Hypertriglyceridemia and Its Pharmacologic Treatment Among US Adults FREE

The relationship between hypertriglyceridemia and cardiovascular disease has elicited considerable debate. The authors of a meta-analysis of 17 prospective studies in 1996 concluded that an elevated triglyceride (TG) concentration was a risk factor for cardiovascular disease and that the effect was stronger among women than among men.¹ A pooled analysis of 26 prospective cohorts reached similar conclusions.² Finally, a recent meta-analysis using 29 prospective studies conducted in Western populations again noted that TG concentrations were an independent risk factor for coronary heart disease but reported that the associations were similar for men and women.³ Change in TG concentrations has also been shown to result in change in the risk of developing incident coronary heart disease.⁴

If TG concentrations are indeed a risk factor for cardiovascular disease, then it becomes important to establish the prevalence of hypertriglyceridemia in the US population and to learn about the degree of pharmacologic management of this risk factor. Previously, it was reported that the geometric mean TG concentrations among US adults aged 20 to 74 years were 114 mg/dL (to convert to millimoles per liter, multiply by 0.0113) for the periods 1976 through 1980, 116 mg/dL for the period 1988 through 1994, and 122 mg/dL for the period 1999 through 2002.⁵ However, the prevalence of hypertriglyceridemia and the prevalence of the use of medications to lower TG concentration were not reported. The objective of this study was to examine the prevalence of hypertriglyceridemia among US adults 20 years or older and to estimate the degree to which adults with hypertriglyceridemia used prescription medications to manage this hyperlipidemia.

We used data from National Health and Nutrition Examination Survey, 1999 to 2004 (NHANES 1999-2004). During each consecutive 2-year cycle from 1999 to 2004, a national sample was recruited using a multistage, stratified sampling design. The surveys were designed to produce results representative of the civilian, noninstitutionalized US population. The participants were interviewed at home and were invited to attend a mobile examination center, where they were asked to complete additional questionnaires, undergo various examinations, and provide a blood sample. The study received approval from an institutional review board at the Centers for Disease Control and Prevention, and participants were asked to sign an informed consent form. Details about the survey may be found elsewhere.⁶

Fewer than half of the participants were asked to attend a morning examination after an overnight fast. For participants who had fasted at least 8.5 hours and less than 24 hours, serum TG concentrations were measured enzymatically on a Hitachi 704 Analyzer (Roche Diagnostics [formerly Boehringer-Mannheim Diagnostics], Indianapolis, Indiana) at the Lipid Laboratory at the Johns Hopkins University School of Medicine, Baltimore, Maryland, after a series of steps that included hydrolyzing these lipids to glycerol, oxidizing glycerol with glycerol oxidase, and converting hydrogen peroxide, one of the reaction products, to a phenazone with peroxidase. Triglyceride concentrations were categorized as lower than 150 mg/dL, 150 to lower than 200 mg/dL, 200 to lower than 500 mg/dL, 500 mg/dL or higher, and 1000 mg/dL or higher. In addition, we also dichotomized TG concentrations into categories of lower than 150 mg/dL and 150 mg/dL or higher as well as lower than 200 mg/dL and 200 mg/dL or higher for some analyses. The 2001 National Cholesterol Education Program (NCEP) report defined borderline high TG concentrations as 150 to 199 mg/dL and high TG concentrations as 200 to 499 mg/dL.⁷

Participants could report up to 20 prescription medications and were asked to show the containers of these medications to the interviewer. Participants who reported using fenofibrate, gemfibrozil, or niacin were considered to be receiving treatment consistent with lowering TG concentrations.

Covariates included in the analyses were age, sex, race/ethnicity (white, African American, Mexican American, and other), education (<high school, high school diploma or general equivalency diploma, and >high school), insurance status, smoking status (current, former, and never), body mass index, physical activity, alcohol use, intake of carbohydrates (≤60% of total energy intake and >60% of total energy intake), diagnosed diabetes (yes and no), and use of cholesterol concentration–lowering medications. Health insurance status was determined from the following question: “Are you covered by health insurance or some other kind of health care plan? [Include health insurance obtained through employment or purchased directly as well as government programs like Medicare and Medicaid that provide medical care or help pay medical bills.]” Participants answering yes were considered to have health insurance. Participants who had smoked at least 100 cigarettes during their lifetime and were currently smoking were designated as current smokers. Those who had smoked at least 100 cigarettes during their lifetime and were not currently smoking were designated as former smokers. Those who had not smoked at least 100 cigarettes during their lifetime were designated as never having smoked. Body mass index was calculated as weight in kilograms divided by height in meters squared and categorized as lower than 18.5, 18.5 to lower than 25, 25 to lower than 30, and 30 or higher. Alcohol use was derived from a series of questions about the use of alcoholic beverages during the past 12 months. Weekly times spent in transportation-related physical activities (in response to the question “Over the past 30 days, have you walked or bicycled as part of getting to and from work or school or to do errands?”), house- or yard-related activities, or exercise, sports, and physically active hobbies were summed and dichotomized into categories of 150 minutes per week or more and less than 150 minutes per week.⁸ Carbohydrate intake was determined from a single 24-hour dietary recall for 1999-2000 and 2001-2002 and from two 24-hour recalls for 2003-2004. Participants who responded affirmatively to the question “Have you ever been told by a doctor or health professional you have diabetes or sugar diabetes?” were considered to have diagnosed diabetes. Those who answered that they had not been so told or that they had borderline diabetes were not considered to have diagnosed diabetes. Participants who reported using a statin (atorvastatin, cerivastatin, fluvastatin, lovastatin, pravastatin, rosuvastatin, or simvastatin) or ezetimibe were considered to be receiving medications to lower cholesterol concentration.

The analyses were limited to men and women 20 years or older. Differences in means or percentages were calculated by using unpaired t tests or χ² tests, respectively. Tests for linear trend were conducted by using orthogonal linear contrasts. Prevalence rate ratios using the log-binomial method were calculated to assess the independent association between the presence of hypertriglyceridemia (dependent variable) and various study variables. For the analyses to account for the complex sampling design, SUDAAN (Software for the Statistical Analysis of Correlated Data; Research Triangle Institute, Research Triangle Park, North Carolina) was used.

Of the 15 332 participants 20 years or older in NHANES 1999-2004, 6036 were selected for the morning examination and 5934 had values for TG concentration. The exclusion of pregnant women reduced the sample to 5610 participants. The analytic sample included 2837 men and 2773 women. Approximately 9% of women reported currently using oral contraception (16% among women aged 20-50 years), and 14% of women reported using hormone therapy (28% among women 51 years or older). Triglyceride concentrations ranged from 26 to 3780 mg/dL, with a mean of 147.4 mg/dL, a median of 116.7 mg/dL, and a geometric mean of 121.3 mg/dL.

Overall, the percentage (SE) of participants with at TG concentration of 150 mg/dL or higher was 33.1% (0.8%); a concentration of 200 mg/dL or higher, 17.9% (0.7%); a concentration of 500 mg/dL or higher 1.7% (0.2%), and a concentration of 1000 mg/dL or higher, 0.4% (0.1%) (Table 1). Defining hypertriglyceridemia as an elevated TG concentration or the use of prescription medications used to lower TG concentration (ie, fenofibrate, gemfibrozil, or niacin) increased the mean (SE) prevalence of hypertriglyceridemia to 33.5% (0.8%) for a concentration of 150 mg/dL or higher and 18.5% (0.7%) for a concentration of 200 mg/dL or higher.

We also estimated the prevalence of hypertriglyceridemia assuming that the use of statins can decrease TG concentration. Assuming that the use of a statin results in a mean reduction of 7%, the prevalence was 33.7% for of an elevated TG concentration of 150 mg/dL or higher and 18.3% for an elevated concentration of 200 mg/dL or higher. If the use of fenofibrate, gemfibrozil, or niacin was counted as representing hypertriglyceridemia, the prevalence increased to 34.0% and 18.9%, respectively. Assuming that the use of a statin results in a mean reduction of 30%, the prevalence was 35.4% for elevated TG concentration of 150 mg/dL or higher and 20.8% for an elevated concentration of 200 mg/dL or higher. Counting the use of fenofibrate, gemfibrozil, or niacin as representing hypertriglyceridemia increased the prevalence to 35.7% and 21.3%, respectively.

The percentage (SE) of men with a TG concentration of 150 mg/dL or higher was 36.7% (1.4%); a concentration of 200 mg/dL or higher, 21.5% (1.1%); and a concentration of 500 mg/dL or higher, 2.8% (0.4%). Among women, these percentages were 29.6% (1.0%) (P value vs men, <.001), 14.4% (0.7%) (P value vs men, <.001), and 0.8% (0.2%) (P value vs men, <.001), respectively. Among men, a smaller percentage of African Americans (19.5%) had a TG concentration of 150 mg/dL or higher compared with whites (39.0%) or Mexican Americans (40.4%). Similarly, among women, a smaller percentage of African Americans (13.6%) had a TG concentration of 150 mg/dL or higher compared with whites (31.6%) or Mexican Americans (34.9%).

Based on the results of the regression analyses, compared with participants who did not have hypertriglyceridemia (≥150 mg/dL), those who did were older; less likely to be African American than white; more likely to have completed high school than to have pursued education beyond high school; more likely to be a current smoker than a never smoker and be overweight or obese; and more likely to report having diabetes and using a TG concentration–lowering medication, statins, or ezetimibe (Table 2). Hypertriglyceridemia was not significantly associated with insurance status, physical activity, alcohol use, or intake of carbohydrates. When we defined hypertriglyceridemia as a TG concentration of 200 mg/dL or higher, the associations were for the most part similar. However, participants who reported engaging in less than 150 minutes of physical activity per week had a higher prevalence of hypertriglyceridemia than those who did 150 minutes or more per week, and participants consuming any alcohol were less likely to have hypertriglyceridemia. In addition, the use of statins or ezetimibe was not significantly associated with hypertriglyceridemia.

We also examined the percentages (SEs) of participants with elevated total cholesterol concentrations (≥200 mg/dL) or low high-density lipoprotein cholesterol concentrations (men, <40 mg/dL, and women, <50 mg/dL) by TG concentration. Overall, 47.7% (1.0%) of participants had a total cholesterol concentration of 200 mg/dL or higher; this prevalence was 39.1% (1.2%) among participants with a TG concentration lower than 150 mg/dL, 58.6% (2.2%) among those with a TG concentration of 150 to lower than 200 mg/dL, and 70.5% (1.6%) among those with a TG concentration of 200 mg/dL or higher. A low-density lipoprotein cholesterol concentration of 130 mg/dL or higher was present among 37.1% (1.0%) of all participants; this prevalence was 33.6 (1.1%) among those with a TG concentration lower than 150 mg/dL, 45.8% (2.3%) among those with a TG concentration of 150 to lower than 200 mg/dL, and 43.6% (1.8%) among those with a TG concentration of 200 to lower than 400 mg/dL. Furthermore, 33.8% (0.9%) of all participants had a low high-density lipoprotein cholesterol concentration; this prevalence was 23.8% (1.0%) among those with a TG concentration of lower than 150 mg/dL, 44.2% (2.5%) among those with a TG concentration of 150 to lower than 200 mg/dL, and 62.1% (1.9%) among those with a TG concentration of 200 mg/dL or higher. (To convert cholesterols to millimoles per liter, multiply by 0.0259.)

In addition, we calculated the percentage (SE) of 5610 participants who had an elevated total cholesterol concentration, low high-density lipoprotein cholesterol concentration (men, <40 mg/dL, and women, <50 mg/dL), and hypertriglyceridemia (≥150 mg/dL). When only considering lipid values, 10.5% (0.4%) of participants had all 3 lipid abnormalities. When considering the use of cholesterol concentration–lowering medications (self-reported) as also indicating hypercholesterolemia and the use of TG concentration–lowering medications (ie, fenofibrate, gemfibrozil, and niacin) as also indicating hypertriglyceridemia, 11.9% (0.4%) of participants had all 3 abnormalities.

Overall, the percentage (SE) of participants who used one of the prescription medications used to lower TG concentration (ie, fenofibrate, gemfibrozil, or niacin) during 1999 to 2004 was 1.3% (0.2%). Only 2.6% (0.4%) of participants with a TG concentration of 150 mg/dL or higher and 3.6% (0.7%) of participants with a TG concentration of 200 mg/dL or higher reported having prescription medications. The small number of participants who used such medications prevented a detailed analysis of correlates of such use. Among those using TG concentration–lowering medications, 32.8% (6.6%) had a TG concentration lower than 150 mg/dL (41.0% [8.3%] among men and 11.3% [7.9%] among women).

The percentage of all participants who used one of the TG concentration–lowering medications increased during successive 2-year cycles: 0.6% during 1999-2000, 1.3% during 2001-2002, and 1.8% during 2003-2004 (P value for trend, .005). Among participants with a TG concentration of 150 mg/dL or higher, the percentages were 1.7%, 2.4%, and 3.5%, respectively (P value for trend, .10). Among participants with a TG concentration of 200 mg/dL or higher, the percentages were 1.4%, 3.6%, and 5.1%, respectively (P value for trend, .02).

Because cholesterol concentration–lowering medications such as statins can also lower TG concentration, we examined the percentage of participants with elevated TG concentrations who were also taking cholesterol concentration–lowering medications. Overall, the percentage (SE) of 5605 participants who reported using a statin or ezetimibe during 1999 to 2004 was 9.8% (0.6%); this percentage was 7.0% (0.6) among participants with a TG concentration of lower than 150 mg/dL, 16.0% (1.4) among participants with a TG concentration of 150 to lower than 200 mg/dL, and 15.2% (1.5) among participants with a concentration of 200 mg/dL or higher.

As evidence about the pathophysiologic role of TGs in cardiovascular disease continues to accumulate, an estimate of the magnitude of hypertriglyceridemia in the US population is helpful in understanding the size of the population that may require management of their hyperlipidemia, especially in view of the escalating prevalence of obesity, a cause of elevated TG concentrations, in the United States. Almost 1 in 3 US adults had a TG concentration of 150 mg/dL or higher, and almost 1 in 5 US adults had a TG concentration of 200 mg/dL or higher. Establishing levels of treatment helps in securing a foundation against which to measure future change in the pharmacologic treatment of hypertriglyceridemia. During 1999 to 2004, the use of 3 agents used to treat hypertriglyceridemia was lower than 4% among people who had a TG concentration of 200 mg/dL or higher.

An important approach to the patient with hypertriglyceridemia is the implementation of nonpharmacological interventions. The high percentages of participants with a TG concentration of 200 mg/dL or higher who were overweight or obese, who failed to engage in physical activity for approximately 150 minutes per week, or who smoked attest to the challenge confronting health care providers in working with their patients to implement therapeutic lifestyle changes.

The low percentage of people with hypertriglyceridemia receiving pharmacologic treatment for this abnormality perhaps reflects the uncertainty about the need to treat elevated TG concentrations. When concentrations are exceedingly high (≥500 mg/dL), thus increasing a patient's risk for pancreatitis, the need for both nonpharmacologic and pharmacologic treatment is clear. For patients with a TG concentration of 150 to lower than 200 mg/dL, the 2001 NCEP report recommended therapeutic lifestyle change as the treatment of choice and did not see a need for pharmacologic management.⁷ For TG concentrations in the range of 200 to lower then 500 mg/dL, the 2001 NCEP report recommended therapeutic lifestyle change as the initial therapy.⁷ If therapeutic lifestyle change fails to achieve the desired results, the use of pharmacologic agents is an option. The 2001 NCEP report set no specific treatment goals for TG concentration as it does for low-density lipoprotein cholesterol concentration. Rather, it expressed its therapeutic goals in terms of non–high-density lipoprotein cholesterol target concentrations.

In a review of 17 randomized controlled trials, the use of fibrates conferred no benefit on all-cause mortality (relative risk [RR], 1.00; 95% confidence interval [CI], 0.91-1.11).⁹ Of concern was the increased risk of dying from a cause other than cardiovascular disease among those using fibrates (RR, 1.13; 95% CI, 1.01-1.27). However, the authors did not provide results for the different fibrates, and only 1 of the 17 trials used fenofibrate. In 3 trials for the primary prevention of coronary heart disease, the group using fibrates experienced increased all-cause mortality (RR, 1.25; 95% CI, 1.05-1.48). On the basis of 2 trials examining the effect of niacin on the secondary prevention of coronary heart disease, there was little difference in all-cause mortality among participants in the treatment and placebo arms (RR, 0.96; 95% CI, 0.86-1.08).⁹

A couple of trials have tested the use of fenofibrate on cardiovascular disease outcomes. In the Diabetes Atherosclerosis Intervention Study, diabetic participants assigned to the fenofibrate group had less progression of coronary artery disease than those assigned to placebo.¹⁰ The study was not sufficiently powered to study clinical end points. The rate of all-cause mortality was 2.9% in the treatment group and 4.3% in the placebo group. In the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study, diabetic participants in the treatment group had a nonsignificant 11% reduction in the primary end point of first myocardial infarction or coronary heart disease.¹¹ Treatment favored several secondary outcomes such as nonfatal myocardial infarction (hazard ratio [HR], 0.76; 95% CI, 0.62-0.94), total cardiovascular disease events (HR, 0.89; 95% CI, 0.80-0.99), coronary revascularization (HR, 0.79; 95% CI, 0.68-0.93), and all revascularizations (HR, 0.80; 95% CI, 0.70-0.92). Total mortality, however, was increased by a nonsignificant 11% in the treatment group (HR, 1.11; 95% CI, 0.95-1.29).

Gemfibrozil has also been the subject of study in trials. In the Helsinki Heart Study, asymptomatic men with dyslipidemia who were treated with gemfibrozil experienced a significant 34% reduction in the incidence of fatal and nonfatal myocardial infarction and cardiac death.¹² All-cause mortality did not differ significantly between the treatment group (21.9%) and the placebo group (20.7%). In the Veterans Affairs Cooperative Studies Program High-Density Lipoprotein Cholesterol Intervention Trial (VA-HIT), participants with coronary heart disease who were treated with gemfibrozil experienced a significant 22% reduction in nonfatal myocardial infarction or death from coronary heart disease.¹³ All-cause mortality was lower by a nonsignificant 11%.

Measurements of TG concentrations in the present study were determined in fasting serum specimens. However, several studies have pointed out the potential value of nonfasting measurements of TG concentrations.^14- 16 Hopefully, future work will sort out the utility of fasting and nonfasting TG concentration measurements in the prediction of risk for cardiovascular disease.

The 3 medications included in the present study—fenofibrate, gemfibrozil, and niacin—are also used to raise high-density lipoprotein cholesterol concentrations. Therefore, we cannot be certain that the intended use was for lowering elevated TG concentrations rather than raising high-density lipoprotein cholesterol concentrations. If their use was indicated for the latter reason, then the percentage of pharmacologic treatment for hypertriglyceridemia would be lower than our reported estimate.

Our analyses indicated that women had a lower prevalence of hypertriglyceridemia than men. The use of oral contraceptives and hormone therapy can increase TG concentrations, although the effects may vary by route of administration and composition of the medications.^17- 18 Thus, assuming that the use of oral contraceptives and hormone therapy raised TG concentrations in the women included in our analyses, the true difference in the prevalence of hypertriglyceridemia between women and men was likely even larger than what we noted. The potential health risks from the elevation in TG concentrations attributable to exogenous sources of estrogens and progestagens among women remain uncertain.

Our results showed that the prevalence of hypertriglyceridemia was substantially lower among African Americans than among whites or Mexican Americans. These findings are consistent with those from other studies.^19- 25 Compared with whites, a smaller amount of visceral adipose tissue among African Americans for a given level of total body fat, an increased activity of postheparin lipoprotein lipase among African Americans, and a more favorable postprandial TG response among African Americans may at least partially explain why African Americans enjoy more favorable TG concentrations compared with whites and other ethnicities.^26- 27 Limited data are available on the relationship between TG concentrations and cardiovascular disease outcomes from prospective studies among African Americans. In at least 1 study, TG concentrations were equally predictive of mortality from cardiovascular disease among African Americans and whites.²⁸ Assuming that TGs are an equally potent risk factor for cardiovascular disease among African Americans as among other ethnicities, lower TG concentrations among African Americans suggest that the population-attributable risk fraction for cardiovascular disease is less among African American than among other ethnicities.

In conclusion, the prevalence of hypertriglyceridemia is high among US adults, the use of pharmacologic treatment is low, and the prevalence of modifiable causes of hypertriglyceridemia, such as physical inactivity and overweight or obesity, is high. Because measuring TG concentrations is routinely performed in clinical practice, physicians have to regularly decide on the need for treatment in many of their patients. As research clarifies uncertainties in the relation between TG concentration and cardiovascular disease, guidelines to treat hypertriglyceridemia will likely be modified.

Correspondence: Earl S. Ford, MD, MPH, Division of Adult and Community Health, Centers for Disease Control and Prevention, 4770 Buford Hwy, MS K66, Atlanta, GA 30341 (eford@cdc.gov).

Accepted for Publication: October 12, 2008.

Author Contributions: Dr Ford had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Ford and Mokdad. Analysis and interpretation of data: Ford, Li, Zhao, and Pearson. Drafting of the manuscript: Ford. Critical revision of the manuscript for important intellectual content: Ford, Li, Zhao, Pearson, and Mokdad. Statistical analysis: Ford, Li, and Pearson. Study supervision: Ford and Mokdad.

Disclaimer: The findings and conclusions in this article are those of the authors and do not represent the official position of the Centers for Disease Control and Prevention.

Financial Disclosure: None reported.