Triglycerides and Risk for Coronary Heart Disease

Evidence that elevated serum triglyceride levels are associated with increased risk for atherosclerotic events is increasing. In this issue of JAMA, 2 large, long-term prospective cohort studies conducted in different populations by Bansal and colleagues¹ and by Nordestgaard and colleagues² support the role of nonfasting triglyceride levels as a significant risk factor for coronary heart disease (CHD) events. However, a high serum triglyceride level is associated with abnormal lipoprotein metabolism, as well as with other CHD risk factors including obesity, insulin resistance, diabetes mellitus, and lowered levels of high-density lipoprotein cholesterol (HDL-C).³ When determining CHD risk, how important is it to know which came first—high serum triglyceride levels or the risk factors that cause high levels?

Elevated triglyceride levels have a substantial effect on lipoprotein metabolism, which explains much of the controversy about the role of serum triglycerides as a risk factor for CHD.⁴ Even though many studies have demonstrated that triglycerides are associated with CHD risk, the association has been limited due to the interactive effects of high triglyceride levels on other lipoproteins such as HDL-C, levels of which generally are inversely related to those of triglycerides.⁵ This occurs because normal HDL-C particles are produced in the conversion of triglyceride-rich lipoproteins to low-density lipoprotein cholesterol (LDL-C) when serum triglyceride levels are normal. In addition, in the presence of high serum levels of triglycerides, LDL-C and HDL-C become small and dense, creating a highly atherogenic state.⁶ Therefore the question still remains whether high triglyceride levels, or the risk factors associated with high levels, are most important.

Many studies support the finding that triglyceride-rich lipoproteins are equivalent in risk to LDL-C.^{2 ,4 - 6} In normal lipoprotein metabolism, triglyceride-rich lipoproteins are converted into smaller lipoproteins such as LDL-C through the removal of triglycerides. However, not all triglyceride-rich lipoproteins are associated with atherosclerosis.^{2 ,4} Minimal atherosclerotic risk is reported for patients with hyperchylomicronemia or type V hypertriglyceridemia, even though triglyceride levels may exceed 1000 mg/dL (to convert to millimoles per liter, multiply by 0.0113), most likely due to the presence of large lipoprotein particles associated with these disorders.^{2 ,4 ,7} The primary risk associated with triglyceride levels greater than 1000 mg/dL is pancreatitis due to high serum hyperviscosity.

Several factors may account for the variability of risk associated with different lipoprotein disorders characterized by high triglyceride levels. The metabolic abnormalities associated with moderate hypertriglyceridemia (levels of 150-800 mg/dL) are likely related to the types of triglyceride-rich lipoproteins and the presence of small, dense LDL-C and HDL-C particles.^{4 ,6} Triglyceride-rich remnant particles and small, dense LDL-C are highly atherogenic, and small HDL-C particles may not function to reduce CHD risk.⁶ Small, dense LDL-C particles are more likely to penetrate the arterial endothelium and to be taken up by macrophages in the arterial wall than are large LDL-C particles.⁶ Several prospective cohort studies have found that the number of small, dense LDL-C particles is a greater predictor of CHD risk than measured levels of serum LDL-C.^{6 ,8} Thus, risk associated with elevated triglyceride levels may be more a function of the associated lipoprotein disorder than a direct numerical correlation with triglycerides, unlike the linear relationship of risk with LDL-C.⁷

Risk reductions achieved in clinical trials using statins mostly have been correlated with lowering LDL-C levels, with moderate changes in triglyceride levels and modest effects on levels of HDL-C. The majority of patients with premature CHD have lipoprotein disorders that have a combination of elevated triglyceride levels, low levels of HDL-C, and atherogenic LDL-C particles, referred to as the “atherogenic lipoprotein phenotype” due to a strong association with CHD risk.⁹ The atherogenic lipoprotein phenotype is associated with truncal obesity and insulin resistance, referred to as the metabolic syndrome. This syndrome, in which those affected have dyslipidemia, hypertension, and glucose intolerance or diabetes mellitus, has a prevalence of 24% in US adults and 43% of adults older than 60 years.¹⁰

Atherosclerosis is described in some research studies as a postprandial phenomenon. The 2 long-term prospective cohort studies^{1 - 2} published in this issue of JAMA support the concept that nonfasting triglyceride levels more strongly predict CHD risk than levels measured after a 12- to 14-hour fast.^{1 - 2} Postprandial lipoproteins are generally triglyceride-rich, and if an individual has a predisposition to producing remnant particles or small, dense LDL-C and HDL-C particles, or has insulin-resistance, then clearance of these lipoprotein particles can be delayed as long as 12 hours or more.^{2 ,4} Prolonged exposure of the patient's endothelium to triglyceride-rich, atherogenic remnant particles, or the associated states in which atherogenic lipoprotein particles occur (eg, obesity, the metabolic syndrome), may account for why postprandial increases in triglyceride levels account for greater CHD risk.^{1 - 2}

The data from both studies^{1 - 2} in this issue suggest that women have greater risk associated with hypertriglyceridemia than men, as has been suggested,^{9 ,11} but the risk of elevated postprandial triglyceride levels increase risk for both sexes.^{1 - 2} The patients in both studies were not randomized and the nonfasting state was not standardized, which are important concerns meriting replication of studies on the association of triglyceride levels and cardiovascular risk. In addition, the usual limitations of cohort studies, such as lack of specific data on the risk factor mechanisms, lack of direct measurement of insulin resistance, or the possibility of other reasons for elevated triglyceride levels (such as obesity, insulin resistance, and the metabolic syndrome), exist for both studies.^{1 - 2} For example, the metabolic syndrome and increased triglyceride-rich lipoproteins are also associated with a proinflammatory and prothrombotic state due to the presence of atherogenic lipoproteins, clotting factors, and increased plasma viscosity.¹²

Further research is needed to clarify the role of postprandial triglyceride levels and their use in clinical practice. The studies by Bansal et al¹ and by Nordestgaard et al² suggest that using 2- to 4-hour postprandial triglyceride measurements may be more predictive than LDL-C calculated using the Friedewald equation, which requires fasting triglyceride levels.^{1 - 2 ,13} However, fasting triglyceride measurements may be more reliable due to controlled conditions. Postprandial triglyceride levels may need to be measured under specific conditions to improve test reliability, which could add complexity to integrating this test into practice. It would likely be easier for patients and clinicians to assimilate into practice any nonfasting triglyceride level, but a random level may not be as reliable.¹ In addition, clinical trials testing treatment for elevated triglyceride levels may need to include the effects of both baseline and postprandial levels and to measure the effect of specific treatments on reducing postprandial lipoproteins.¹

A simpler choice may be the use of non–HDL-C, as suggested by the final report from the Third National Cholesterol Education Program Adult Treatment Panel (NCEP-ATP III).¹⁴ Non–HDL-C is simply total cholesterol minus HDL-C, which is accurate and reliable in a nonfasting state and would be simple to incorporate into clinical practice.¹⁴ Non–HDL-C is more predictive of CHD risk than LDL-C levels when triglyceride levels are elevated because this measure is a sum of all atherogenic lipoproteins.^{14 - 15} The NCEP-ATP III guidelines recommend the use of non–HDL-C for risk assessment and as a secondary therapeutic goal when the serum triglyceride level is 200 to 499 mg/dL.¹⁴

In the end, is it the triglyceride levels or the associated changes in metabolism that explains the high risk associated with postprandial triglyceride levels? This question is important scientifically, but in clinical practice the argument may be as academic as the debate about which came first, the chicken or the egg. For clinicians, it is important to recognize that when triglyceride levels are between 150 and 1000 mg/dL, the risk for atherosclerosis-related events is significantly increased.^{1 - 2 ,7} Therefore, it is important to aggressively and comprehensively treat patients with dyslipidemias that include high levels of triglycerides, low levels of HDL-C, and the presence of small LDL-C particles, using both lifestyle change and medications if necessary.

This approach is consistent with that recommended in the NCEP-ATP III report, which recommended both lifestyle interventions and pharmacological treatment for triglyceride disorders.¹⁴ Clinical trial evidence increasingly supports this approach, including treating patients with elevated triglyceride levels and low HDL-C levels,^{16 - 17} and using combination therapy (eg, statin plus niacin) to treat combined dyslipidemias.¹⁸ In both the Helsinki Heart Study¹⁶ and the VA-HIT trial,¹⁷ gemfibrozil treatment achieved significant risk reductions for patients with elevated triglyceride levels and low HDL-C levels, without significantly lowering levels of LDL-C. In the Helsinki Heart Study, gemfibrozil reduced the incidence of CHD events in middle-aged men by 71% if the fasting triglyceride level was greater than 200 mg/dL.¹⁶ Trials currently under way, such as the AIM-HIGH trial funded by the National Heart, Lung, and Blood Institute,¹⁹ will further evaluate the role of combination therapy in patients with combined dyslipidemias.