments. Piperine-mediated exposure enhancement is not evidence of safe pathway targeting.

A universal stack of curcumin/piperine, pomegranate, high-dose omega-3, NAC and resveratrol is therefore unsupported as a CEC-repair protocol. Likewise, a HOMA-IR target below 2 cannot be used as a universal criterion for restored HDL function. Insulin resistance, smoking, renal disease and inflammation deserve appropriate clinical management on their own evidence, with CEC used only for the question that the validated local research method can answer.

## Interpreting measurements without invented treatment cutoffs

| Measurement | What it contributes | Essential limitation |
|---|---|---|
| Total CEC or a normalized efflux index | Acceptor performance in a specified cell/tracer assay | No universal healthy range or intervention threshold; not whole-body RCT |
| ABCA1-related component | Operational pathway estimate under defined stimulation or control conditions | Units and contrast must be reported; not automatically a hourly rate |
| ApoA-I and HDL-C | Protein abundance and cholesterol cargo | Neither diagnoses complete HDL function; ranges are laboratory and population dependent |
| PON1 activity | Hydrolysis of the named assay substrate | Substrate, temperature, genotype and units matter; antioxidant function is broader |
| SAA | Acute-phase or chronic inflammatory context | Does not by itself prove impaired CEC or uniformly harmful HDL |
| ApoA-I chlorotyrosine/carbamyllysine | Specialized evidence of chemical modification | Plasma and lesion measures differ; no universal treatment cutoff |
| HDL-associated S1P | Selected signalling cargo | Isolation and normalization matter; not an established routine efficacy target |

An HDL-C value above an arbitrary boundary such as 2.3 mmol/L does not alone diagnose dysfunction. A CEC index below 0.8 may describe relative performance in one assay but is not an internationally validated disease threshold. PON1, SAA, chlorotyrosine and S1P measurements also require method-specific and population-specific interpretation. A responsible report states the method, unit, reference preparation, reproducibility and population context before interpreting the number.

CEC enriches the scientific account of HDL by moving attention from cholesterol cargo toward an experimentally measurable function. Its strongest use is a clearly defined mechanistic or research question. Clinical protection depends on the entire atherogenic exposure and transport network, and the evidence does not yet justify universal CEC-directed treatment or a supplement programme promising arterial cholesterol clearance.

## How this article was prepared

This narrative scientific synthesis was prepared on 8 October 2026. Targeted checks of primary publications addressed transporter structures, assay interpretation, clinical cohorts and randomized interventions, including AEGIS-II. Accessible journal or PubMed abstracts were used to verify study populations and reported primary outcomes; accessible full texts were examined for selected mechanisms and methodological distinctions. Availability differed across papers, and an abstract check does not establish a complete audit of every analysis or supplement. Reviews supplied biochemical context where appropriate.

This was not a systematic literature search or a GRADE assessment, and no independent external peer review or clinical validation of the article was performed. Cell, animal, observational and randomized evidence are distinguished throughout. Experimental effects and trial doses are reported in their study context and do not define a universal prevention protocol or a validated CEC treatment target.

## FAQ

### Does CEC measure arterial clearance?

No. It measures ex vivo acceptor performance; hepatic uptake, intestinal pathways and net mass flux are separate quantities.

### Does a CEC index below 0.8 diagnose disease?

There is no universal validated cutoff. The index depends on the assay and reference preparation.

### Why does high HDL-C not guarantee protection?

It measures cargo rather than every function or transport flux. Genetic and intervention findings are heterogeneous.

### Is a supplement combination proven to repair CEC?

No universal combination or CEC treatment target is established. Experimental effects and trial doses do not replace clinical outcomes.

## Sources

- [Khera AV et al. Cholesterol efflux capacity, high-density lipoprotein function, and atherosclerosis. N Engl J Med.2011;364:127–135.](https://pubmed.ncbi.nlm.nih.gov/21226578/)
- [Rohatgi A et al. HDL cholesterol efflux capacity and incident cardiovascular events. N Engl J Med.2014;371:2383–2393.](https://pubmed.ncbi.nlm.nih.gov/25404125/)
- [Saleheen D et al. Association of HDL cholesterol efflux capacity with incident coronary heart disease events: a prospective case-control study. Lancet Diabetes Endocrinol.2015;3:507–513.](https://pubmed.ncbi.nlm.nih.gov/26025389/)
- [Kuusisto S et al. Genetic and observational evidence: No independent role for cholesterol efflux over static high-density lipoprotein concentration measures in coronary heart disease risk assessment. J Intern Med.2022.](https://doi.org/10.1111/joim.13479)
- [Huang R et al. Apolipoprotein A-I structural organization in high-density lipoproteins isolated from human plasma. Nat Struct Mol Biol.2011;18:416–422.](https://doi.org/10.1038/nsmb.2028)
- [Camont L et al. Small, dense high-density lipoprotein-3 particles are enriched in negatively charged phospholipids: relevance to cellular cholesterol efflux, antioxidative, antithrombotic, anti-inflammatory, and antiapoptotic functionalities. ATVB.2013;33:2715–2723.](https://pubmed.ncbi.nlm.nih.gov/24092747/)
- [Phillips MC. Molecular mechanisms of cellular cholesterol efflux. J Biol Chem.2014;289:24020–24029.](https://pubmed.ncbi.nlm.nih.gov/25074931/)
- [Qian H et al. Structure of the human lipid exporter ABCA1. Cell.2017;169:1228–1239.e10.](https://doi.org/10.1016/j.cell.2017.05.020)
- [Segrest JP et al. ABCA1 is an extracellular phospholipid translocase. Nat Commun.2022;13:4812.](https://doi.org/10.1038/s41467-022-32437-3)
- [Sankaranarayanan S et al. A sensitive assay for ABCA1-mediated cholesterol efflux using BODIPY-cholesterol. J Lipid Res.2011.](https://doi.org/10.1194/jlr.D018051)
- [Mackness M, Mackness B. Human paraoxonase-1 (PON1): Gene structure and expression, promiscuous activities and multiple physiological roles. Gene.2015;567:12–21.](https://pubmed.ncbi.nlm.nih.gov/25965560/)
- [Christoffersen C et al. Endothelium-protective sphingosine-1-phosphate provided by HDL-associated apolipoprotein M. PNAS.2011;108:9613–9618.](https://pubmed.ncbi.nlm.nih.gov/21606363/)
- [Voight BF et al. Plasma HDL cholesterol and risk of myocardial infarction: a mendelian randomisation study. Lancet.2012;380:572–580.](https://doi.org/10.1016/S0140-6736(12)60312-2)
- [Barter PJ et al. Effects of torcetrapib in patients at high risk for coronary events. N Engl J Med.2007;357:2109–2122.](https://pubmed.ncbi.nlm.nih.gov/17984165/)
- [Schwartz GG et al. Effects of dalcetrapib in patients with a recent acute coronary syndrome. N Engl J Med.2012;367:2089–2099.](https://pubmed.ncbi.nlm.nih.gov/23126252/)
- [Lincoff AM et al. Evacetrapib and cardiovascular outcomes in high-risk vascular disease. N Engl J Med.2017;376:1933–1942.](https://pubmed.ncbi.nlm.nih.gov/28514624/)
- [HPS3/TIMI55-REVEAL Collaborative Group. Effects of Anacetrapib in Patients with Atherosclerotic Vascular Disease. N Engl J Med.2017;377:1217–1227.](https://doi.org/10.1056/NEJMoa1706444)
- [AIM-HIGH Investigators. Niacin in Patients with Low HDL Cholesterol Levels Receiving Intensive Statin Therapy. N Engl J Med.2011;365:2255–2267.](https://doi.org/10.1056/NEJMoa1107579)
- [HPS2-THRIVE Collaborative Group. Effects of extended-release niacin with laropiprant in high-risk patients. N Engl J Med.2014;371:203–212.](https://pubmed.ncbi.nlm.nih.gov/25014686/)
- [Zheng L et al. Apolipoprotein A-I is a selective target for myeloperoxidase-catalyzed oxidation and functional impairment in subjects with cardiovascular disease. J Clin Invest.2004;114:529–541.](https://pubmed.ncbi.nlm.nih.gov/15314690/)
- [Shao B, Cavigiolio G, Brot N, Oda MN, Heinecke JW. Methionine oxidation impairs reverse cholesterol transport by apolipoprotein A-I. PNAS.2008;105:12224–12229.](https://pubmed.ncbi.nlm.nih.gov/18719109/)
- [Holzer M et al. Protein carbamylation renders high-density lipoprotein dysfunctional. Antioxid Redox Signal.2011.](https://pubmed.ncbi.nlm.nih.gov/21235354/)
- [Vaisar T et al. Inflammatory remodeling of the HDL proteome impairs cholesterol efflux capacity. J Lipid Res.2015;56:1519–1530.](https://doi.org/10.1194/jlr.M059089)
- [de Beer MC et al. The Impairment of Macrophage-to-Feces Reverse Cholesterol Transport during Inflammation Does Not Depend on Serum Amyloid A. J Lipids.2013:283486.](https://doi.org/10.1155/2013/283486)
- [Besler C et al. Mechanisms underlying adverse effects of HDL on eNOS-activating pathways in patients with coronary artery disease. J Clin Invest.2011;121:2693–2708.](https://doi.org/10.1172/JCI42946)
- [Tabet F et al. HDL-transferred microRNA-223 regulates ICAM-1 expression in endothelial cells. Nat Commun.2014;5:3292.](https://doi.org/10.1038/ncomms4292)
- [Goodwin B et al. Differential regulation of rat and human CYP7A1 by the nuclear oxysterol receptor liver X receptor-alpha. Mol Endocrinol.2003;17:386–394.](https://pubmed.ncbi.nlm.nih.gov/12554795/)
- [Zanoni P et al. Rare variant in scavenger receptor BI raises HDL cholesterol and increases risk of coronary heart disease. Science.2016.](https://pubmed.ncbi.nlm.nih.gov/26965621/)
- [Nissen SE et al. Effect of recombinant ApoA-I Milano on coronary atherosclerosis in patients with acute coronary syndromes: a randomized controlled trial. JAMA.2003;290:2292–2300.](https://pubmed.ncbi.nlm.nih.gov/14600188/)
- [Nicholls SJ et al. Effect of Infusion of HDL Mimetic Containing Recombinant ApoA-I Milano on Coronary Disease in Patients With an Acute Coronary Syndrome in the MILANO-PILOT Trial: A Randomized Clinical Trial. JAMA Cardiol.2018;3:806–814.](https://pubmed.ncbi.nlm.nih.gov/30046837/)
- [Gibson CM et al. Safety and Tolerability of CSL112, a Reconstituted, Infusible, Plasma-Derived Apolipoprotein A-I, After Acute Myocardial Infarction: The AEGIS-I Trial. Circulation.2016;134:1918–1930.](https://pubmed.ncbi.nlm.nih.gov/27881559/)
- [Gibson CM et al. Apolipoprotein A1 Infusions and Cardiovascular Outcomes after Acute Myocardial Infarction. N Engl J Med.2024.](https://pubmed.ncbi.nlm.nih.gov/38587254/)
- [Das Pradhan A et al. Triglyceride Lowering with Pemafibrate to Reduce Cardiovascular Risk. N Engl J Med.2022.](https://doi.org/10.1056/NEJMoa2210645)
- [Hernáez A et al. Mediterranean Diet Improves High-Density Lipoprotein Function in High-Cardiovascular-Risk Individuals: A Randomized Controlled Trial. Circulation.2017;135:633–643.](https://pubmed.ncbi.nlm.nih.gov/28193797/)
- [Bhatt DL et al. Cardiovascular Risk Reduction with Icosapent Ethyl for Hypertriglyceridemia. N Engl J Med.2019;380:11–22.](https://pubmed.ncbi.nlm.nih.gov/30415628/)
- [Zhu Y et al. Anthocyanin supplementation improves HDL-associated paraoxonase 1 activity and enhances cholesterol efflux capacity in subjects with hypercholesterolemia. J Clin Endocrinol Metab.2014;99:561–569.](https://pubmed.ncbi.nlm.nih.gov/24285687/)