If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Association of High-Density Lipoprotein Cholesterol With Coronary Heart Disease Risk Across Categories of Low-Density Lipoprotein Cholesterol: The Atherosclerosis Risk in Communities Study
Department of Epidemiology, Ryals School of Public Health, University of Alabama at Birmingham, 1665 University Boulevard, Suite 230J, Birmingham, AL 35205.
National cholesterol treatment guidelines include a low level of high-density lipoprotein (HDL) cholesterol (<40 mg/dL) as a major risk factor for coronary heart disease (CHD) that should be considered when making decisions on treatment of low- density lipoprotein (LDL) cholesterol.
Methods
We investigated the association of HDL and LDL-cholesterol with incident CHD events (fatal or nonfatal CHD) over 14 years of follow-up among 13,615 adults aged 45 to 64 years in the Atherosclerosis Risk in Communities study.
Results
A total of 966 (7.1%) participants had a CHD event during follow-up. After adjustment for age, race, sex, diabetes, smoking, alcohol consumption, systolic blood pressure, waist circumference, chronic kidney disease and physical activity, a graded association was present between progressively lower levels of HDL-cholesterol and higher CHD risk, overall (P < 0.001) and within each level of LDL-cholesterol (<100, 100–129, 130–159, 160–189 and ≥190 mg/ dL) investigated (all P < 0.05). In addition, after multivariable adjustment including LDL-cholesterol, each standard deviation higher HDL- cholesterol (18 mg/dL) was associated with a hazard ratio of incident CHD of 0.70 (95% Cl: 0.63–0.77).
Conclusions
These data suggest a graded association exists between lower levels of HDL-cholesterol and CHD across the full range of LDL-cholesterol levels. As interventions targeting HDL levels are developed, the combinatorial effects of lower HDL levels with various levels of LDL-cholesterol should be examined.
The Third report of the National Cholesterol Education Program Adult Treatment Panel (ATP-III) recommends low-density lipoprotein (LDL) cholesterol as the primary target of clinical lipid management and reducing coronary heart disease (CHD) risk.
Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults
Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III).
In addition, the ATP-III guidelines recommend initial pharmacologic treatment for high LDL-cholesterol includes treatment with 3-hydroxy-3-methylglutaryl-co- enzyme A reductase inhibitors (ie, statins). Although this recommendation is justified by extensive evidence, many CHD events still occur among patients treated with statin therapy.
Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/TexCAPS. Air Force/ Texas Coronary Atherosclerosis Prevention Study.
Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: the Atherosclerosis Risk in Communities (ARIC) Stud.
High density lipoprotein cholesterol and myocardial infarction or sudden coronary death: a prospective case-control study in middle-aged men of the Oslo study.
Relation of high-density lipoprotein cholesterol and triglycerides to incidence of atherosclerotic coronary artery disease (the PROCAM experience). Prospective Cardiovascular Munster study.
However, HDL-cholesterol levels are correlated with LDL-cholesterol, and little evidence on the association of HDL-cholesterol, independent of LDL-cholesterol, on CHD risk is available. As several new treatments are being developed to raise HDL-cholesterol, it is important to understand association of lower levels of HDL-cholesterol with CHD risk across the range of LDL-cholesterol levels.
Therefore, we investigated the association of HDL-cholesterol with CHD events over 14 years of follow-up in the Atherosclerosis Risk in Communities (ARIC) study.
METHODS
Study Population and Data Collection
The ARIC study, a population-based prospective cohort study of atherosclerosis and its risk factors, has been described in detail previously.
In brief, between 1986 and 1989, 15,792 study participants aged 45 to 64 years were enrolled from 4 U.S. communities (Forsyth County, North Carolina; Jackson, Mississippi; the northwest suburbs of Minneapolis, Minnesota; and Washington County, Maryland). Of permanent residents in the 4 study areas, potential participants were excluded if they were deemed, in the judgment of the interviewer, physically or mentally incapable of full participation in the study or were planning on permanently relocating. A baseline and 3 follow-up clinic examinations were conducted at 3-year intervals. The current analyses were limited to participants with valid covariate information who were not taking cholesterol-lowering medications at baseline and fasted for ≥12 hours before the baseline visit and had valid HDL and LDL-cholesterol measurements. Participants with a history of CHD at baseline [history of myocardial infarction (MI) based on self-report or electrocardiogram evidence at the baseline study visit, previous heart or arterial surgery, coronary bypass, or balloon angioplasty] or with triglycerides ≥400 mg/dL were excluded from all analyses. After these exclusions, the final study population included 13,615 ARIC participants.
Clinical Examinations
Study visits were conducted in the morning, and participants were asked to fast 12 hours before the blood was drawn, and actual fasting times were recorded. Blood was drawn from an antecubital vein of seated participants, serum was centrifuged, and frozen samples were shipped to central laboratories for analysis. Of relevance to the current analysis, baseline (1986–1989) total cholesterol and triglycerides were measured by enzymatic methods, and total HDL-cholesterol was measured after dextran-magnesium precipitation. LDL-cholesterol was calculated using the Friedewald equation. Aliquots from 7% of all samples were stored an additional week. Analysis of these “blind” duplicates provided a measure of variability that includes processing, storage and shipping effects. Blind duplicate coefficients of variation for LDL-cholesterol and HDL- cholesterol were 5% and 7%, respectively. Reliability estimates, based on an intraindividual variability study with 3 measurements taken 1- to 2-week intervals, were excellent for total cholesterol (0.94) and HDL-cholesterol (0.94). Glucose was measured by the hexokinase/glucose-6 phosphate dehydrogenase method. Diabetes mellitus was defined as a fasting glucose of ≥126 mg/dL, a self-reported history of diabetes or a use of glucose-lowering medications. Glomerular filtration rate was estimated using a formula derived by the Modification of Diet in Renal Disease, and chronic kidney disease was defined as <60 mL/min/1.73 m2.
ARIC technicians trained and certified in the use of a random-zero sphygmomanometer took 3 blood pressure measurements according to a standardized protocol; an average of the second and third measurements was used to estimate blood pressure. The presence of hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or use of antihypertensive medications. Trained technicians measured waist circumference to the nearest centimeter at the umbilical level following a standardized protocol. Abdominal obesity was defined as a waist circumference >88 and >102 cm in women and men, respectively. Family history of CHD, current cigarette smoking and physical activity were determined through the use of standardized questionnaires. Participants who reported having smoked >400 cigarettes during their lifetime and responded affirmatively to “Do you now smoke cigarettes?” were classified as current smokers. Physical activity was defined as participating in ≥1 hours of sports per week for ≥10 months during the previous year. All participants had a standard 12-lead electrocardiogram at baseline.
Outcome Definition and Assessment
The primary outcome for this analysis was the incidence of CHD from the baseline ARIC visit through December 31, 2002. Several methods were used to ascertain incident CHD events among ARIC participants. Participants were contacted annually through telephone to identify all hospitalizations or deaths or both. ARIC study staff members also surveyed death certificates and discharge lists from local hospitals to identify additional CHD events. For hospitalizations of ARIC participants, the signs and symptoms at presentation and related clinical information were abstracted from charts by trained and certified study staff. Out of hospital deaths were validated using death certificate data and, when possible, interviews with next of kin and the participant’s physician. When available, autopsy reports were used for further validation. For the current analysis, CHD incidence was defined as a definite or probable MI or a definite CHD death. An ARIC Morbidity and Mortality Classification Committee used published criteria to review and adjudicate all potential CHD events.
Statistical Analysis
Baseline characteristics of the ARIC study population were calculated, overall and by level of HDL-cholesterol (<40, 40– 49, 50–59 and ≥60 mg/dL), as means, for continuous variables and, prevalences, for dichotomous variables. The statistical significance of trends in continuous and dichotomous characteristics across category of HDL-cholesterol was determined using linear and logistic regression models, respectively. The Pearson correlation coefficient between HDL-cholesterol and LDL-cholesterol was calculated. Cumulative CHD incidence during follow-up, for each level of LDL-cholesterol (<100, 100–129, 130–159, 160–189 and ≥190 mg/dL) separately, was graphed by HDL-cholesterol level. For each LDL- cholesterol level, cumulative CHD incidence was compared across HDL-cholesterol levels using log-rank tests for trends. Next, incidence rates and the multivariable-adjusted hazard ratio of CHD events associated with level of HDL-cholesterol, overall and stratified by level of LDL-cholesterol were calculated using Cox-proportional hazards regression models. Initial Cox regression models included adjustment for age, race and sex with subsequent models including additional adjustment for ARIC field center, current and former cigarette smoking, current alcohol consumption, physical activity, waist circumference, systolic blood pressure, diabetes mellitus and chronic kidney disease. The combined effects of levels of HDL-cholesterol and LDL-cholesterol were determined by classifying participants into 1 of 20 HDL-LDL groupings (eg, HDL-cholesterol 50-59 mg/dL and LDL-cholesterol 100-129 mg/dL) and calculating each groups’ multivariable-adjusted hazard ratio of CHD with participants having an HDL-cholesterol ≥60 mg/dL and LDL-cholesterol <100 mg/dL serving as the reference.
Next, the multivariable-adjusted hazard ratio of CHD associated with 1 standard deviation (SD) higher HDL-cholesterol (18 mg/dL) and LDL-cholesterol (39 mg/dL) was calculated. This analysis included HDL-cholesterol and LDL-cholesterol simultaneously in a Cox regression model and was performed for the overall population and subgroups defined by demographics, physical activity, cigarette smoking, alcohol consumption, abdominal obesity, hypertension, diabetes and chronic kidney disease.
The proportionality assumption of the Cox model was confirmed using Schoenfeld residuals. All P-values are 2 sided and <0.05 was considered statistically significant. All data management and analysis were conducted using SAS 8.1 (Cary, NC).
RESULTS
ARIC participants with lower HDL-cholesterol were more likely to be male and white (Table 1). In addition, current smokers, persons with hypertension, diabetes, chronic kidney disease and prevalent CHD were more likely to have lower HDL-cholesterol, whereas individuals who consumed alcohol had higher HDL-cholesterol levels. Higher systolic blood pressure, waist circumference, serum triglycerides and LDL-cholesterol were associated with progressively lower HDL-cholesterol. In contrast, higher HDL-cholesterol was associated with higher total cholesterol. The correlation between HDL- and LDL-cholesterols was –0.22 (P < 0.001).
Table 1Baseline characteristics of the ARIC study population by HDL-cholesterol level
Lower HDL-cholesterol was associated with a higher CHD incidence within each level of LDL-cholesterol (Figure 1). This association remained present after adjustment for age, race and sex (Table 2). Although somewhat attenuated after additional multivariable adjustment, a graded association was present between lower HDL-cholesterol levels and higher hazard ratios of CHD incidence (Table 3). This association was present overall and within each LDL-cholesterol level.
Figure 1Cumulative incidence of CHD associated with HDL-cholesterol stratified by level of LDL-cholesterol
Table 2Incidence rates and age-, race-, sex-adjusted hazard ratio of CHD events associated with HDL-cholesterol, overall and stratified by LDL-cholesterol
Table 3Multivariable-adjusted hazard ratio of CHD associated with HDL-cholesterol by level of LDL-cholesterol
HDL-cholesterol (mg/dL)
<40
40–49
50–59
≥60
P-trend
Overall
2.28 (1.84-2.83)
1.74 (1.41-2.15)
1.65 (1.33-2.05)
1 (ref)
<0.001
LDL-cholesterol level (mg/dL) <100
2.43 (1.31–4.52)
1.85 (1.00–3.42)
1.44 (0.75–2.74)
1 (ref)
<0.001
100–129
2.20 (1.42-3.40)
1.79 (1.18-2.74)
1.84 (1.21-2.80)
1 (ref)
<0.001
130–159
1.69 (1.12-2.55)
1.50 (1.01–2.21)
1.43 (0.95-2.14)
1 (ref)
0.020
160–189
2.77 (1.62–4.73)
1.36 (0.79–2.35)
1.76 (1.02–3.03)
1 (ref)
<0.001
≥190
1.65 (0.92-2.95)
1.50 (0.86-2.59)
1.05 (0.58-1.90)
1 (ref)
0.038
Numbers in table represent hazard ratio (95% CI), adjusted for age, race, sex, ARIC field center, current and former cigarette smoking, alcohol consumption, physical activity, waist circumference, systolic blood pressure, diabetes mellitus and chronic kidney disease.
Similar results were present when the hazard ratios of CHD incidence were calculated by HDL and LDL-cholesterol jointly (Figure 2). Specifically, within each LDL- cholesterol category, the hazard ratios of CHD increased at progressively lower HDL-cholesterol. Moreover, a higher CHD incidence was present at higher LDL-cholesterol within each HDL-cholesterol grouping. Overall, compared with their counterparts with an HDL-cholesterol ≥60 mg/dL and LDL-cholesterol <100 mg/dL, ARIC participants with an HDL- cholesterol <40 mg/dL and LDL-cholesterol ≥190 mg/dL were 4.38 (95% CI: 2.71–7.07) times more likely to have a CHD event during follow-up.
Figure 2Adjusted relative hazard of CHD (defined as MI or fatal CHD, adjusted for age, race, sex, ARIC field center, current and former cigarette smoking, alcohol consumption, physical activity, waist circumference, systolic blood pressure, diabetes mellitus and chronic kidney disease) associated with level of HDL-cholesterol and LDL-cholesterol.
After multivariable adjustment, including LDL-cholesterol, each SD higher HDL-cholesterol (18 mg/dL) was associated with a hazard ratio of incident CHD of 0.70 (95% CI: 0.63–0.77; Figure 3). Similarly, each SD higher LDL- cholesterol (39 mg/dL) was associated with a hazard ratio of 1.29 (95% CI: 1.21–1.38). These associations remained markedly similar for all subgroups investigated with the exception of participants with chronic kidney disease.
Figure 3Adjusted hazard ratio of CHD associated with 1 SD higher HDL-cholesterol (18 mg/dL) and 1 SD higher LDL-cholesterol (39 mg/dL) within subgroups of the ARIC study population. HR, hazard ratio (represented by the square); CI represented by the bar/arrow. Models for subgroup analyses included adjustment for age, race, sex, physical activity, smoking status, alcohol consumption, systolic blood pressure, waist circumference, diabetes mellitus and chronic kidney disease as appropriate (eg, the model for individuals without chronic kidney disease did not include adjustment for chronic kidney disease).
The goal of the current study was to assess CHD risk associated with HDL-cholesterol across levels of LDL-cholesterol. Overall, a strong and graded association was present. Importantly, significant trends of increased CHD risk associated with progressively lower HDL-cholesterol were present within each level of LDL-cholesterol investigated. Compared with their counterparts with HDL-cholesterol levels ≥ 60 mg/dL, at LDL-cholesterol levels of <100, 100 to 129, 130 to 159, 160 to 189 and ≥190 mg/dL, ARIC participants with HDL-cholesterol <40 mg/dL were 2.43, 2.20, 1.69, 2.77 and 1.65 times more likely to have a CHD event during follow-up, respectively. These results highlight the impact of low HDL-cholesterol, independent of LDL- cholesterol level, on CHD risk.
Low HDL-cholesterol has been reported to be the most common lipid abnormality in patients with known CHD.
In 1 study, low HDL-cholesterol was reported in approximately 50% of patients with CHD. Furthermore, data from the general U.S. population collected in the National Health and Nutrition Examination 1999-2002 indicate low HDL-cholesterol (HDL-cholesterol <40 mg/dL) is more common among U.S. adults (22.2%) than LDL-cholesterol ≥160 mg/dL (13.8%) and total cholesterol ≥240 mg/dL (15.1%).
Therefore, at the population level, identifying safe and effective ways to raise HDL-cholesterol may have an important public health impact.
Despite the aggressive treatment of LDL-cholesterol, CHD remains the leading cause of death in the United States. Although LDL-cholesterol is an important target in CHD prevention, 40% of CHD events occur among persons with LDL-cholesterol <130 mg/dL. As demonstrated in the current study, HDL-cholesterol remains a predictor of CHD incidence for persons with LDL-cholesterol in this range. Given its independent association with CHD incidence, raising HDL-cholesterol provides an additional target for CHD prevention therapy.
Low HDL-cholesterol was proposed initially as a risk factor for CHD in the 1950s.
High density lipoprotein cholesterol and myocardial infarction or sudden coronary death: a prospective case-control study in middle-aged men of the Oslo study.
Relation of high-density lipoprotein cholesterol and triglycerides to incidence of atherosclerotic coronary artery disease (the PROCAM experience). Prospective Cardiovascular Munster study.
analyzed data from 5 large longitudinal studies and found a consistent inverse relation between HDL-cholesterol and CHD risk. These results have been confirmed in many observational studies. In addition, results from the Department of Veterans Affairs’ HDL- cholesterol Intervention Trial demonstrated the benefit of the fibrate therapy gemfibrozil in the secondary prevention of CHD events.
Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of high- density lipoprotein cholesterol. Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial Study Grou.
Although patients randomized to gemfibrozil did not experience a change in LDL-cholesterol, they experienced a statistically significant 6% increase in HDL-cholesterol and 31% reduction in serum triglycerides. More importantly, compared with their counterparts randomized to receive a placebo, patients who were randomized to gemfibrozil experienced a 22% reduction in the primary outcome of nonfatal MI or death caused by CHD. In this study, the increase in HDL-cholesterol, but not the reduction in serum triglycerides, was associated with the reduction in CHD. These results are consistent with subgroup analyses of the Helsinki Heart Study, a primary CHD prevention trial.
It is well accepted that women and blacks have higher HDL-cholesterol than men and whites, respectively.
Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: the Atherosclerosis Risk in Communities (ARIC) Stud.
High density lipoprotein cholesterol levels among US adults by selected demographic and socioeconomic variables. The Second National Health and Nutrition Examination Survey 1976-1980.
High density lipoprotein cholesterol levels among US adults by selected demographic and socioeconomic variables. The Second National Health and Nutrition Examination Survey 1976-1980.
These include overweight, physical inactivity and cigarette smoking. Therapeutic lifestyle changes including weight loss, increased physical activity and, for current smokers, smoking cessation have been demonstrated to increase HDL-cholesterol.
Changes in body weight and serum lipid profile in obese patients treated with orlistat in addition to a hypocaloric diet: a systematic review of randomized clinical trials.
Coronary atherosclerosis has been prevented in animals expressing high apolipoprotein A-I (apoA-I) through oral administration or somatic gene transfer of human apoA-I. Although not entirely known, 1 proposed mechanism is reverse cholesterol transport. This occurs whereby apoA-I removes excess cholesterol from macrophages in the atherosclerotic plaque. This cholesterol is returned to the liver for excretion in bile. Although the administration of cholesteryl ester transfer protein inhibition has been showed to increase HDL-cholesterol, data on its impact on CHD events, per se, have not been positive. Specifically, the Investigation of Lipid Level Management to Understand its Impact in Atherosclerotic Events (ILLUMINATE) trial was halted early because of adverse outcomes among patients randomized to receive torcetrapib, an cholesteryl ester transfer protein inhibitor, and atorvastatin compared with their counterparts who were randomized to receive atorvastatin alone. Although future trials will ultimately determine the benefit of cholesterol ester transport protein inhibitors, HDL-cholesterol has also been proposed to have other antiatherosclerosis properties. These include being an antioxidant, anti-inflammatory, nitric oxide promoter and platelet inhibitor.
In Cox proportional hazards models including adjustment for LDL-cholesterol, each SD higher HDL-cholesterol (18 mg/dL) was associated with a 0.70 (95% CI: 0.63-0.77) risk of CHD. In previous studies, which generally have not reported adjustment for LDL-cholesterol, each 1 mg/dL higher HDL-cholesterol is associated with a 2% to 3% lower CHD risk. The current results are consistent with the previous findings. A hazard ratio of 0.70 per each 18 mg/dL higher HDL-cholesterol translates into a 2% lower CHD risk for each 1 mg/dL higher HDL-cholesterol [ie, this is calculated as ln(0.70)/18]. Therefore, on this basis, it can be inferred that much of the CHD risk associated with HDL- cholesterol is not explained by LDL-cholesterol.
Although the current study provides important data on the association of HDL-cholesterol with CHD incidence after controlling for LDL-cholesterol, it needs to be interpreted in the context of its limitations. The most notable limitation is the use of a single HDL-cholesterol measurement. Although remeasuring HDL-cholesterol and using the average of 2 measurements was not feasible because of the regression to the mean, the association between HDL-cholesterol based on single measurement and CHD incidence reported is likely underestimated. As with all large epidemiological studies, the current analysis only included events for which medical charts were retrieved and outcomes adjudicated. Despite these limitations, the current study had several strengths. These include a large sample size that included blacks and whites. This permitted the assessment of CHD risk associated with HDL-cholesterol by level of LDL-cholesterol and within important population subgroups. Additional strengths include the longitudinal study design with up to 14 years of follow-up with rigorous participant tracking and events ascertainment. Specifically, <5% of the study population was not traced, and events were confirmed by a committee that was blinded to participants’ study measurements.
In summary, the current study demonstrates a strong association between HDL-cholesterol and CHD events regardless of LDL-cholesterol levels. This association was markedly consistent in the present analyses. Specifically, the association was present within categories of LDL-cholesterol, after statistical adjustment for LDL-cholesterol, and when HDL-cholesterol was modeled as a categorical and as a continuous variable. In addition, even after multivariable adjustment including LDL-cholesterol, an inverse relationship was present in all subgroups, with the exception of a small subgroup of ARIC participants with chronic kidney disease (n = 129). Based on the findings of the current study, raising HDL-cholesterol may have important cardioprotective benefits. Therapeutic lifestyle changes should be strongly encouraged, and drug therapy should be considered for all adults with low HDL-cholesterol to increase HDL- cholesterol and lower CHD risk.
REFERENCES
Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults
Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III).
Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/TexCAPS. Air Force/ Texas Coronary Atherosclerosis Prevention Study.
Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: the Atherosclerosis Risk in Communities (ARIC) Stud.
High density lipoprotein cholesterol and myocardial infarction or sudden coronary death: a prospective case-control study in middle-aged men of the Oslo study.
Relation of high-density lipoprotein cholesterol and triglycerides to incidence of atherosclerotic coronary artery disease (the PROCAM experience). Prospective Cardiovascular Munster study.
Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of high- density lipoprotein cholesterol. Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial Study Grou.
High density lipoprotein cholesterol levels among US adults by selected demographic and socioeconomic variables. The Second National Health and Nutrition Examination Survey 1976-1980.
Changes in body weight and serum lipid profile in obese patients treated with orlistat in addition to a hypocaloric diet: a systematic review of randomized clinical trials.
Pfize did not have any role in the design and conduct of the study, the collection, the management, the data analysis, or the interpretation of the data, or the preparation or approval of the article.
This study was supported by NHLBI grants N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021 and N01-HC-55022 and by Pfize grant (to PM, BCA).