Remnants of the Triglyceride-Rich Lipoproteins, Diabetes, and Cardiovascular Disease

Both type 1 (T1DM) and type 2 (T2DM) diabetes mellitus increase the risk of cardiovascular disease (CVD) (1,2). The causal relationship of LDL cholesterol (LDL-C) with increased CVD risk is well established, and drugs that lower LDL-C levels by increasing LDL receptor activity are widely used to prevent CVD. However, despite the benefit of statins and marked reduction of circulating LDL-C levels, patients with diabetes continue to have more CVD events than patients without diabetes, indicating significant residual CVD risk. A recent new classification system of T2DM, based on different patient characteristics and risk of diabetes complications, divides adults into five different clusters of diabetes (3). Such a substratification of T2DM, which takes insulin resistance and several other factors into account, could perhaps help tailor treatment to those who would benefit the most.

There are likely to be a number of reasons why cholesterol reduction in patients with diabetes is not sufficient to reduce CVD events to the levels found in patients without diabetes. Atherosclerotic lesions in patients with diabetes tend to be more inflammatory, i.e., have greater numbers of macrophages (4). Moreover, after cholesterol reduction, intravascular ultrasound studies show less regression of lesions in people with diabetes (5). Residual CVD risk appears to be at least partly linked to elevated plasma triglycerides and abnormal metabolism of triglyceride-rich lipoproteins (TRLs) (6,7), which are conventionally considered to consist of chylomicrons, VLDLs, and their respective remnant lipoproteins (remnant lipoprotein particles [RLPs]), many of which are present in intermediate-density lipoproteins (IDLs).

Normal fasting levels of plasma triglycerides are defined by current clinical guidelines as <1.69 mmol/L (<150 mg/dL). The definitions of elevated triglyceride levels vary, but fasting triglyceride levels of 1.69–2.25 mmol/L (150–199 mg/dL) are often considered moderately elevated and fasting triglycerides >2.26 or 2.83 mmol/L (200 or 250 mg/dL) are considered high and >5.65 mmol/L (500 mg/dL) severely elevated (8). Elevated plasma triglycerides can be due to increased triglyceride production, decreased lipolysis of triglycerides, and/or reduced clearance of TRLs (7). Moderate hypertriglyceridemia is common in subjects at increased risk of CVD, such as patients with T2DM. However, hypertriglyceridemia strongly associates with a host of other potential risk factors, including obesity, insulin resistance, increased levels of all apolipoprotein (apo)B particles, including RLPs and small-dense LDL, and low levels of HDL cholesterol (HDL-C) (6,9).

More than 40 years ago studies of familial forms of hypertriglyceridemia linked elevated triglyceride levels to an increased risk of CVD (10). Abundant epidemiological evidence associated triglyceride levels with CVD risk (11,12), although the risk was often attenuated when adjusted for potential confounders, such as HDL-C (13). Furthermore, clinical trials of fibrates and n-3 fatty acids (polyunsaturated fatty acids, also often referred to as fish oils) have, with the exception of REDUCE-IT (Reduction of Cardiovascular Events With Icosapent Ethyl–Intervention Trial) and JELIS (Japan EPA Lipid Intervention Study), been mostly negative for CVD benefit (7). The STRENGTH (Outcomes Study to Assess STatin Residual Risk Reduction With EpaNova in HiGh CV Risk PatienTs With Hypertriglyceridemia) trial was recently stopped by the data monitoring board due to a low likelihood of success, but the results have yet to be published. These negative results could have resulted from inclusion of individuals without hypertriglyceridemia in most of the fibrate trials, use of inadequate doses or form (ester versus free fatty acid) in most of the n-3 fatty acid trials, or heterogeneity in the atherogenicity of the triglyceride-carrying lipoproteins, at least in terms of triglyceride as a marker of CVD risk, in the study cohorts. Therefore, there is a need to better define the relationship between plasma triglyceride levels, the apoB lipoprotein particles that carry triglycerides and cholesterol (chylomicrons, VLDL, IDL, and RLPs), and the relative atherogenicity of those lipoproteins versus LDL. Increased understanding of these difficult issues will offer critical insights needed to facilitate our search for additional approaches to CVD prevention and treatment options for hypertriglyceridemic patients, especially those with diabetes.

Genetic studies over the past several years provide strong evidence that elevated triglyceride levels are indeed an independent risk factor for atherosclerosis. Thus, genes involved in TRL metabolism such as lipoprotein lipase (LPL) and genes that control LPL’s activity appear to strongly associate with CVD risk (14). A key early observation was that a null mutation in APOC3, the gene encoding apoC-III (apoC3), which acts in part by inhibiting LPL, associated with reduced triglyceride levels and reduced CVD risk in an Amish population (15). Subsequent studies have confirmed this observation and demonstrated that apoC3 as a CVD risk factor is largely independent of LDL-C (16,17). Other genes linked to elevated levels of triglycerides and CVD risk independently of LDL-C include APOA5 and angiopoietin-like protein 4 (ANGPTL4), an activator and an inhibitor of LPL, respectively (14). Loss-of-function mutations in ANGPTL3, another gene in the angiopoietin-like protein family, is associated with reduced CVD risk and lower triglyceride levels as well as lower levels of LDL-C and HDL-C (familial combined hypolipidemia) (4). It is now widely accepted that hypertriglyceridemia is a risk factor for CVD, although whether elevated triglycerides might be only a biomarker or a direct mediator of CVD remains unclear.

In patients with T1DM, fasting triglyceride levels are usually normal (<150 mg/dL) unless diabetes control is poor (9). There are uncertainties about the role of TRLs in CVD risk in these patients and in patients with T2DM and normal triglyceride levels. However, data from DCCT/EDIC (Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications) suggest that total serum apoC3 and apoC3 in HDL are important predictive biomarkers for CVD in adults with T1DM (18). Moreover, we recently showed that apoC3 is a CVD risk factor in a cohort of T1DM subjects with median triglyceride levels in the normal range (17). Although there was a strong positive correlation between triglyceride levels and serum apoC3 in this cohort, apoC3 appeared to be a stronger risk factor. These observations suggest that apoC3 and triglycerides promote atherosclerosis in T1DM subjects—and perhaps in other subjects with triglycerides generally in the normal range. One possible explanation is that accumulation of lipolytic products of chylomicrons and VLDL, i.e., cholesterol-enriched RLPs, is the reason for elevated CVD risk rather than the more triglyceride-rich precursors of RLPs, chylomicrons and VLDL. Moreover, this hypothesis implies that partial defects in the TRL lipolysis pathway, e.g., due to heterozygous loss of LPL or polygenic causes of reduced lipolysis, would be more atherogenic than total loss of LPL. Thus, very active lipolysis appears to lead to a rapid removal of both chylomicrons and remnants from the blood stream. Heterozygous deficiency of LPL leads to greater postprandial lipemia including the accumulation of more IDLs (19) and chylomicron remnants (20), whereas total LPL deficiency actually prevents the formation of RLPs (21). It is therefore possible that patients with LPL deficiency due to mutations that lead to very low activity levels could be harmed by strategies that activate the small amount of residual activity associated with these mutations. How these mechanisms relate to the residual CVD risk associated with diabetes remains to be established.

The lipoproteins that predominantly transport triglycerides are chylomicrons and VLDL. Chylomicrons are large triglyceride-rich particles secreted by the intestines following ingestion of dietary fat. Their major apolipoprotein is apoB48, a truncated form of apoB100, and their primary role is the delivery of ingested energy, in the form of dietary triglycerides, to peripheral tissues for storage or utilization. In contrast to apoB100 present in VLDL and LDL, apoB48 does not bind to the LDL receptor. Instead, chylomicrons and their smaller remnants are cleared through interactions between apoE, an apo that is abundant in chylomicrons as well as other lipoprotein particles, and several hepatic receptors described below (Fig. 1). VLDLs, which are much smaller than chylomicrons (Fig. 2), are synthesized by the liver. Similar to chylomicrons, VLDL transport triglycerides from dietary sources, but most of VLDL triglyceride derives from lipolysis of adipose tissue and from the conversion of glucose to triglyceride in the liver. VLDL delivers these triglycerides to peripheral tissues in the fasted state. As noted above, VLDL contains apoB100, the major ligand for the LDL receptor, as its major apo. High levels of chylomicrons inhibit clearance of VLDL by substrate competition for LPL. As a result, although apoB48-containing chylomicrons and their remnants contribute the large majority of the rise of triglycerides in the postprandial state, TRLs containing apoB100 account for most of the increase in total apoB lipoproteins observed postprandially (22). Alternatively, rapid lipolysis of chylomicrons and hepatic uptake of exogenously derived triglyceride fatty acids could stimulate increased postprandial hepatic production of VLDL.

Hydrolysis of triglycerides in the core of chylomicrons and VLDL by LPL leads to the formation of triglyceride-depleted particles termed chylomicron remnants and VLDL remnants, respectively (Fig. 1). These RLPs also become depleted in apoCs and relatively enriched in apoE compared with their parent TRLs. The mechanism for these changes appears to be the reduction in volume and surface area of the nascent chylomicrons and VLDL by LPL resulting in the preferential movement of apoCs to HDL as well as the direct addition of apoE that is secreted by the liver (23,24). It is important to keep in mind that RLPs are dynamic products of the ongoing lipolysis of both apoB48-containing chylomicrons and apoB100-containing VLDL and, therefore, consist of a heterogeneous population of particles, varying in size and composition.

Early studies in rats demonstrated the importance of the liver in RLP removal, since hepatectomy resulted in remnant accumulation (25). RLPs are cleared by the liver primarily by an apoE-dependent interaction with the hepatic LDL receptor and LDL receptor–related protein 1 (LRP1). The three apoE isoforms have different affinities for LRP1 and the LDL receptor. In subjects with T1DM and the apoE2/2 isoform, which has a low affinity for LRP1, most of the lipoproteins isolated in the LDL density range are composed of dense remnants and some lipoprotein(a) rather than terminal LDL particles (26), suggesting reduced clearance of RLPs. Furthermore, when apoE2/2 subjects become hyperlipidemic as a result of another primary or secondary abnormality of lipid metabolism, RLPs with both apoB100 and apoB48 accumulate in what is known as remnant removal disease, also known as type III hyperlipoproteinemia, broad band β disease, or dysbetalipoproteinemia (27). In mice, the proteoglycan syndecan 1 also plays an important role in the clearance of both hepatic and intestinally derived TRLs (28). A majority of chylomicron-derived RLPs are cleared by the liver. Between 25% and 75% of VLDL-derived RLPs undergo additional lipolysis and are converted to LDL (Fig. 1).

We define RLPs as postlipolytic partially triglyceride-depleted particles. If derived from chylomicrons, these RLPs contain apoB48, and if derived from VLDL they contain apoB100. As hydrolysis proceeds, the RLPs become progressively smaller and more triglyceride depleted and relatively enriched in cholesterol as a result. Additionally, the longer the RLPs remain in circulation, the greater their enrichment with cholesteryl esters due to the transfer of this lipid species from HDL via cholesterol ester transfer protein (CETP). At the same time, they also become enriched in apoE, as discussed above.

There is little consensus as to how to accurately measure RLPs. This in part is due to the fact that RLPs are defined by metabolic processes that alter the lipid and apolipoprotein composition of chylomicrons and VLDL. Therefore, the numbers and composition of RLPs are constantly changing as a result of ongoing hydrolysis, CETP lipid transfer, and uptake by hepatic receptors. Adding complexity, a majority of chylomicron remnants are removed by the liver, whereas, as noted earlier, 25–75% of VLDL remnants are not directly removed by the liver but, rather, are converted to LDL. Moreover, the accumulation of remnants is likely to differ in people with differing baseline triglyceride levels and is likely to increase with increasing triglyceride levels due to greater competition between chylomicrons and VLDL for both lipolysis and hepatic clearance. Because of the heterogeneity of RLPs and widely varying lipid and protein composition resulting from different degrees of hydrolysis, current physicochemical analytical techniques used for lipoprotein separation do not allow easy separation of chylomicron-derived RLPs from those derived from VLDL. Moreover, these methods cannot completely separate RLPs from their precursors, chylomicrons, or VLDL. Based on the definition used in this review, RLPs should 1) be undetectable or dramatically reduced in individuals with LPL deficiency, 2) have properties similar to the lipoprotein particles that accumulate in the disorder remnant removal disease (type III hyperlipoproteinemia, dyslipoproteinemia, broad band β disease), and 3) be relatively enriched in apoE. The various methods by which RLPs have been identified and quantified so far are described in the next sections.

RLPs often are assessed by ultracentrifugation (29), which relies on relative flotation based on density. Due to their relative triglyceride depletion, RLPs are smaller and denser than their parent particles. Remnants derived from VLDL are found mainly in the IDL (d = 1.006–1.019 g/mL) range after ultracentrifugation, although some are also present in the VLDL density range (d < 1.006 g/mL). Remnants derived from chylomicrons are found in the chylomicron, VLDL, or IDL range (Fig. 2). Thus, there is overlap with other lipoproteins, and ultracentrifugation is therefore an imperfect way to isolate and quantify RLPs (Table 1).

RLPs also can have electrophoretic mobility different from that of their precursor lipoproteins (Fig. 2) due to differences in charge. Plasma from individuals with remnant removal disease exhibit a broad band of material that migrates between β-lipoproteins (LDL) and pre-β-lipoproteins (VLDL) on electrophoresis (30). Hence, the term broad β disease also has been used to describe remnant removal disease (31). β-Migrating VLDLs also are believed to be enriched in RLPs. Gradient gel electrophoresis resolves IDL into two major bands that overlap in size and density. Discontinuous nonequilibrium density gradient ultracentrifugation was able to isolate a series of fractions containing progressively smaller lipoproteins, which showed progressive enrichment in cholesteryl esters, depletion of triglycerides, and slower migration on agarose gels (29), consistent with enrichment of RLPs.

A widely employed immunoaffinity method of measuring RLPs uses a monoclonal antibody to apoB100 that apparently does not bind an apoE-rich population of VLDL containing apoB100 as well as TRLs containing apoB48 (32). This antibody together with a monoclonal antibody to apoA1 is used to bind most of the apoB100-containing lipoproteins (namely, LDL and VLDL) and apoA1-containing lipoproteins (namely, chylomicrons and HDL), leaving behind an unbound fraction of TRLs, including RLPs derived from VLDL and chylomicrons, both of which are enriched in apoE. Cholesterol or triglyceride is measured in the unbound fraction to quantify RLPs (33). Measurement of RLPs by this immunoaffinity method correlates both with triglycerides in the combined VLDL and IDL ultracentrifugal fractions and with RLPs determined electrophoretically. Changes in RLP cholesterol determined by this immunoaffinity method also correlate highly with increment in triglycerides postprandially (34). Although this method has been widely used to determine RLPs in human subjects, the nature of the epitope that allows for precipitation of apoB in VLDL, but not in remnant particles, is not fully characterized, and it is not clear whether the method detects all types of RLPs (Table 1). Although ELISAs have been developed for RLP measurements (35), the validity of these ELISAs has been questioned (36).

Nuclear magnetic resonance (NMR), which is based on computational deconvolution of the proton NMR signal of plasma lipid methyl groups characteristic of lipoproteins, has been used to quantify six subclasses of VLDL and IDL (37). It also has been used recently to measure RLPs (38). Although smaller VLDL subspecies and IDL are enriched in cholesterol relative to large VLDL, inclusion of all VLDL classes is unlikely to selectively quantify RLPs and leads to much greater estimates of “remnant lipoproteins” than other methods, presumably because it also contains nonremnant lipoproteins.

Differential ion mobility analysis can directly determine the size of a broad range of lipoprotein particles (39) and can be calibrated to provide information on particle concentrations and sizes (40). The differential ion mobility analysis method is based on the principle that particles of a given size and charge behave in a predictable manner when carried in a laminar flow of air and subjected to an electric field (39). However, the utility of this approach for determining the sizes and particle concentrations of RLPs has not been established. Differential ion mobility analysis may offer a powerful approach for quantifying RLPs because 1) most remnant particles range in size between LDL and VLDL particles and 2) this approach should readily distinguish between RLPs of different sizes.

Specific measurement of apoB48 in fasting plasma as distinct from apoB100 has been used as a measure of chylomicron remnants derived from the intestines (41). However, measurements of fasting plasma apoB48 do not measure remnants derived from VLDL.

Another method used as a simple approximation of RLPs in nonfasting plasma relies on calculating the sum of VLDL, IDL, and chylomicron cholesterol, using the formula RLP = total cholesterol − HDL-C − LDL-C, but requires specific isolation of LDL by ultracentrifugation (42). However, the “RLP” fraction can also contain chylomicrons, depending on the time after the last meal and the meal’s fat content, as well as IDL and nonremnant VLDL from which the RLPs were derived. While this calculation method is easy and readily applied to large population studies, it is not a specific measure of RLPs.

From the foregoing, it is clear that there is no ideal way of specifically quantifying RLPs. Development of accurate, reproducible methods for the quantification of RLPs would provide a powerful tool for determining the real role of this complex class of lipoproteins in the pathogenesis of atherosclerosis. Achieving this goal depends, however, on agreement of what the size, density, and the lipid and apolipoprotein content of the RLP are in people without the APOE2/2 genotype.

Subjects with remnant removal disease are at a greatly increased risk for CVD (43,44). Remnant removal disease usually requires homozygosity for the APOE2 genotype, resulting in impaired hepatic uptake of APOE-containing lipoproteins (45). In the majority of cases, remnant removal disease is an autosomal recessive disorder that associates with the APOE2/E2 genotype. However, in the absence of additional genetic, hormonal, or environmental factors, RLPs do not accumulate to a degree sufficient to cause hyperlipidemia in APOE2 homozygotes; in fact, hypolipidemia is commonly seen in this situation. RLP accumulation results when the APOE2/2 genotype is accompanied by a second genetic or acquired defect that causes overproduction of VLDL such as obesity or diabetes (45,46), a decrease in RLP clearance, or a reduction in LDL receptor activity (e.g., hypothyroidism) (47). Thus, full phenotypic expression of remnant removal disease requires the presence of other environmental or genetic factors. Interestingly, the reduced uptake of RLPs by the liver results not only in accumulation of RLPs but also, for unclear reasons, in reduced conversion of VLDL and IDL to LDL (27,48). The high prevalence of premature coronary artery disease (43,44) in remnant removal disease is strongly supportive of a direct role for RLPs in atherogenesis. Patients with remnant removal disease, unlike those with familial hypercholesterolemia due to increased LDL, also have increased incidence of peripheral arterial disease (49), as do patients with diabetes.

Increased circulating RLPs are almost certainly related to increased CVD even in the absence of the specific remnant removal disorder. In patients with abnormally high triglyceride levels (e.g., obese patients with T2DM), RLPs have been claimed to be a risk factor for CVD distinct from LDL (6,50,51). Because RLPs are depleted in triglycerides as compared with chylomicrons and VLDL, elevated plasma triglyceride levels are not necessarily an indication of elevated RLPs. Hence, it is possible that increased concentrations of RLPs contribute to atherogenic risk even in subjects with normal triglyceride levels.

Cholesterol is the major lipid that accumulates in atherosclerotic lesions. RLPs have a greater cholesterol content per particle than LDL (7). Chylomicrons and VLDL particles are likely too large to effectively cross the endothelium that lines the artery wall and noneroded lesions of atherosclerosis, in contrast to the much smaller RLPs. Although RLPs are larger than LDL, they likely are small enough to cross the endothelial barrier, where they could be trapped by proteoglycans in the subendothelial space (7). Already in the early 1970s Zilversmit proposed that TRLs and remnants accumulate at arterial foci in proportion to the local concentrations of sulfated polysaccharides (21). In contrast to unmodified LDL, native RLPs promote rapid cholesterol accumulation by macrophages (52,53), suggesting that RLPs are much more potent on a per particle basis than LDL at producing macrophage foam cells in vivo. This might be due to increased uptake of the apoE enriched remnants by non-LDL receptor pathways (52). In summary, while these are good reasons to invoke RLPs in promoting atherosclerosis, uncertainty still exists as to their precise role in promoting atherosclerosis (Fig. 1, bottom).

The major differences in lipid and lipoprotein profiles between T1DM and T2DM have previously been reviewed (54). The hallmark of dyslipidemia in T2MD is high plasma triglycerides and low HDL-C. LDL-C concentrations are normal or only slightly elevated, but the LDL particles are characteristically small and dense with triglyceride enrichment (55). T1DM dyslipidemia primarily occurs in the setting of dysglycemia. With diabetic ketoacidosis, absolute insulin deficiency results in hypertriglyceridemia with low HDL-C and LDL-C levels (56). Well-treated patients with T1DM tend to have normal lipid profiles (54).

T2DM is characterized by increased apoB secretion and elevated plasma VLDL triglycerides (9) (Fig. 1). RLPs, measured by ultracentrifugation and the antibody method alluded to earlier, have also been reported to be increased in T2DM independent of the presence of dyslipidemia (57). The increased VLDL secretion is regulated largely by increased hepatic triglycerides rather than by chronically increased insulin signaling (58). Furthermore, diabetes prevents clearance of TRLs by reducing hepatic uptake of apoB-containing lipoproteins, including RLPs (59), and by increasing levels of apoC3 through relative insulin deficiency or perhaps hepatic insulin resistance (17). Furthermore, hepatic insulin resistance has been shown to result in impaired hepatic LRP1 translocation from intracellular vesicles to the plasma membrane, which could contribute to impaired hepatic clearance of TRLs (60,61) (Fig. 1). Accordingly, in postmenopausal women, diabetes has been shown to decrease the catabolism of chylomicron-derived RLPs, resulting in accumulation of RLPs in plasma (62). Some early studies in diabetic rats injected with radiolabeled chylomicrons demonstrated the accumulation of the cholesteryl ester but not the triglyceride components in plasma, suggesting that chylomicron-derived RLPs accumulate in diabetes (63). Elevated levels of RLPs (detected by electrophoresis) also were observed in cholesterol-fed diabetic dogs (64). However, the role of RLPs per se in diabetic dyslipidemia is unclear, in part due to the difficulty in specifically measuring RLPs. For the same reasons, the potential causative role for RLPs in mediating increased CVD risk associated with diabetes also remains unclear.

How could RLPs accelerate CVD in the setting of diabetes? Early studies identified increased binding of chylomicrons/chylomicron remnants to the artery wall despite a lack of detectable accumulation of RLPs in plasma (65). These findings are consistent with those of a subsequent study showing that retention of injected radiolabeled chylomicron remnants by the arterial intima in diabetic rabbits and rats correlated with hyperglycemia due to insulin deficiency (66). These early studies suggested that entrapment of RLPs within the subendothelial space could in part explain the increased prevalence of atherogenesis in diabetes. Furthermore, macrophage uptake of RLPs from diabetic patients, isolated by immunoaffinity gel using the method described above (33) followed by ultracentrifugation, correlated with the degree of glycemic control in those subjects (67). However, although review articles suggest a role of RLPs in the relationship of diabetic dyslipidemia with the accelerated CVD seen in diabetes (6,13), definite studies supporting the role of RLPs are needed because most studies do not differentiate RLPs from TRLs, for which there is strong evidence of an association with CVD (13). Moreover, very little is known about RLP levels in normolipidemic subjects with or without diabetes, other than those with the apoE2/E2 isoform discussed earlier.

Recent studies have revealed that apoC3 predicts CVD risk in two independent cohorts of subjects with T1DM and fasting plasma triglyceride levels close to the normal range (17,18). Increased plasma apoC3 was associated with a significant increase in CVD risk; this risk was independent of diabetes duration and HbA_1c as well as LDL-C and HDL-C but not of plasma triglycerides (17). Moreover, apoC3, apoE, and apoB accumulated in the atherosclerotic lesions of diabetic mice, consistent with a role for RLPs in atherogenesis (17) (Fig. 1). Macrophages isolated from the peritoneum of diabetic mice contained increased levels of cholesteryl ester, and this accumulation was prevented by suppression of hepatic apoC3 expression (17). Taken together, these observations raise the possibility that apoC3 is an important risk factor for CVD in patients with diabetes and that RLPs containing apoC3 and apoE may play a key role in atherogenesis.

A causative role of circulating triglyceride levels and CVD has been debated for at least 50 years. Recent genetics studies and analyses of subgroups in fibric acid trials are supportive of the triglyceride/CVD hypothesis (7). In addition, beneficial effects of n-3 fatty acids in REDUCE-IT (68) and JELIS (69) have been interpreted in this light. However, n-3 fatty acids are likely to affect processes other than triglyceride levels, and in REDUCE-IT, benefits occurred in subjects whose triglyceride levels were not elevated. Moreover, aside from reducing triglycerides (primarily in VLDL), it is likely that these treatments also reduce RLPs.

Despite decades of experimental data linking RLPs to atherogenesis in animal models and humans (70), the impact of RLPs as causal mediators in CVD associated with hypertriglyceridemia in patients with T2DM is still uncertain. A major limitation of addressing the role of RLPs in CVD is the difficulty in accurately quantifying levels of these lipoproteins, and therefore the number of studies of RLPs’ causal role in diabetic CVD is limited. Once a suitable and specific method becomes available, it will be important to also assess their role as a CVD risk factor in subjects with normal levels of cholesterol and triglycerides. Furthermore, it is becoming clear that not all hypertriglyceridemia is associated with elevated levels of RLPs and that RLPs may be elevated even in people with normal triglycerides and cholesterol (e.g., those with T1DM). New methods are critically needed that can specifically and accurately quantify RLPs derived from different metabolic pathways and in subjects with relatively normal lipid levels to test the long-standing hypothesis that RLPs mediate residual CVD risk associated with diabetes and in other disorders.

Funding. Research in the authors’ laboratories is funded in part by National Heart, Lung, and Blood Institute, National Institutes of Health (NIH), grants R01HL126028, R01HL127694, P01HL092969, R01HL45095, R01HL73029, P01HL128203, R01HL144558, and R35HL135833 and National Institute of Diabetes and Digestive and Kidney Diseases, NIH, grants P30DK017047 and DP3DK108209.

Duality of Interest. K.E.B. has received research grants from Novo Nordisk A/S for unrelated projects. I.J.G. has received support from Arrowhead Pharma. A.C., H.N.G., and T.V. have received funding from MedImmune LLC for unrelated projects. No other potential conflicts of interest relevant to this article were reported.